ABSTRACT
Title of dissertation: THE INFLUENCE OF MATERNAL SENSITIVITY
AND MATERNAL STIMULATION ON LATER
DEVELOPMENT OF EXECUTIVE FUNCTIONING
VIA STRUCTURAL EQUATION MODELING
Jessica Elaine Emick, Doctor of Philosophy, 2007
Dissertation directed by: Professor William Strein
Department of Counseling and Personnel Services
This study investigated the relations between early maternal behaviors,
maternal sensitivity and maternal stimulation, and the later development of executive
function. It was hypothesized that maternal behaviors could influence the
development of executive function either directly or indirectly by influencing a
child?s language or attentional abilities. This study attempted to model these
relationships using archival data from phase I and phase II from the Study of Early
Child Care and Youth Development (SECCYD). Structural equation modeling was
used with data from 470 participants on measures of SES, maternal sensitivity,
maternal stimulation, language, attention, and executive function. From existing
literature three nested models were proposed to examine how maternal behaviors
influenced the later development of executive function. While there were significant
differences between the three proposed models it is important to recognize the overall
poor fit of the models. The differences between the models suggest that maternal
sensitivity and maternal stimulation do not directly influence executive functioning in
the 1
st
grade but instead influence the development of executive functioning through
assisting the child in development of attention and language skills. Interestingly, the
model also indicated verbal ability played an important role in the development of
executive function. Secondly the study attempted to examine multi-group differences
in the proposed models (Caucasian and African American). While small sample size
precluded this analysis, examining the effect size differences between the two groups
indicated that within the current sample ethnicity, language ability, and SES are
deeply entangled. The results of the current study highlight the potential role of
language ability in the development of executive function and the need for cleaner
measures of executive function that are developmentally appropriate.
THE INFLUENCE OF MATERNAL SENSITIVITY AND MATERNAL
STIMULATION ON LATER DEVELOPMENT OF EXECUTIVE FUNCTIONING
VIA STRUCTURAL EQUATION MODELING
by
Jessica Elaine Emick
Dissertation submitted to the Faculty of the Graduate School of the
University of Maryland, College Park in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
2007
Advisory Committee:
Professor William Strein, Chair
Professor Alfred Amado
Professor Natasha Cabrera
Professor Ellin Scholnick
Professor Marilyn Welsh, University of Northern Colorado
? Copyright by
Jessica Elaine Emick
2007
ii
Acknowledgments
I would like to recognize individuals who have made important contributions to this
dissertation:
First and foremost, I want to recognize the National Institute of Child Health and Human
Development for making the Study of Early Child Care and Youth Development
available to qualified researchers for scientific collaboration. Secondly, I want to
acknowledge my dissertation committee for their thoughtful comments and support.
Finally, I want to thank Dr. Karen Samuelsen for her statistical support and for opening
her door for my many questions.
I would like to recognize several important contributors to my research career:
Dr. Robert Haaf, University of Toledo, for introducing me to developmental research and
making my first research experience so positive. Dr. Marilyn Welsh, University of
Northern Colorado, for sharing her immense knowledge about the executive functioning
and acting as sounding board for many ideas. Dr. William Strein, University of
Maryland, for introducing me to archival datasets and encouraging me to pursue my own
line of research. Dr. Rebecca Kopriva, University of Maryland, teaching me how to think
more creatively about research and approach research questions from different angles.
I would like to recognize the important individuals in my personnel life:
My nephews for making me laugh, my brother for making me smile, my parents for
providing me a strong sense of self, my sister and brother-in-law for opening their home
to me, my friends for letting me live vicariously through them, and Todd for his unfailing
support. I would not have wanted to do this without any of you!
Caleb, age 5, as his aunt was leaving to go the coffee shop to work on her dissertation, ?If
you want to work with children, why do you go always go the coffee shop and work? Are
there children at the coffee shop??
iii
Table of Contents
Acknowledgements?????????????????????...?????.ii
Table of Contents?????????????????????????..??.iii
List of Tables?????????????????????????????..iv
List of Figures?????????????????????????????.v
CHAPTER I........................................................................................................................ 1
Introduction..................................................................................................................... 1
Statement of Purpose .................................................................................................. 5
Research Questions..................................................................................................... 7
CHAPTER II?????????????????????????????....8
Literature Review............................................................................................................ 8
Overview of Mother-Child Interactions...................................................................... 8
Measuring Mother-Child Interactions...................................................................... 19
Overview of Executive Functions.............................................................................. 28
Measuring Executive Function ................................................................................. 35
Mother-Child Interactions and Executive Function ................................................. 46
CHAPTER III ................................................................................................................... 54
Methodology................................................................................................................. 54
Participants............................................................................................................... 54
Procedures and Materials......................................................................................... 58
Data Analysis............................................................................................................ 72
Current Models ......................................................................................................... 75
CHAPTER IV ................................................................................................................... 79
Results........................................................................................................................... 79
Descriptive Statistics................................................................................................. 79
Question 1 ................................................................................................................. 86
Question 2 ............................................................................................................... 108
Question 3 ............................................................................................................... 112
CHAPTER V .................................................................................................................. 117
Discussion................................................................................................................... 117
Question 1 ............................................................................................................... 118
Question 2 ............................................................................................................... 127
Socioeconomic Status and Executive Function....................................................... 128
Question 3 ............................................................................................................... 133
Limitations .............................................................................................................. 134
Future Research...................................................................................................... 136
Appendices????????????????????????????.......138
References?????????????????????????????....141
iv
LIST OF TABLES
TABLE
1. Demographic Variables for participants and families in the Current Study sample?57
2. Variables Identified for Inclusion in the Structural Equation Modeling?????.78
3. Descriptive Statistics and Effect Sizes for Indicator Variables??????.??..80
4. Descriptive Statistics and Effect Sizes for Indicator Variables from current study and
excluded data?????????????????????????????.83
5. Descriptive Statistics for Indicator Variables???????????????..85
6. Summary of Maximum Likelihood (and) Robust Model-Fit Statistics?????...88
7. Unstandardized, Standardized, and Significance Levels for Measurement Model.....89
8. Decomposition of Standardized Effects for Structural Paths of Indirect Model?.?96
9. Intercorrelations Among Measures of Maternal Sensitivity?????????..102
10. Intercorrelations Among measures of Maternal Stimulation?????????.103
11. Intercorrelations Among Language Measures???????...???????105
12. Intercorrelations Among measures of Attention??????????????106
13. Intercorrelations Among Executive Function Measures at 1
st
grade??????.107
14. Descriptive Statistics for Caucasian (N=427) and African American (N=26)?..?110
v
LIST OF FIGURES
FIGURE
1. Measurement Relationship between Maternal Behaviors, Executive Function,
Attention, and Language????...?????????????????92
2. No Relationship between Maternal Behaviors at 54 months and Executive
Function at 1
st
grade???????????????????????...94
3. Indirect Relationship between Maternal Behaviors at 54 months and Executive
Function at 1
st
grade???????..????????????????.98
4. Indirect and Direct Relationship between Maternal Behaviors at 54 months and
Executive Function at 1
st
grade???..??????????????.?100
5. Direct and Indirect Relationships via Maternal Verbal Ability between SES and
Executive Function????????..???????????????113
6. Proposed Relationship between Maternal Verbal Ability in place of
SES?????????????????????????????..115
7. Proposed Relationship between Indirect Relationship via Maternal Verbal ability
between SES and Executive Function??.??????????????116
8. Possible Model of Executive Function??...?????????????.136
1
1
The original NICHD SECCYD study was conducted by the NICHD Early Child
Care Research Network supported in part by NICHD and was made available to
qualified researchers for scientific collaboration. The author gratefully acknowledges
this collaboration.
The Influence of Maternal Sensitivity and Maternal Stimulation on Later
Development of Executive Functioning Via Structural Equation Modeling
1
CHAPTER I
Introduction
Research has frequently noted a relationship among cognitive, academic, and
social skills and parental factors (Stevens, Blake, Viatle, & Macdonald, 1998; Tamis-
LeMonda, Shannon, Cabrera, & Lamb, 2004). More specifically a foundation has
been laid that supports the impact of maternal influence on the future development of
problem solving skills (Freund, 1990; Hess & Shipman, 1965; Hubbs-Tait, Culp,
Culp, & Miller, 2002). Dodici, Draper, and Peterson (2003) note a range of factors
were related to future cognitive and social development; specifically they report
relationships between emotional tone, parental talk, engagement, parental guidance
style, parental responsivitity, and parental sensitivity. Although at times the strength
of that relationship has been challenged, Burchinal, Campbell, Bryant, Wasik, and
Ramey (1997) write that, ?the extent to which characteristics of the child and family
influence cognitive performance is still subject to debate because different analytic
methods yield varying results? (p. 935). Nevertheless, the debate of the influence of
socialization on cognitive outcomes has moved from questioning its existence to
examining the mechanisms and mediators of change (NICHD Early Child Care
Research Network, 2000).
2
Prior research has examined maternal influence during a collaborative task,
typically a joint play, and reported that maternal sensitivity and maternal stimulation
during a child?s infant and preschool years influences later cognitive abilities (NICHD
Early Child Care Research Network, 2005a). One such study, Landry, Smith, Swank, and
Miller-Loncar (2000), examined rural, low SES families (n=228) to see if children?s later
functioning was facilitated by certain maternal behaviors, namely maintaining children?s
focus, responsiveness, and directiveness, during both collaborative play and daily
activities. The authors reported that maternal interaction style, as observed when the
children were 24 months old during two interactions (one play interaction and one daily
activity interaction), indirectly influenced the cognitive skills of children at 4 ? years by
directly influencing cognitive and social skills at 2? and 3 ? years. The authors report,
?when parents identify critical elements of the task on which the child should focus,
children are better able to sustain attention and organize strategies to attempt to solve the
problem? (2000, p. 17). This suggests that certain maternal behaviors may be more
influential during earlier years (2 and 3 years) than later years (4 and 5 years).
Although this relationship between specific maternal behaviors and future
cognitive development is fairly well established, there have been few studies to examine
the influence of maternal behaviors, specifically sensitivity and stimulation, on the
development of executive functioning (Assel, Landry, Swank, Smith, & Steelman, 2003;
Landry, Miller-Loncar, Smith, & Swank, 2002). Of the studies that have examined the
influence of maternal behaviors on the broad concept of cognitive skills, the majority
have examined cognitive skills immediately following a play task, and thus, the
3
relationship appears to be most established following an immediate time frame and not
examining how maternal behaviors influences long-term cognitive development.
As Landry and associates (2000) note, ?an important developmental goal for
children during the preschool years is the ability to interact independently with their
world without a high degree of structure and support from others? (p. 358). This requires
the development of a range of cognitive processes (e.g., inhibition) and the coordination
of those processes (e.g., executive functioning). As cognitive development proceeds, it is
expected that these cognitive regulatory processes provided by the parent should become
more internalized until the child can complete the task autonomously. For the purposes of
the current study executive functioning will be defined as a, ?set of cognitive processes
that are necessary for goal-directed behavior and that are mediated by the prefrontal
cortex of the brain? (Welsh, Friedman, & Spieker, 2006, p. 5) and measured via
memory, inhibition, and planning (Welsh & Huizinga, 2001).
The underlying theoretical framework for the current question comes from
developmental and neuropsychological research. It is important to note that this is
secondary analysis of data from the Study of Early Child Care and Youth Development
(SECCYD) (NICHD, 2000). The original NICHD SECCYD research study was
developed from an ecological model and considered two ecological mechanisms that
were thought to influence cognitive and intellectual development (NICHD, 2000). The
first mechanism suggests that children?s cognitive and intellectual development may be
the result of an enriched and varied environment. As children develop they would begin
to actively organize their own experiences through exploration of their environment. This
mechanism would result in parents with greater resources being better equipped to
4
provide cognitively rich stimulation which results in greater intellectual development.
The second mechanism suggests that it is through adult-child interactions and
tutored/scaffolded experiences that children develop intellectually (Vygotsky, 1978).
More specifically, children gain the most intellectually when they are taught using
appropriate, structured experiences.
As previously noted there is limited empirical research relating maternal
behaviors during joint play tasks and future development of executive functioning.
Although the available research appears to clearly note the importance of parental
involvement, the research is less clear about the differential impact of two specific
components, maternal sensitivity and maternal stimulation, of maternal involvement.
These different component skills may be of variable importance for the future
development of executive functioning, depending on the child?s age. This research will
examine the relationship between maternal sensitivity and maternal stimulation and the
later development of executive functions. Within the context of this study, mother-child
interaction is recognized as ?mother and child actively engaged in the same friendly play
or work together? (Gardner, Ward, Burton & Wilson, 2003, p. 368). The current study
defines maternal stimulation as, ?the degree to which the parent tries to foster his/her
child's cognitive and mental development? (NICHD, 2000) and maternal sensitivity as,
?the degree to which parents adapt to children?s needs and abilities? (Beckwith &
Rodning, 1996 as cited in Dodici et al., 2003, p. 125).
Although executive functioning is central to everyday experiences and theorists
have postulated that contextual factors influence the development of executive functions,
as of yet there is little empirical research available to support (or reject) these theories
5
(Henderson & Moore, 1980; Welsh & Pennington, 1988). To establish the motivation
and justification for the present study, the literature review that follows will first examine
the literature on mother-child interactions, specifically stimulation and sensitivity, and
secondly review relevant literature related to executive function, specifically the
constructs theorized to underlie the executive functioning (Simon, 1975; Welsh, 1991;
Welsh & Huizinga, 2005).
Statement of Purpose
The purpose of this study is to add to the small body of empirical research about
the influence of maternal behaviors on the development of executive function. Because
executive function influences nearly every aspect of behavior, the results of this study
could have both theoretical and practical importance. Dodici, Draper, and Peterson
(2003) write that it, ?seems imperative that parents should be made aware of the
influence of their everyday interactions may have on their children?s success in
school.....Parents need to seize teachable moments as they occur all day long, and many
parents may need education and/or support to do this effectively?? (p. 134). This may
have implications for supporting parental involvement in various early intervention
programs.
This study will examine via structural equation modeling the relationship between
early maternal sensitivity and maternal stimulation and later development of executive
functions. Specifically, this study will be examining different factors that have been
either empirically or theoretically linked to maternal behaviors and executive functioning:
early vs. late maternal stimulation and maternal sensitivity, ethnicity (Caucasian and
African American), and maternal verbal ability. A socio-cultural perspective of child
6
development would suggest that children learn through adult interactions but few
researchers have empirically examined how early vs. later maternal stimulation and
sensitivity may influence later executive functioning development and if ethnicity
moderates and maternal verbal ability mediates socioeconomic factors related to the
development of executive functioning.
First, and the major goal of the study, is to examine the relationship between
maternal behaviors on later development of executive functioning. Specifically, this study
will examine whether the influence of maternal stimulation and maternal sensitivity on
executive functioning was direct or indirect through the effect on children?s early
language and attentional skills. Prior research has suggested that certain types of
maternal support is most beneficial during early development by supporting earlier
acquisition of language and cognitive abilities which in turn are prerequisites for the
development of executive functions (Landry et al., 2000).
Secondly this study will examine ethnicity as a moderator variable that influences
the development of executive functions. Ethnicity (i.e., Caucasian and African American)
was selected as a potential moderator because of the strong cross-cultural theory and
research that suggests maternal interactions can be culturally bound. Specifically,
research suggests that African American mothers are more directive in their assistance
during collaborative tasks and ask fewer scaffolding questions, this may result in less
cognitive stimulation and a different mechanism for the development of executive
functioning for different ethnic groups (Connell & Prinz, 2002). This study will examine
whether estimates of model parameters remained invariant across two ethnic groups
7
(Caucasian and African American) or whether group membership moderated the
relationships specified in the model (Frazier, Tix, & Barron, 2004).
Finally, this study will examine the role of maternal verbal ability, as measured by
maternal performance on the PPVT-R, as a possible mediator between SES and executive
functions using structural equation modeling. The meditational hypothesis being tested
suggests that maternal verbal ability mediates the relationship between SES and
executive function. It is expected that increases in SES will result in increased maternal
verbal ability which in turn better predict the development of executive function above
and beyond SES alone.
Research Questions
1. What is the relationship between maternal sensitivity and maternal stimulation
and the later development of executive functioning?
2. Does ethnicity moderate the strength of the relationship between maternal
behaviors and executive functioning?
3. Does maternal verbal ability mediate the relationship between SES and
executive functioning?
Three competing nested models will be developed to examine research question 1.
Nested multiple group models will also be used to examine if ethnic group membership
moderates the relationship between maternal behaviors and executive functioning
(research question 2). Non-nested model comparison will be used to analyze maternal
verbal ability as a mediator between SES and executive functioning (research question 3).
The goal of this study is to identify a parsimonious, substantively meaningful model that
fits the observed data adequately well.
8
CHAPTER II
Literature Review
Overview of Mother-Child Interactions
Development of Mother-Child Interactions
Joint attention, which is process that underlies mother-child interactions, begins to
emerge between 9 to 12 months. Neurocognitive researchers have implicated the
ventromedial prefrontal cortex, which specializes in processing social information, in the
development of joint attention (Nichols, Fox, & Mundy, 2005). Infants begin interacting
dyadically with objects and other people at about six months of age. But as early as 9
months of age, infants are capable of engaging tridyadically between themselves, a
parent, and a third object of interest. Joint attention during infancy is typically assessed
by measuring children?s initiation of joint attention (e.g., gestures and eye contact) and
children?s response to parental attempts to engage (e.g. time on task). Vaughn and
colleagues (2003) report significant individual differences in infant?s joint attention
regulatory and initiation abilities. Developmental knowledge about a child?s ability to
attend and interact has also stemmed from autism research. Individuals with autism
frequently have a decreased ability to initiate or respond to joint interaction, in fact
failure to initiate joint attention is frequently related to the differential diagnosis of autism
(Baird et al., 2000).
Around age two children are first able to understand the basic content of
communication but still require explicit directives related to their focus of interest
(Wertsch, 1979). By age three, children still need assistance to focus on an area of
interest but are able to better express their goals. As cognitive development proceeds it is
9
expected that these cognitive regulatory processes provided by the parent should become
more internalized until the child can autonomously complete the task. For example,
researchers reported that during a collaborative interaction task that required planning
mothers assumed greater responsibility when planning with 4 and 5-year-olds then when
planning with 6 and 7-year-olds (de la Ossa & Gauvain, 2001).
Conceptually, developmental psychologists suggest that self-regulation is learned
as part of the reciprocal process in which social interactions shape current and future
behavior. Vygotsky?s (1978) much cited theory of ?proximal development? emphasizes
the importance of the context in which development occurs. The theory suggests that as
children begin to develop cognitively support can be critical to assist the child in working
above their current autonomous level. During mother-child interactions the zone of
proximal development is typically conceptualized as ?regulating the child?s behavioral
responses, such as directing the child?s attention, maintaining goal direction, reminding
the child where he or she is in the task, and/or evaluating the success or failure of the
child?s task behaviors? that are required for autonomous performance (Freund, 1990, p.
113).
Wood and Middleton (1975) describe how effective instruction supports a child
such that, ?As he enacts and perfects such isolated task constituents, uncertainty about
what to do and what to anticipate as a consequence of his actions diminishes, at least with
regard to subset of the task. This further frees the child to consider the wider or related
task constraints and operations. At best, this process continues until he becomes
acquainted with and skilled in all aspects of task activity to the point where he can initiate
and control his own behavior in the absence of an instructor? (p. 190). Current theory
10
suggests that maternal support (e.g., cognitive stimulation) may be more influential
during certain time periods. During these developmental transitions, certain contexts can
play lesser or greater roles in future development. Specifically, researchers suggest that
certain types of maternal behaviors would be more effective for younger children than for
older children (Landry et al., 2002).
In particular some researchers has found that sensitive and stimulating mother-
child interactions are found to be most beneficial during early development by supporting
earlier acquisition of language and cognitive abilities which in turn are prerequisites for
future cognitive and social development (Landry, Smith, Miller-Loncar, & Swank, 1997).
Other researchers using the NICHD SECCYD dataset (NICHD Early Child Care
Research Network, 2005) have reported dissimilar findings using hierarchical regression
and suggest that a child?s experiences and environment throughout their early and late
childhood make important and roughly equivalent contributions to their cognitive
outcomes. The current study will attempt to further examine these findings using
structural equation modeling.
Mother-Child Interaction and Culture
When considering the development of mother-child interactions, it is also
necessary to consider cultural influences. Valenzuela (1997) writes that, ?cultural
differences may be responsible for the lack of an association between maternal sensitivity
in the home and mother and child behavior during play interactions? (p. 851). Cross-
cultural research suggests significant parental variability in attitudes and behaviors
related to the maternal role (Bornstein, Haynes, Pascual, Painter, & Galperin, 1999;
Roopnarine, Johnson, & Hooper, 1994). Researchers go further than noting the cultural
11
variation and suggest that mother-child interactions are a mechanism for transmitting
culture. Mother-child interactions have been viewed as a microgenic mechanism, which
transmits culture through the use of culturally relevant everyday family tasks and play
(Vandermaas-Peeler, Way, & Umplevy, 2003).
As Connell & Prinz (2002) note, ?Research has generally supported the findings
that parent-child interactions characterized as warm, structured, and emotionally
responsive are related to positive cognitive and behavioral gains in children, regardless of
racial/ethnic group or social class? (p. 189). But other researchers, namely Baldwin,
Baldwin, and Cole (1990), suggest that for African American children more directive
maternal interactions results in more positive development. Thus the current study will
examine the role of ethnicity (Caucasian and African American) in moderating the
influence of mother-child interactions on later development of executive functions.
In one recent study, the authors reported that sensitive mother-child interactions
were, ?found to be a potentially buffering factor for children from families at risk for
school difficulties, such as ethnic minority families? (Morrison, Rimm-Kauffman, &
Pianta, 2003, p. 197). The study examined how mother-child interactions during 2 four-
minute problem solving tasks in kindergarten were related to a number of positive and
negative outcome measures in middle school. Specifically the authors reported that non-
White children, predominantly African American children, who engaged in less sensitive
mother-child interactions, showed more problem behavior in middle school compared
with non-While children who engaged in more sensitive mother-child interactions. Also
non-While children who engaged in more sensitive mother-child interactions showed
similar levels of negative behaviors as their peers. This suggests that maternal sensitivity
12
may have a buffering effect for at-risk children thus sensitive and stimulating parenting
may be more important (i.e., have greater cognitive, social, and behavior benefits) for
certain ethnic groups.
These findings were similar to the work of Buchinal and associates (1997) which
suggested that between-family differences can have a greater impact for certain ethnic
groups, specifically African American children from low income families, as compared
with Caucasian children from middle class families. Also importantly, Nabors, Evans, &
Strickland (2000) note for certain cultures, specifically African American cultures, it is
difficult to disentangle the influence of culture from social class and parental education as
they are highly related with a large percentage of African Americans having lower
socioeconomic status and less education as compared to Caucasians. Morrison, Rimm-
Kauffma, & Pianta (2003) write that, ?These demographic variables (referring to
ethnicity, gender, ses) may be indices of larger cultural variations in social processes,
which together create risk status for the child. Thus, attempts to evaluate the predictive
power of process variables such as mother-child relationships must account for these
demographic predictors? (p. 188). The current model examined if ethnicity (Caucasian
vs. African American) moderated the strength of the relationship between maternal
behaviors and executive functioning.
Maternal-Child Interactions during Joint Play
Parent-child play has been the focus of a considerable amount of research thus a
overview of parent-child interactions will be reviewed here. An important note to the
current study is the exclusion of fathers/male caregivers in the study. This is due to
pragmatic limitations in the NICHD dataset that does not contain samples adequate for
13
the proposed SEM analyses. This is an acknowledged limitation in the current study as
other researchers (e.g., Tamis-LeMonda et al., 2004) have shown that fathers make
unique contributions to children?s cognitive and social outcomes and would be a valuable
avenue for future research.
In one recent study, Gardner, Ward, Burton, and Wilson (2003) examined the
relationship between mother-child interactions during joint play tasks and the
development of conduct disorders. They reported that positive parenting practices as
measured during joint play tasks, makes a unique contribution to the early development
of conduct disorder. Specifically they note that when children engaged in pretend play
with their mothers, the play included more integration and higher-order activities. Fein
and Fryer (1995) report that the effectiveness of maternal play styles represents an
inverted-U-shaped function such that, ?Mothers who are distant or indirect have little
influence on their children; mothers who are intrusive and tutorial have a negative
influence. Mothers who offer direct suggestions, solicit pretend behaviors from their
children, and participate in pretend exchanges have a positive influence on their children?
(p. 378). From this the general conclusion is that mothers are most effective when they
provide moderate support. This U-shaped function is similar to the relationship suggested
by researchers of the effectiveness of teachers. The available literature appears to suggest
that the relationship between maternal support during play activities is complex and that
optimal support requires a knowledge of the child?s abilities, typical development, and
appropriate support strategies.
14
Conceptualizations of Sensitivity and Stimulation
The greatest commonality across mother-child interaction definitions is the basic
idea that adult-child interactions result in a shared task and a common engagement. What
is less clear or defined are the features of adult-child interaction that explicitly influence
children?s learning. Rogoff (1990) notes five theoretically supported features of this
scaffolding like process across conceptualizations that are thought to underlie positive
parent-child interactions. First, each of the conceptualizations includes an expert tutor,
typically a parent, who can bridge the knowledge gap for the novice learner, a child.
Secondly, the expert tutor provides structuring support so that the overall goal of the
project can be met. The third principle states that expert tutor ensures active participation
and contribution from the novice. In conjunction with this concept, the next principle
suggests that effective guidance must involve a transferring or increase of responsibility
from the expert tutor to the novice learner. Finally, and germane to the current study,
these scaffolding encounters frequently occur in a more naturalistic context for children
through everyday play behavior. Following a developmental perspective it could then be
expected that a child would learn specific skills and then generalize those skills (Rogoff,
1990).
Although Rogoff?s research has recognized the core supported tenets of positive
scaffolding support, there is still significant variance between those principles and the
ways they are operationalized and labeled across studies. Because there is frequently
overlap between the proposed mechanisms and factors and the terms used to describe
them, there is confusion within the field as to what exactly is the nature of the help or
support being provided. For example, Vandermaas-Peeler, Way, and Umpleby (2003)
15
report that guidance, engagement, and frequency of interaction varied as independent
factors in their study of parental scaffolding. In another study examining factors that
support childhood outcomes, the researchers report that responsivity/sensitivity,
emotional tone, engagement, parental talk, and parental guidance are theoretically or
empirically recognized as individual but interrelated factors (Dodici et al., 2003). Others,
namely Rogoff (2003), describe parental interactions in terms of ?guided participation? or
?cognitive apprenticeship? while Brown & Campione (1990) use the term ?reciprocal
teaching? to describe this concept. The current study examined two forms of parenting
support, maternal sensitivity and maternal stimulation. These concepts were chosen, in
part, because each broadly encompassed a range of parenting behaviors and in part based
on the previous research which empirically linked them to positive childhood outcomes
(Landry et al., 2002; NICHD, 2005).
The Development of Sensitivity and Stimulation
Developmental psychologists have long recognized the importance of sensitive
and stimulating maternal interactions during early childhood (Dodici et al., 2003; Landry
et al., 2002) but only more recently has there been closer examination of the
developmental unfolding of those sensitive and stimulating interactions (Dallaire &
Weinraub, 2005). It is important to separate parenting practices, which are subject to
change and grow in developmentally appropriate ways over time, from parenting styles,
which have shown relative stability over time. Darling and Steinberg (1993 as cited in
Dallaire & Weinraub, 2005) suggest that the practices or behaviors that parents use to
display their parenting styles (e.g., sensitivity) should change over time. For example
during infancy, sensitive and stimulating parenting is practiced by attuning to the infants
16
non-verbal cues and assisting the infant in maintaining joint attention with the goal of
developing a secure attachment. Keller, Lohaus, Volker, Elben, and Ball (2003) report
that sensitivity (i.e. prompt response to child?s need) is critical during infancy as infants
have limited working memory. For toddlers, sensitive and stimulating parenting
behaviors are focused on self-regulation and guidance with the goal of developing
autonomy and compliance. For preschoolers, sensitive and stimulating parenting is
focused on the development of academic and social competency through both warmth
and setting of limits/control. It is important to note that although the relative stability of
stimulation and sensitive parenting has been supported, the actual behaviors and goals of
these constructs changes based on age appropriate competencies (Dallaire & Weinraub,
2005).
Sensitivity
As noted above, there are a number of positive, neutral, and negative techniques
that parents use to provide structure and support during these interactions. Sensitivity has
been defined many ways such as, ?the degree to which parents adapt to children?s needs
and abilities? (Beckwith & Rodning, 1996, as cited in Dodici et al., 2003, p. 125). Other
researchers conceptualize sensitive parenting more broadly along two dimensions:
warmth and contingency (Keller et al., 2003). Because the current study is longitudinal
in nature and because how a parent displays sensitivity over time should change to meet
the developmental needs of the child, sensitivity is conceptualized more broadly than
many studies which are looking for specific behaviors at a single point in time, The
current study will use two scales to measure sensitivity: supportive presence and respect
for autonomy. These scales will be discussed more fully later in the paper.
17
In Wood and Middleton?s study (1975), they examined effective levels of
sensitivity by mothers. They hypothesized that effective instruction would require a child
to complete an extra operation or decision above and beyond what could be completed
independently. This ?region of sensitivity to instruction? was thought to be the most
effective level of intervention and the researchers suggested the most effective mothers
provide this level of intervention. The researchers examined ways that mother?s
intervened with children and suggested 5 levels of maternal intervention (general verbal
instruction, specific verbal instruction, direct intervention by indicating materials, direct
intervention by assembling material, and demonstrates on operation). The author?s
suggested that, ?the sensitive instructor continually modifies his or her approach to the
teaching task on the basis of the tutees responses. Each instruction can be likened to a
hypothesis; a hypothesis about the most effective level or intervention? (p. 190). More
recently, the NICHD Early Child Care Research Network (1999) reported that maternal
sensitivity was influenced by the amount of time that the child was in child care such that
an increase the amount of hours in childcare resulted in a decrease in sensitivity in
children under 36 months. They also reported that having a martial partner resulted in
higher levels of maternal sensitivity.
Stimulation
As Hubbs-Tait and associates note, ?Although the relation of children?s cognitive
functioning to specific levels of such parental cognitive stimulation varies across
samples, in general, the greater parental use of statements that challenge children to use
representational thought the better children?s cognitive performance? (2002, p. 110).
Sigel, Stinson, & Flaugher?s (1991) operationalization of this concept of cognitive
18
stimulation, which they call cognitive distancing, resulted in the level of cognitive
stimulation being directly related to the type of challenging nature of the questions being
asked by the mother.
This concept was similar to Bernstein?s (1961 as cited in Hess and Shipman
(1965) description of communication styles during parent-child interactions as either
restricted or elaborated. Restricted communication style are stereotyped, scripted, and
lack specificity or differentiation by task or individual. This style of communication is
explicit, directive, and short. On the other hand an elaborated communication style is
more individualized to task and individual. It emphasizes a more complex range of
thoughts and diverse communication styles. For example, if a parent asked a child to pick
up their toys using an elaborated style of communication they would say, ?Johnny, please
pick up these toys and put them in the toy box so we can go to the park? or if they were
using a restricted style a mother would say ?Pick up the toys?. In the first statement, the
child is being asked to integrate several pieces of information: a time dimension, more
complex vocabulary, and to consider the impact of his behavior on future behavior and
others. The second statement the child is expected to comply with only one, simple
request. Other researchers (Landry et al., 2002) use the terms maternal scaffolding and
maternal stimulation interchangeably and operationlaize them more broadly by
examining a number of verbal and non-verbal behaviors that occurred during a 60 minute
interaction.
The current study will measure stimulation using scales at 24 (cognitive
stimulation) and 36 months (mothers as mediators of cognitive development and
cognitive stimulation) and at 54 months (cognitive stimulation, goal directed partnership,
19
and quality of assistance). As noted the operationalization of these measures is important
to understanding the exact nature of what is being measured, thus the next section will
focus broadly on issues related to measuring mother-child interactions and then provide
more specific details about how the current study operationalized maternal sensitivity and
maternal stimulation.
Measuring Mother-Child Interactions
There are several methodological issues to consider when measuring mother-child
interactions. Specifically researchers must consider the reliability, validity, and the utility
of the chosen methodology in relation to the question of interest. Based on the current
research and the available literature, there are many compelling reasons why it is most
appropriate to measure mother-child interactions using direct observations. Namely,
direct observation allow for a fuller understanding of complex behaviors as compared to
parent reports which are subject to systematic bias based on expectations, mood, and
social economic status. (Eddy, Dishon, & Stoolmiller, 1998). Aspland and Gardner
(2003) note the many ways that direct observations can be coded. ?Parent-child
interactions can be interpreted on many different levels including content (e.g., the
semantic content of verbalizations), interactional quality (e.g., affect, sensitivity,
reciprocity), frequency of specified behaviors (e.g. suggestions), and behavioral
contingencies (e.g. praise contingent on compliance)? (Aspland & Gardner, 2003, p.
137). This suggests that direct observations can result in a range of relevant information.
While direct observation techniques are not without their limitations (i.e., time
consuming and expensive), overall they appear to be the most reliable and valid method
for examining mother-child interactions. As a result of the intensiveness and expense it is
20
more uncommon for studies with larger samples sizes to use direct observations. Thus,
the sample sizes for studies which use direct observations are traditionally smaller than
the sample size of studies using self-rating scales. Additionally, because direct
observations are more time consuming and expensive, researchers tend to carry out fewer
observations, which can lead to low/reduced stability of data (Aspland & Gardner, 2003).
Fortunately, the sample size the current study is quite large with multiple time points
which allows for greater generalizability and more stability.
Another concern related to reliability is that, ?high levels of day-to-day variability
in behavior, can lead to problems such as low occurrence of some behaviors, or findings
that do not represent stable estimates of behaviors of interest? (Gardner, 2000, p. 188).
Thus an assumption of the current study is that the, ?behavior during the task bears some
relationship to participants? style of interacting in more natural settings? (Gardner, 2000,
p. 192). This assumption is supported by Dallaire and Weinraub (2005) who reports that
within the NICHD data set maternal stimulation and sensitivity displayed relative
stability (i.e., parents who were stimulating at 6 months were more likely to be
stimulating at 6 years) over the course of birth to age 6. The authors also report that
negative maternal behaviors displayed less stability over time. Finally, the authors report
that the correlations between stimulation and sensitivity got stronger over time suggesting
that there is a greater chance for intervention at younger ages.
Gardner (2000) notes the importance of ensuring that parental behaviors observed
in clinic settings are similar to those observed during natural/home settings. Specifically,
it is necessary to consider the role of reactivity in the current study as observers are
present during the mother-child interactions which occur in a lab setting. One way this
21
study decreased reactivity?s threat was by habituating participants to being observed.
Because the current study is part of a larger longitudinal study in which participants are
observed numerous times both in the lab and at home, it is expected, based on previous
research, that over time and numerous observations participant?s reactivity would
decrease (Gardner, 2000). Several studies have examined systematic factors that
influence reactivity and found that mothers demonstrate less reactivity than fathers and
older children are more reactive than younger children (Lewis, Kier, Hyder, Prenderville,
Pullen, & Stephens, 1996). Other studies have compared methods of recording
observations (e.g., paper and pencil vs. audio) and reported no significant differences
between method of recording and level of reactivity (Gardner, 2000).
Another consideration for measuring mother-child interactions is the type of task
chosen. As Vandermaas-Peeler, Way, and Umpleby (2003) note there are inherent
advantages and disadvantages to observing parent-child interactions using structured vs.
unstructured tasks. The use of structured tasks (e.g., block building using standardized
toys) allows for more quantifiable results that are more easily generalized across
individuals, thereby increasing the reliability of the findings. It also allows you to assess
the associations between the frequency or quality of certain behaviors and specific
outcomes. The use of structured tasks is also valuable because the mother-child
experiences will vary.
The joint play tasks chosen for the current study are common activities (e.g.,
playing with an Etch-a-sketch) for parents and children and are developmentally
appropriate. The current study uses a global rating scheme from which observes
measured behaviors over a continuous observation time. By utilizing global rating scales,
22
it allows researchers to summarize across a set of related behaviors to provide a single
molar rating (Aspland & Gardner, 2003). It is also important to note that, ?interactive
parent?child tasks that involve problem solving or planning offer opportunities for
researchers to capture both the controlling and the warmth dimensions of parenting?
which allows for simultaneous measurement of maternal sensitivity and maternal
stimulation (Dallaire & Weinraub, 2005, p. 203).
There are, of course, limitations to using more structured tasks in that there is an
underlying assumption that all behaviors are equivalent and the tasks are designed to
limit the kind of behavior being elicited. It is also suggested that structured tasks may
overestimate behaviors in general because the tasks are geared to draw out particular
behaviors (Gardner, 2000). Naturalistic observations are also believed to allow for more
authentic culturally relevant results. Another important consideration in measuring joint
play interactions is that it is uncommon to have a normal distribution of positive and
negative behaviors. Research (Gardner et al., 2003) noted that in general adult-child
interactions are positive or neutral. The relationship between negative, harsh, or
inconsistent parent-child interactions and negative childhood outcomes (i.e. increase in
rates of behavioral disorders) is well-established (Gardner, 2000). But research within
naturalistic settings has indicated that these negative interactions typically occur in only
about 20% of interactions (Gardner et al., 2003). In the current study, it is expected that
mother-child interactions will most frequently occur in the neutral or positive category.
Recent research into the measurement of collaborative problem-solving tasks has
reported that structured joint play tasks (vs. unstructured joint play tasks) best
differentiated hyperactive and non-hyperactive children (Befera & Barkley, 1985).
23
Specifically, the authors reported that structured problem-solving tasks in which the child
could not complete the activity without parental assistance best discriminated between
hyperactive and non-hyperactive groups of children. Webster-Stratton (1985 as cited in
Gardner, 2000) reported that mothers provided more directive support and provided more
frequent positive praise within a laboratory setting and the children displayed more
behavioral problems with the home setting. But, more importantly, they reported
moderate to high correlations between maternal behaviors across settings.
There are a number of tasks used to measure mother-child interactions. Flick and
McSweeney (1987) compared three measures of mother-child interactions (Barnard?s
Teaching Scale, Schaefer?s Attachment Inventory, and Clarke-Stewart?s Rating Scales)
and reported that the three measures were interrelated but did not share common
underlying factors. This lack of a common single underlying factor is not unexpected
considering they were each designed to measure different constructs (e.g., attachment).
Gardner and others (2003) developed a system that measured mother-child interactions in
a naturalistic, home setting by coding the occurrence of joint play based on 30-second
intervals. This allowed researchers to examine the amount of time spent in a joint play
task during a typical 1 hour observation. But as the authors acknowledge a limitation to
this method of measuring mother-child interactions is that, ?we did not have a measure of
the quality of joint play, which would help elucidate more precisely the parenting and/or
child mechanisms contributing to outcome? (2003, p. 374).
The Parent-Child Interaction System (PARCHISY) is more similar to the current
study and consists of 13 different 7-point likert scales (Deater-Deckard, Pylas, & Perill,
1997). The parent-child interactions are observed during both unstructured home play
24
and a structured Etch-a-Sketch task similar to the Etch-a-Sketch task used in the current
study. The authors report that PARCHISY scales were associated with children?s social-
emotional status. In sum, Gardner (2000) notes ?Observational techniques can
summarize the relevant aspects of these complex interchanges and thus test hypotheses
about how behavior unfolds over, and how it is influenced by social conditions, including
the behavioral triggers and reactions of others? (p. 187).
The Mother-Child Interaction Task
Because the current study will be examining data collected as part of a larger
study, the goal of the NICHD researchers in using the mother-child interaction task was
to ?assess qualities of parenting such as supportive presence, respect for autonomy,
hostility, and quality of assistance? (1995, p. 18). The children were presented with three
age-appropriate tasks and two of the three tasks were designed and chosen because there
were too difficult to be carried out independently by the child and required maternal
assistance and support. The tasks selected were similar to tasks used in several other
studies (Deater-Deckard et al., 1997; Hess & Shipman, 1965). From these taped
interactions data were coded using two different methods: Mother-Child Interaction Task
Scoring and Mother?s as Mediators of Cognitive Development Scoring.
The first method, mother-child interaction task scoring, was developed
specifically for this study. The scales were adapted for the current study based on the
work of Pianta (1994) and Egeland and Hiester (1993). Using similar scales, Pianta and
Harbers (1996) reported that principal component analysis of maternal-child interaction
factors resulted in a single factor with three maternal facets of interaction (quality of
instruction, supportive presence, and respect for child?s autonomy) and five child aspects
25
interaction (task orientation, affection, self-esteem, negative affect, reliance on mother).
The scoring system at 36 and 54 months consisted of mother-child interactions being
rated by trained coders on a 7-point global rating scale with 1 = ?Very Low? to 7 = ?Very
High? with six scales reflecting mother?s behavior. The scoring system for 24 months
consisted of mother-child interactions being rated by trained coders on a 4 point scale
with 1 = ?Not at all characteristic? to 4 = ?Highly characteristic? with seven scales
reflecting mother?s behavior.
The second method, mothers as mediators of cognitive development scoring,
involved only the 36 month data and examined 28 positive and negative behaviors (e.g.,
labeling object, narrating child?s play, asks question, etc) related to mother?s mediating
cognitive development. This coding system, which was developed by Friedman &
Sherman (1995) for the NICHD study, was applied to 5 minutes of the 15 minutes
videotaped session of mother child interaction. Coding was started when the mother and
child started using the second box. The coding was carried out using a time sampling
technique of observing and then coding 20, 15-second segments of the videotape. From
this system a behavioral composite was developed (Total Positive Scaffolding). This
composite was computed using the weighted sum of the items that promoted positive
cognitive development and included 25 positive behaviors.
The current study will operationalize mother-child interactions by measures of
both maternal stimulation and maternal sensitivity. Early stimulation was measured by
two scale scores (stimulation of development at 24 months and cognitive stimulation at
36 months) from the mother-child interaction task scoring and one composite score (total
positive scaffolding at 36 months) from the mother as a mediator of cognitive
26
development scoring. Late stimulation was measured using three scale scores (cognitive
stimulation, goal directed partnership, and quality of assistance) from the mother-child
interaction task scoring. Early sensitivity was measured using three scale scores
(supportive presence at 36 months, respect for autonomy at 36 months, and sensitivity to
distress at 24 months) from the mother-child interaction task scoring. Late sensitivity was
measured with two scale scores (supportive presence at 54 months and respect for
autonomy at 54 months). The next section will more examine each scale in more detail.
The cognitive stimulation scale is purported to measure the, ?degree to which the
parent tries to foster his/her child's cognitive and mental development? (NICHD, 1995, p.
23). Parents who receive high scores on this scale, ?(a) help their children acquire or
master new skills, (b) illustrate or teach concepts or principles, (c) ask questions that
encourage problem solving, (d) encourage or reinforce sophisticated pretend play
sequences, and so on? (NICHD, 1995, p. 23). Additionally the level of support must be
appropriate based on the child?s cognitive and developmental ability (e.g., illustrating
how to write letters for a 12-month-old would be scored negatively).
Quality of assistance was designed to measure ?how well the parent structures the
situations so that the child knows what the task objectives are and receives hints or
corrections while solving the problems that are: (a) timely to his/her current focus,(b)
paced at a rate that allows comprehension and use of each hint, (c) graded in logical steps
that the child can understand, and (d) stated clearly without unnecessary digressions to
unrelated phenomena or aspects of the task that might only confuse the child? (NICHD,
1995, p. 24). Mothers who receive high scores demonstrate a plan to assist their child but
flexibility to change that plan based on the child?s abilities.
27
The total positive scaffolding score was designed to measure, ?how the mother
conveys cognitive information to her child and the specific kind of information she
conveys? (NICHD, 2001, p. 4). Mothers with higher scores represent more frequent and
more complex types of cognitive assistance. Seven general areas of behavior were coded:
type of mother question, response to child question, reinforce, type of mother directive,
type of demonstration/how things function, goals, plans as statements, and type of mother
statement and the scores were summed.
Maternal sensitivity is based on three scales. The supportive presence scale
measures positive regard and emotional support toward the child. Parents who receive
high scores recognize the child?s accomplishments through verbal and physical praise
and positive regard and provide support and reassurance to children who are struggling
with tasks.
The respect for child?s autonomy scale measures, ?the degree to which the parent
acted in a way that recognized and respected the validity of the child's individuality,
motives, and perspectives in the session? (NICHD, 1995, p. 21). Parents who score high
on this scale, ?acknowledge the child's perspectives and desires as a valid part of the
child's individual identity? (NICHD, 1995, p. 21). This acknowledgement is done through
explicitly recognizing and collaborating with the child about the task and roles and
validating the child?s intentions, desires, and individuality. The sensitivity to distress
scale focuses on, ?how the mother observes and responds to the child's social gestures,
expressions, and signals. The key defining characteristic of sensitive interaction is that it
is child centered. The sensitive mother is tuned to the child and manifests awareness of
28
the child's needs, moods, interests, and capabilities, and allows this awareness to guide
his/her interaction? (NICHD, 1993, p. 38).
Overview of Executive Functions
Recently there has been a flood of research examining executive functions due, in
part, to the paramount role the prefrontal lobes play in every aspect of human
functioning. Unfortunately, there has yet to be one accepted definition. The term
executive cognition is defined by Duffy and Campbell (2001) as ?the qualitative
organizing principles necessary to navigate the fluctuating and ambiguous challenges
confronted in autonomous social behavior? (p. 114). This is compared to Welsh (2002)
who defines executive function as, ?a cluster of skills that are necessary for efficient and
effective future-oriented behavior...planning, inhibition, monitoring, and flexibility? (p.
143). The first definition appears to speak more to the self-monitoring and set-changing
where as the second definition focuses on attention and set preservation.
Miyake, Friedman, Emerson, Witzki, Howerter, and Wager (2000) examined a
model based on three executive function processes: mental set shifting, inhibition, and
updating (i.e., working-memory). This model focuses more on the changes in attention
and inhibition than the previous models. Welsh and others (2002) examined a five-factor
model consisting of working memory, inhibition, planning, flexibility, and rule induction.
Working memory and inhibition jointly influence flexibility, but only working memory
covaries with planning and rule induction. Tower of Hanoi-Revised and Matrix
Reasoning both loaded on to the rule inductive factor.
Miyake et al. (2000) conclusions are compared to the findings of Senn, Espy, and
Kaufmann (2004) who used path analysis to determine the relationship between
29
preschool aged children?s performance (n=117) on a complex problem solving task
(TOH) and measures of working memory, inhibition, and set shifting. The best fitting
model included paths between working memory and inhibition which accounted for 29%
of the overall variance in the TOH scores. Interestingly, inhibition more strongly
predicted performance in children under age four and working memory was a better
predictor in children over age four indicating a protracted course of development. The
authors suggest that older children?s better developed working memory allow them to
solve more complex TOH problems that require greater working memory. This suggests
that the underlying components of EF have a complex relationship. As Senn, Espy, and
Kaufmann (2004) note, ?Clearly, longitudinal studies are necessary to characterize these
ontogenetic relations in young children and how they map onto subsequent executive
functions during the school-age years? (p. 459).
History of Executive Functions
In 1928, American neurologist Tilney called the period of human evolution the
?age of the frontal lobe.? The frontal lobes, which are located in the anterior third of the
cerebral hemispheres, are evolutionarily the most recent. Although concrete data
concerning the frontal lobes are just recently emerging, the frontal lobes have been shown
to play a role in nearly every aspect of human functioning from fine motor movement to
intricate social situations. Originally researchers attributed executive functions to the
dorsolateral and orbital-frontal prefrontal subsystems (Iversen & Dunnett, 1990). But
more recent conceptualizations have parsed the executive functions even further with the
dorsolateral prefrontal cortex driving planning, working memory, and stragetic reasoning
and the ventromedial/orbitofrontal prefrontal cortex responsible for self-monitoring and
30
emotional responses (Banfield, Wyland, & Macrae, 2004). Therefore it receives sensory
information from nearly every modality, including somatic, auditory, visual, olfactory,
and gustatory senses, and integrates all of these cross-sensory modalities and generates
responses (Malloy & Richardson, 2001). The frontal lobes contain some of the most
complex reciprocal circuitry. Based on these observations, the human frontal lobes,
specifically the prefrontal cortex, are theorized to play key roles in higher-order cognitive
processes or executive function.
The frontal lobes account for the largest portion of the total human cortex. The
prefrontal cortex results from growth and convergence of primordial fields of the
parahippocampal, insular area, and the motor cortex area (Fuster, 1997). The prefrontal
cortex is histologically heterogeneous and extensively filled with connections. The
cortex of the prefrontal lobe is typically six-layered with a distinct internal granular layer.
The prefrontal cortex has rich connections with the limbic system, which play an active
role in higher order thinking. Damage to the prefrontal lobes causes deficits in
coordination and management of behavior, which is manifested as an inability to
integrate sensory information, impoverished behavioral responses, loss of attention,
inflexibility, and an inability to self-monitor.
Development of Executive Functions
Executive functions including attentional control, problem solving, and self-
regulation are recognized as early as infancy, however at the younger ages executive
functions are, ?variable, fragile, and bound to external stimulus situation, with increasing
stability achieved between 18 and 30 months of age? (Ruff & Rothbarth, 1996, as cited in
Isquith, Gioia, & Espy, 2004, p. 405). Specific cognitive process, such as working
31
memory, flexibility, and inhibition, can be accurately measured in preschoolers but more
complex problem solving and planning develop along a more protracted course (Espy,
Kaufmann, Glisky, & McDiarmid, 2001). It is also reported that, ?the emergence and
development of executive functions likely varies across individuals in terms of age and
the shape of the trajectory of skill acquisition? (Isquith et al., 2004, p. 405).
Researchers who have examined the development of executive functions across 4
different age groups (6-year-olds, 8-year-olds, 10 ?year-olds, and 12-year-olds) reported
that generally the most active period of development occurred between the ages of 6 and
8 years of age, although inhibition related to mnemonic representations developed around
12 years of age (Becker, Isaac, & Hynd, 1987; Passler, Isaac, & Hynd, 1985). Other
researchers (Welsh, Pennington, & Groisser, 1991) examined the development of a
greater range of executive functions tasks in children aged 3 to 12 and reported that age at
which adult like performance on the tasks was achieved was task dependent. Specifically,
they noted that children?s performance on simple planning tasks displayed adult like
performance around age 6, tasks that required inhibition skills displayed adult
performance at age 10, and finally tasks that required complex planning and had a
memory component displayed adult like performance at age 12 (Welsh et al., 1991).
These studies suggest that the development of executive functions is similar to most
complex psychological processes in that it occurs in a multistage fashion with certain
simple problem solving skills developing around age 6.
General systems models are one systematic way to identify variables that
influence development over time (i.e., executive functions) and to understand how
mechanisms may differ by a number of characteristics (e.g., SES). Within a general
32
systems model, there appears to be a hierarchy of factors that promote or prevent
development (i.e., risk or protective factors) (Bronfenbrenner & Ceci, 1994). For example
researchers have identified specific child characteristics (i.e., temperament, task
orientation, gender) related to cognitive outcomes. Development is thought to occur
through a dynamic iterative process in which the environment changes in relation to the
child?s responses.
Burchinal and associates (1997) following a general systems model for cognitive
performance theorize that cognitive level of parent has a direct genetic effect on a child?s
cognitive performance and an indirect influence through care giving characteristics (e.g.,
interaction styles). Following a general systems model, they suggest that, ?frequent
responsive and stimulating interactions with caregivers are believed to facilitate
children?s cognitive development because responsive interactions should provide
scaffolded learning experiences? (p. 936). Although there is no general systems model for
the development of executive functions, the current study will attempt to understand the
indirect and direct parental influences on the development of executive functioning.
Maternal Verbal Ability
An important goal of the current study is to examine the influence of maternal
verbal ability on the relationship between SES and executive functions. Specifically, this
study will examine if maternal verbal ability, as measured by the Peabody Picture
Vocabulary Test ?Revised (Dunn & Dunn, 1981) mediates that relationship. It is
important to recognize the limitation of using the PPVT-R as it is very culturally bound
and more related to an individual?s fund of knowledge than to the more fluid, process
based measures. Researchers (NICHD Early Child Care Network, 2005) have examined
33
the relationship between attention, memory, and planning from home and school
environments used the PPVT-R as a proxy of maternal verbal ability (NICHD, 2005).
The researchers reported that, ?to the extent that maternal vocabulary controls for
heritability, the findings cannot be explained by the shared intelligence of mothers and
children, which could be a source of both environmental and cognitive variations among
children? (NICHD, 2005, p. 110). The results of the study suggest that after controlling
for maternal verbal intelligence, early and late family, classroom, and childcare
environments still predicted performance on measures of attention, memory, and
planning, although the strength of that relationship was weakened. These findings support
the use of PPVT-R as a proxy for maternal verbal ability in the current study. The
current study is similar to the NICHD study in that ?ideally, we would have liked to
control for parents? performance on executive function?, unfortunately, due the pragmatic
limitations of the data that information is not available (NICHD, 2005, p. 110).
In another study that examined maternal interactions and future development of
executive functions, the authors wrote that, ?the absence of any measure of mothers?
intellectual abilities, genetic backgrounds, or both in the model may raise questions as to
whether children?s cognitive outcomes are related, in part, to shared genotypes between
parents and children? (Landry et al., 2002, p. 37). Based on prior research, this study will
examine if maternal verbal intelligence, as measured by the PPVT-R mediates the
relationship between SES and later executive functions.
Culture and Measures of Executive Function
As Campbell, Dollaghan, Needleman, and Janosky (1997) note there are
conceptual differences between knowledge-based and processing-basing measures,
34
specifically theorizing that knowledge-based (i.e., crystallized) assessments are more
influenced by opportunity to learn. These cultural expectations are tied to specific
knowledge bases that make it particularly difficult to generalize and interpret the results.
Examining the cultural aspects of the TOH specifically, it is purported to be less
culturally bound than other measures of EF (Wallner, 1996). For example, the above
mentioned grocery-shopping task is similar to the TOH in that it requires configuration of
the problem state but differs in that the grocery shopping task requires using culturally
based cues. Also the items being purchased are ethnocentrically biased.
Welsh and associates report that ?very young children can plan a birthday party or
perform well on an errand-running task but have more difficulty on the abstract Tower of
Hanoi task? (2006, p.33). Because the TOH task is a novel task with few culturally
bound requirements, it allows for a more pure measure of executive functioning without
the influence of cultural schemas. As Wallner (1996) notes the TOH requires little
domain knowledge as it is based on arbitrary, artificial rules thus when participants
encounter the TOH they cannot use specific knowledge they have acquired from past
experiences (i.e. they lack a script from which to represent the problem). Thus it is
expected because TOH is a novel, processing-based task that cultural biases would be
minimized. This is compared to participant?s performance on the current measure of
short-term memory which is presented verbally and has a crystallized component. It is
anticipated that this measure (WJ-R Memory for Names) will be more confounded as a
measure of EF because of the lack of distinction been crystallized and fluid knowledge
bases.
35
Measuring Executive Function
One of the greatest hindrances to research examining frontal lobe functioning is
the lack of adequate and sensitive tests. The majority of the available executive function
tasks were designed to be sensitive to brain damage but lacked a theoretical and empirical
foundation within typically developing populations. The lack of consensus regarding the
definition of executive functions results in diverse measures of executive functions which
potentially map onto many diverse cognitive processes therefore executive functioning
could be task dependent (Brocki & Bohlin, 2004). The coordination of these diverse
cognitive processes makes executive functions unique. As Espy, McDiarmid, Cwik,
Senn, Hamby, and Stalets (2004) note many of the current measures of executive
function are quite diverse and should not be considered the same based solely on face
validity. As Welsh, Friedman, and Spieker (2006) write, ?tests of executive function
should vary on a continuum between mind twisters that are completely unfamiliar to the
person examined and tests that model real life tasks? (p.33).
Another limitation of many of the current executive function tasks is that they
were developed based on principles of frontal lobe functioning in adults as opposed to
developmental frameworks (Welsh & Pennington, 1988). For example, the Wisconsin
Card Sorting Test is a well known neuropsychological task used to assess executive
function (WCST; Miller & Cohen, 2001). This task requires set shifting, creation of
strategies, and information organization (i.e., inductive reasoning), together within one
task. Using the WCST, Milner (1963) presented research from adult patients with frontal
lobe trauma who were able to state the correct rule but unable to complete the task. This
compelling evidence led researchers to make a teleological error. It became a common
36
belief that because children were unable to successfully complete the WCST task then
they were incapable of executive functions.
Therefore prior to the 1980?s, most neurospychologists believed that until
puberty executive functions were not engaged (Golden, 1981). As Isquith, Gioia, and
Espy note, ?Historically they (young children) are perceived as lacking inhibitory control,
exhibiting significant distractibility, being inflexible in their ability to solve problems,
and not organizing, planning, or monitoring their problem-solving behaviors (2004, p.
404). Since that time, researchers have clearly dispelled this idea and demonstrated that
children do in fact possess executive functions (for a review of this literature see Welsh et
al, 2006). Currently, literature and research suggests that the development of planful,
future-directed, intentional behavior begins in infancy and can be assessed using
measures of object permanence (Diamond, 1990). It is also recently recognized that a
number of disorders involving executive dysfunction are manifested in preschool years
and change over time including ADHD, autism spectrum disorder, obsessive-compulsive
disorder, and PKU (Welsh, 2002). Historically speaking the measurement of executive
functioning in children is a relatively new area of research.
There are several executive function tasks commonly used with infants and
children. Although complex coordination of skills is typically not evident until at least 36
months, Piaget?s classic A-not-B task of object permanence is the considered the earliest
(12 months) measure of emerging executive function and is reported to be sensitive to
frontal lobe damage (Welsh, 2002). Currently the NEPSY Developmental
Neuropsychological Assessment is one of the only validated, standardized measures of
executive function for preschool and early childhood. It is important to note that the
37
Tower of London, which is a tower task that is similar, but not isomorphic, to the TOH, is
one of the subtests administered under the attention-executive function domain. Another
recently developed assessment is the Behavior Rating of Inventory of Executive Function
(BRIEF; Gioia, Isquith, Guy, & Kenworthy, 2000). This parent and teacher rating scale
attempts to include eco-behavioral validity while assessing a child?s inhibition, ability to
shift, display of emotional control, ability to initiate, working memory, ability to
plan/organize, and ability to monitor. While there are a number of measures of executive
function, the next section will describe and examine relevant literature related to
measures used in the current study.
Measuring Executive Function in the Current Study
The multi-faceted view of executive function supports differential patterns of
development across factors (Welsh et al, 2006). Thus the ?measures that are included to
tap a single executive function construct inevitably demand multiple executive function
abilities for proficient performance? (Senn et al., 2004, p. 469). While the debate about
the multi-faceted nature of executive function continues, the empirical research to this
point suggests that executive functions are correlated but separable. More specifically, it
has been found that different aspects of executive function differentially predict
performance on frontal lobe tasks. Thus the current study views executive function from
both a unitary and a multi-faceted perspective in that different aspects of executive
function work together but can be measured separately. After examining the current
theoretical and empirical research and recognizing the limitations of using an archival
data set, the current study will measure executive function using measures of
planning/problem solving (TOH), inhibition (CPT ? commission errors), and memory
38
(WJ-R memory for names and WJ-R memory for faces). Each of these concepts and
relevant literature will be discussed below. Planning/Problem Solving
Within the current study, planning is defined as, ?the predetermination of a
sequence of actions aimed at achieving a goal? (Wallner, 1996, p. 1). Viewed as a future
oriented process, planning requires both problem-solving and metacognition. Scholnick
and Friedman (1987) suggest that successful planning requires the following components:
the ability to construct a representation of the environment, select an appropriate goal,
recognize when it is necessary to plan, formulate a successful plan, execute the plan,
monitor or change the plan as necessary, and evaluate the effectiveness of the plan.
There is an obvious conceptual relationship between the definition of executive function
and planning (Wallner, 1996). Specifically, is noted that different executive functions
influence different aspects of the planning process. Wallner (1996) examined the overlap
between measures of planning and executive function and reported that a common
planning task (the grocery store task) accounted for only 25% of the variance on a
measure of executive function (TOH). In the current study, planning/problem solving was
measured using the TOH.
Tower of Hanoi. The TOH is a problem-solving, disk-transfer task that requires
the use of a recursive strategy for optimal performance. The task was originally studied
by Simon in 1975 with other prominent TOH researchers including Jean Piaget. The
TOH has been used as a measure of frontal lobe function with regards to a variety of
neurological-disorders (Malloy & Richardson, 2001) and has demonstrated sensitivity to
prefrontal lobe function and dysfunction (Fuster, 1997; Stuss & Benson, 1986). Wallner
(1996) writes that the TOH requires that, ?individuals must effectively allocate the
39
limited processing resources that are available to construct a sequence of moves in
advance of action? (p. 23). The principal investigators of the current study (NICHD
Early Child Care Research Network, n.d.) note that the TOH task was chosen because of
its extensive research with typically developing adults, sensitivity to age differences and
cognitive disabilities, and lack of cultural specific knowledge. Welsh, Pennington, and
Groisser (1991) report that children as young as age 6 can display adult-like performance
on the simplest TOH problems. But more difficult TOH problems, which typically
include an increased number of moves, require more advanced coordination of working
memory, inhibition, and attention. Hence adult performance is not achieved until age 12
(Welsh et al., 1991).
Although the majority of the TOH research has been examined with adults,
several studies (Bagely, 2002; Scholnick et al., 1997; Welsh, 1991) have examined the
development and use of several versions of the TOH with typically developing school-
age children. Because there are several different child versions of the TOH (see
Scholnick et al., 1997 for a review of differences) and each may tap different underlying
processes and were tested using relatively small non-representative sample sizes, it is
difficult to make generalizable conclusions across versions. Those limitations being
noted, the research with the TOH has indicated in general that the TOH performance
varies systematically in relation to age and cognitive ability (Welsh & Pennington, 1988).
Although there are several versions of the TOH the current study will use a version most
similar to that developed by Welsh (1991) and is briefly described below.
The task consists of moving disks of various sizes on a stand with three equal size
posts. The task has the following constraints: (a) disks may only be moved one at a time;
40
(b) disks may not be set on the table or held in the participant?s hand during the
movement of another disk; and (c) smaller disks must be placed on larger disks. The
object is to move the disks from the start state to the goal state in the fewest number of
moves. This is known as the optimal solution path. The TOH is based on a rule-based
recursive strategy that, once understood, allows the participant to complete any problem
regardless of difficulty level. Goal-recursion is based on cycles of movements necessary
to achieve the goal state. Each task is created by smaller sub-goals that require a repeated
strategy to solve. This repeated cycle of moves has been theorized to be learned through
an implicit or procedural learning process (Bagley et al., 2002; Devine, Welsh, Retzlaff,
Yoh, & Adams, 2001) This recursive strategy, which is unique to the TOH, is theorized
to reduce the impact of other executive function processes such as planning, working
memory, and inhibition (Welsh & Huizinga, 2001).
There are two divergent bodies of literature as to what underlying cognitive
processes the TOH really assess. One body of literature suggests that the TOH is a
neuropsychological measure of executive function and administration should involve a
one trial format with multiple items of increasing difficulty (Bagley et al., 2002;
Goldberg, Saint-Cyr & Weinberger, 1990; Welsh, 1991). These researchers observe
success with achieving the end goal state regardless of time or a decrease in errors across
time. The contrasting view postulates that the TOH is a measure of procedural learning
and improvement in performance over successive trails is based on the acquisition of
implicit skills. The method of administration involves identical single trials given over
multiple sessions to gauge performance improvement. An improvement in performance is
measured based on a decrease in the number of moves needed to eventually achieve the
41
goal state. Thus any improvement is considered to be due to procedural learning (i.e.,
implicit learning). There is much evidence that a variety of cognitive processes are
involved in the TOH task. Some of these suggested processes include working memory,
(Goel, Pullara, & Grafman, 2001; Senn et al., 2004; Welsh et al., 2002) inhibition (Senn
et al., 2004; Welsh, Satterlee-Cartmell, & Stine, 1999), procedural learning/implicit
memory (Bagley et al., 2002; Devine et al., 2001; Goldberg et al., 1990), and fluid
intelligence/explicit reasoning (Devine et al., 2001; Emick & Welsh, 2005; Lock, Welsh,
Adams, & Kurtz, 2002).
Inhibition
For the current study, inhibition is viewed from a neuropsychological perspective
in which it is described as the ability to remove non-relevant information during
reasoning and withholding a prepotent response. This conceptualization of inhibition
covers a broad spectrum of behaviors frequently seen in young children. With regards to
other measures of executive function, inhibition is conceptualized as more than a pure
withholding of a response but requires cognitive flexibility to keep irrelevant information
from interrupting cognitive processes. Inhibition, in other model?s of executive function,
namely (Barkley, 1997), is theorized to be a primary deficit, which contributes to
difficulties with other executive processes that are ?downline?. Upon reviewing the
literature, there are multiple definitions of inhibition. For example, Sergeant and
colleagues (2002) reported 12 different definitions of inhibition. When considering
development and the ability to selectively attend or inhibit, children improve with age.
Schachar and Logan (1990) examined developmental differences on a stop-go inhibition
42
task. They reported on that task that children reached adult level of performance around
2
nd
grade with the greatest variability in performance at the younger ages.
CPT ? commission errors. In the current study inhibition was measured using
CPT-commission errors. While commission errors traditionally reflect disinhibition, the
majority of the research related to this task comes from ADHD research (see Riccio,
Reynolds, & Lowe, 2001 for review). As such, disinhibtion is typically seen as an
underlying factor in ADHD. Schachar and Logan (1990), using a stop-go paradigm,
reported that individual?s with ADHD inhibit less on the task. Research linking inhibition
to the TOH has been inconclusive. For example, Welsh et al., (1999) found a relationship
between the TOH-R and inhibition (CPT), but the relationship was somewhat weak (r =
.31, p=.06). This is compared to other researchers who reported working memory and
inhibition were central to performance on the TOH and accounted for nearly 29% of the
variance (Senn et al., 2004).
Memory
Memory influences nearly all mental processes including language, planning, and
problem solving. Within the psychological literature memory is encoding, processing,
and retrieval of information (Swanson & Cooney, 1991). Memory has been
conceptualized in a number of different ways but has traditionally been broken into long-
term and short-term memory (Atkinson & Schiffrin, 1968). Research suggests that
memory abilities are developmentally relatively stable (Anastasi, 1982). Researchers
have examined the influence of parenting behaviors and the development memory and
found that certain parenting practices (i.e., when parents talk about current events with
their child) contribute to the development of memory ability (Ornstein, Haden, &
43
Hedrick, 2004). The current study includes measures of both short-term and long-term
retrieval.
WJ-R Memory for Sentences (MS). Short-term memory is a capacity limited
system that retains information temporarily (Swanson & Cooney, 1991). Short-term
memory is related to working memory in that both require a participant to hold
information in the conscious awareness but working memory is more future-orientated
and short-term memory is more immediate. Short-term memory tasks, such as the one
used in the current study, typically require the preservation of sequential information.
Within the CHC theory, short-term memory is labeled as Gsm under the memory (MS)
label. Gsm is described as the ability to hold information in conscious or immediate
awareness and then use it within a few seconds. Hutton and Towse (2001) examined the
relationship between short-term memory and cognitive abilities and reported important
developmental differences such that the relationship appears to be stronger in younger
ages compared to adults. There is also an obvious relationship between short-term and
long-term retrieval. Specifically, if short-term memory is deficient than this will interfere
with formation of long-term storage and retrieval.
While much of the research related to executive function is focused on working
memory, there is some research to support the relationship between short term memory
and Raven?s matrices. Specifically, Hutton and Towse (2001) examined the relationships
between short term memory, working memory, and problem solving skills. In the study,
the relationship between the working memory task and the short term memory task was
r=.74. They reported that the both the short term memory task (r=.43) and the working
memory task (r=.53) correlated moderately with the Raven?s matrices task. Importantly,
44
they reported that when processing time was controlled as a factor in the working
memory task then ?the relationships between STM and abilities (referring to Matrices,
reading, and math tasks) were quantitatively superior to the WM scores. Furthermore,
WM measures no longer correlated with matrices or number task after controlling for
processing time? (Hutton & Towse, 2001, p. 391). Examining the literature related to
the WJ-R Memory for Sentences, Dean and Woodcock (1999) examined the predictive
validity of the WJ-R Memory for Sentences with closed head injury and brief loss of
consciousness from those with a psychiatric diagnosis of depression and/or anxiety. They
reported that performance on the Memory for Sentences task significantly discriminated
between groups. It is important to note that there is a crystallized knowledge based
component to this task such that verbal comprehension is secondary but important
requirement of the task. Other researchers, NICHD Early Child Care Network, using the
NICHD SECCYD dataset have reported that both early and late maternal behaviors
influences the development of short-term memory as measured by the WJ-R Memory for
Sentences (2005a). While it may have been preferable to use a measure of working
memory in the current study, there is some empirical literature to suggest a relationship
between executive function and short-term memory that supports its use in the current
study.
WJ-R Memory for Names (MN). Long-term storage retrieval is defined as
information storage that is relatively permanent and resistant to forgetting, although not
always readily recalled (Atkinson & Schiffrin, 1968). The ability to retrieve information
is strongly dependent on a number of factors including an individual?s ability to recall
and adequate storage. While information within long-term storage is typically coded
45
semantically, the current task is visually coded. Long term storage and retrieval, Glr as
conceptualized by the CHC theory, is defined as, ?the ability to store information in and
fluently retrieve new or previously acquired information (e.g., concepts, ideas, and
names) from the long term memory? (Flanagan, McGrew, & Ortiz, 2000). The MN task
does not require knowledge itself but rather the ability to store and to consciously search
for relevant information. Dean and Woodcock (1999) examined results from a sample of
1,315 clinical patients, age 5 to 81 years. They reported that the majority of the clinical
groups experienced their most severe problems with three subtests from the WJ-R with
MN being one of those subtests.
Osmon, Smerz, Braun, and Plambeck (2006) measured the relationship between
executive function and both short term (Gsm) and long term memory (Glr). In the study
executive function was measured by both the Category Test and the Benton Judgment of
Line Orientation. Long-term memory (Glr) was measured using subtests from the WJ-R
(i.e., Memory for Names and Visual-Auditory Learning) and short-term memory (Gsm)
was measured using the Memory for Sentences and Memory for Words. The authors
reported that the measures of short term memory correlated with the Category Test at r=.-
.028 and with the Benton Judgment of Line Orientation at r=.202. Importantly the long-
term memory measures correlated with the Category test at r=.227 and with the Benton
Judgment Line Orientation at r=.371.
Also notably data from the K-ABC-II standardization reported a correlation
between measures of Atlantis and Rover. The Atlantis subtest (Glr) was designed, ?to
measure a child?s ability to learn new information using attractive, intrinsically
interesting stimuli, much like the Memory for Names subtest on the Woodcock-Johnson
46
Psycho-Educational Battery-Revised? (Kaufman & Kaufman, 2004, p. 64). The Rover
subtest was designed primarily, ?to measure the executive functions associated with the
frontal lobe? (Kaufman & Kaufman, 2004, p. 64). The authors reported that the task was
inspired by the Tower of Hanoi. Notably, the correlations between the two tasks varied as
a function of age with the correlation increasing over time. Specifically the relationship
was r=.17 at age 6 and r=.29 at age 15. This suggests that development may be a factor in
the relationship between executive function and long-term memory.
Research related to long-term storage retrieval has suggested that, ?the processes
involved in entering memory trace into long-term store are important sources of ability
group differences in children?s long-term recall? (p. 32). There has been a significant
amount of research observing the connection between the TOH task and procedural
learning which is one type of long-term storage (Bagley et al., 2002; Devine, 2000;
Goldberg et al., 1990). Schmand, Brand, and Kuipers (1992) found that schizophrenic
patients could score within normal range on a version of the TOH even with a loss of
declarative memory and that performance on the TOH could improve based on practice
without explicit knowledge of how to perform the task. Bagley (2001) and Devine
(2000) observed the connection (r = -.315, p<.05, and r = -.511, p<.05 respectively)
between procedural learning (as measured by Mirror Tracing) and the TOH-R.
Mother-Child Interactions and Executive Function
After reviewing the current literature, the parallels between maternal behavior and
executive function become apparent. First it must be noted that mother-child interactions
and executive function share some external similarities. In order to engage in appropriate
shared interactions it requires that children coordinate basic cognitive processes including
47
attention, inhibition, and self-regulation within framework of a social situation. This is
similar to the behaviors associated with executive functioning such as: the ability to
maintain set, the ability to inhibit a response, flexibility, self-monitoring, and planning
(Welsh & Huizinga, 2001). Many of the behaviors associated with executive function are
necessary to engage and sustain mother-child interactions. There are very few studies that
have tested the relationship between maternal behaviors and the development of
executive function. The following four studies have researched that relationship:
Henderson and Moore (1980); Nichols, Fox, and Mundy (2005); NICHD Early Childcare
Network (2005a); and Landry, Miller-Loncar, Smith, and Swank (2002). The direction of
the current study is based, in part, on this past research. It is important to note that across
studies both maternal behaviors and executive functions are measured using a number of
different methods and conceptualizations.
Henderson and Moore (1980) examined the influence of adult interaction,
specifically intrusiveness and responsiveness, during childhood exploratory play.
Complexity in play has been linked to executive functioning. Children (n=48) aged 42
months to 60 months were assigned to either a high curiosity or a low curiosity group
based on play complexity scores from a previous study. The high curiosity and low
curiosity groups were then randomly assigned, controlling for gender and level of
curiosity, to one of three experimental (demonstration, responsive, or nonresponsive)
groups or one control group. These groups varied on two dimensions of adult interaction:
intrusiveness and responsiveness. Each participant then participated in seven 10-minute
play sessions with the seventh session being a measure of generalization that included a
different non-familiar adult.
48
Overall, the researchers reported significant differences between the high
curiosity and low curiosity groups. They found that in general children with high
curiosity levels displayed more complex play than the low curiosity group with the
exception of the non-responsive condition. During the non-responsive condition, the high
curiosity levels did not ask additional questions if they were unanswered as compared to
the high responsive condition. Researchers also noted that there was not an increase in
the complexity of play or curiosity over the six sessions for the low curiosity group. This
suggests that reinforcement of these concepts requires more intensive daily interactions
such as those provided by parents. This study appears to support the notion of individual
differences in children?s play complexity and that certain types of adult-children
interaction styles can influence the complexity and curiosity during play.
Research from neuroscience has also supported the relationship between
executive function and mother-child interactions. In a recent longitudinal study, infants
were assessed at 14, 16, and 18 months and researchers examined the relationship
between mother-child interactions and prefrontal lobe functioning as measured by EEG
indices of left-frontal activity (Nichols et al., 2005). Nonverbal mother-child interactions
were measured using a structured assessment in which the child was seated on the
mother?s lap and presented with various stimuli, a delayed nonmatching to sample
paradigm task was used to measure ventromedial prefrontal cortex functioning, and a
means end task was used to measure dorsalmedial prefrontal cortex functioning. Using
hierarchical regression analysis, the researchers reported that the ventromedial prefrontal
cortex is related to mother-child joint attention. While the current study did not directly
measure brain functioning, several prior studies within the same laboratory (see Mundy,
49
Fox, & Card, 2003) had defined the relationship between these tasks and brain
functioning. The ventromedial prefrontal cortex is also recognized as playing a crucial
role in executive functioning skills such as, ?establishing and inhibiting stimulus-reward
associations and for being flexible in the application and modification of social rewards?
(Nichols et al., 2005, p. 37). These findings show emerging support for the importance of
the ventromedial prefrontal cortex in, ?the capacity to monitor and be aware and be aware
of one?s reactions to the world and then use this self-awareness to stimulate the
awareness of others? which is quite similar to the more recent conceptualizations of
executive functioning (Nichols et al., 2005, p. 47). Although, related to the current study,
it is important to note the TOH task is thought to tap the dorsalateral prefrontal cortex.
In another directly related study, NICHD Early Child Care Research Network
(2005a) examined the link between the quality of children?s early childhood experiences
(including maternal and caregiver sensitivity and stimulation) and the development of
attention, memory, and planning, specifically examining the influence of the first 3 years
of a child?s experiences on their outcomes. Using hierarchical regression the authors
reported that there was a significant cumulative influence of the quality of the child-
rearing environment on attention and memory but not on planning. Also, importantly,
they noted that the family environment, as opposed to the child care and school
environment, was more strongly predictive of attention and memory. Finally, the authors
reported that quality of the child-rearing environment during both early (6 to 36 months)
and later (54 months and 1
st
grade) predicted performance on measures of attention and
memory but not planning.
50
The current study is based, in part, on a study by Landry, Miller-Loncar, Smith,
and Swank (2002) which examined the influence of maternal scaffolding at ages 3 and 4
on the development of executive function at age 6. The study also assessed students?
performance on a set of skills (language, nonverbal problem-solving, and memory) that
are theorized to be perquisite for the development of executive function. Specifically, the
study measured language, memory, and nonverbal problem solving at age 4. The study
was part of a larger longitudinal study that examined parenting behaviors and preterm
infants? outcomes and the study included both term (n=90) and preterm (n=163) children.
Within the current study scaffolding was measured by frequency of occurrence
during a 70 minute observation within the family home. For the first ten minutes of the
observation mothers were instructed to play with toys provided by the researchers. The
remaining sixty minutes of the observation parents were instructed to go about their
typical routine and observers recorded their behavior. Researchers were trained to
consider the nature of the content of maternal verbal support to determine if it qualified
as scaffolding. Maternal verbalizations were coded as scaffolding when, ?they provided
conceptual links between objects, persons, activities, or functions. This form of verbal
input was considered more facilitative of problem-solving skills because it begins to
place more responsibility for problem-solving on the child? (Landry et al., 2002, p. 21).
Children?s receptive and expressive language skills were assessed using the
Clinical Evaluation of Language Fundamentals at ages 4 and 6. Nonverbal problem
solving ability was assessed using the Nonverbal Scale Score from the Stanford Binet
Intelligence Test, fourth edition. Memory skills were assessed using the Memory Scale
Score also from the Stanford Binet. Executive function was assessed using two tasks,
51
Spatial Reversal Task and independent goal directed play behavior task. The Spatial
Reversal Task is a traditional executive function task that requires individuals to identify
the rule, inhibit incorrect responses, and demonstrate flexibility to shift set over time. The
independent goal directed play behavior task required the child to develop play goals and
strategies for achieving those goals. Scores were assigned based on the complexity and
length of play.
Using structural equation modeling, the researchers reported a number of direct
and indirect relationships. For example, they reported that maternal scaffolding at age 3
indirectly influenced executive functioning (i.e., Spatial Reversal Task) by directly
influencing language abilities at age 4. Thus enriched verbal scaffolding at a young age
increased language development which in turn predicted future executive functioning
development. The path from language skills to the independent goal directed play
behaviors task was non-significant possibly due to the fact that the task has fewer rule-
based strategies as compared to the Spatial Reversal Task.
Maternal scaffolding at age 3 also indirectly influenced executive function (i.e.,
independent goal directed play) via nonverbal problem solving skills at age 4. Thus it
appears that the scaffolding at age 3 strengthened nonverbal problem solving. The
authors hypothesize that when a mother is helping the child identify salient features of a
specific problem (e.g., the shape of a puzzle piece is important) those skills are later
generalized into nonverbal play tasks. The path from nonverbal skills to the Spatial
Reversal Task was non-significant possibly because the ability to mentally represent
objects appears to be more closely related to language ability.
52
Another cognitive factor that was examined, but did not influence executive
processing, was short-term memory. Not surprisingly, maternal scaffolding significantly
influenced short-term memory skills. The lack of relationship between executive function
and short-term memory is less clear as other empirical findings have supported a
relationship (Miayke et al., 2000; Senn et al., 2004). Also of significance within the
model was that 4-year old scaffolding did not directly or indirectly influence executive
function. This suggests that there is may be a sensitive period when scaffolding provides
unique support for the development of executive function.
The findings for the Landry et al. (2002) study are relevant to the current study for
several reasons. First, it is one of the few studies to identify, ?common paths of direct
influence from mothers? directness to both goal-directed and initiating skills but different
paths of indirect influence? (Landry et al., 2002, p. 370). Secondly, the study supported
the concept of a sensitive period for development which is in contrast to the results of the
NICHD Early Childcare Network (2005a) study. Based on these studies, the current study
will examine the direct and indirect impact of maternal support at 36 months and 54
months. As mentioned in the previous section of this literature review a major limitation
of the study was, ?the absence of any measure of mothers? intellectual abilities, genetic
backgrounds, or both in the model may raise questions as to whether children?s cognitive
outcomes are related, in part, to shared genotypes between parents and children? (Landry
et al., 2002, p. 37). The current study will include a measure of maternal verbal ability
and consider the role ethnicity in order to better examine the influence of maternal
support on executive function.
53
The current empirical and theoretical research indicates a potential relationship
between the development of maternal behaviors and executive function. Research
supports the concept that the quality and frequency of the maternal interactions in early
childhood has long-term implications. It has been suggested that during these mother-
child interactions children learn problem solving skills. Although executive function is
central to everyday experiences, little is known about what influences its development.
Using data from the NICHD SECCYD and building on the current theoretical and
empirical knowledge, the current study used structural equation modeling to examine the
relationship between the development of executive function and maternal behaviors.
54
CHAPTER III
Methodology
Participants
As part of a 1991 research NICHD SECCYD study, 1364 participants were
recruited from 10 research sites throughout the United States (Little Rock, AR, Orange
County, CA, Lawrence, KS, Boston, MA, Pittsburgh, PA, Philadelphia, PA,
Charlottesville, VA, Seattle, WA, Morgantown, NC, and Madison, WI). Although the
NICHD SECCYD is a longitudinal study that followed individuals from birth through
10
th
grade, the current study only focused on data from birth to 1
st
grade. The overall
NICHD SECCYD sample was selected based on a conditional random sampling plan that
focused on the maternal employment and demographic diversity within each of the 10
research sites. The NICHD SECCYD study was designed to include mothers who
intended to work full-time (60%) or part time (20%) in the first year of the child?s life as
well as mothers who intended to stay at home (20%). Participants were excluded from the
NICHD SECCYD study, and by extension the current study, if their mothers were less
then 18 years of age at the time of birth, if the family did not plan to stay within the site
area for three years, if there were obvious disabilities at birth, if the child was
hospitalized more than seven days following birth, if the mother had a known substance
abuse problem, or if the mother was not sufficiently conversant in English (NICHD Early
Child Care Research Network, 1993). It is important to recognize that the above
conditional sampling plan based on maternal employment and the exclusion criteria were
pre-existing decisions within the archival dataset. Obviously, these pre-existing decisions
within the NICHD SECCYD dataset limited the generalizability of the overall dataset, as
55
it was not designed to be nationally representative, and did not include paternal measures
thus limiting the current study to examining maternal factors.
Four hundred and seventy participant?s data were included in the current study
with roughly equivalent number of females (n=238, 50.6%) and males (n=232, 49.4%).
The ethnicity of the participants in the current study was the following: 87.9% were
White, non-Hispanic, 5.5% were African American, non-Hispanic, 4.3% were Hispanic,
1.5% were Asian-Pacific, less than 1% were American Indian, and .4% were classified as
Other. While the current study has similar proportion of males and females as the overall
NICHD study, there are notable differences in the breakdown of racial/ethnic
membership as compared to the overall NICHD study. The ethnicity of the participants of
the overall NICHD study consisted of 75% White, non-Hispanic, 13% being African
American, non-Hispanic, 6% being Hispanic, 1% being Asian-Pacific, and less than 1%
being American Indian, and 3% being Other. While it is somewhat unclear as to what
caused these differences, there appeared to be differential attrition rates such that more
ethnic minorities, namely African Americans, had either missing data or chose not to
remain in the study. As mentioned above the current study was not designed to be
nationally representative but the small number of participants within each ethnic minority
group makes multi-group comparison difficult and decreases the generalizability of the
current findings.
Examining Table 2 it appears that there were some notable differences between
the current study sample, the total NICHD study sample, and data from 1991 U.S.
Census. As noted above, these differences are attributed to differential attrition and
missing data rates as the sample for the current study only included individuals with full
56
set of data. In a review of the NICHD study (NICHD Early Child Care Research
Network, 2001) researchers compared the demographic characteristics of the families in
the NICHD sample to the 1991 U.S. Census data. The results suggested that the families
in the NICHD study were generally more well-educated, more likely to receive public
assistance, and less likely to include Hispanic participants then the actual U.S. population
(see data in Table 1). The data from participants in the current study represent a further
departure from data reported in the U.S. census such that it under-represents minorities,
particularly African Americans and Hispanics, and over-represents individuals with
higher education.
57
Table 1.
Demographic Variables for participants and families in the Current Study sample
(N=470), NICHD Study sample (N=1364), and the entire United States.
Current
Study sample
NICHD Study
Sample
1991 U.S.
a
Census Data
Female 238 (50.6%) 659(48.3%) N/A
Gender
Male 232 (49.4%) 705(51.7%) N/A
American Indian 2 (.4%) 5(.4%) .9%
Asian American 7 (1.5%) 20(1.5%) 3.6%
Hispanic 20 (4.3%) 90(6.6%) 15.3%
African American 26 (5.5%) 175(12.8%) 15.7%
Caucasian 413 (87.9%) 1023(75.0%) 64.9%
Ethnicity
Other 2 (.4%) 51(3.7%) N/A
Poor 29 (6.2%) 181(13.4%) N/A
Near Poor 65 (13.8%) 218(16.1%) N/A
Income-to-Needs
Ratio
Not Poor 376 (80.0%) 956(70.5%) N/A
Mean Income $43,079.88 $37,781.28 $36,875.31
< 12 years 15 (2.3) 142(10.4%) 24.6%
High School/GED 83 (17.7%) 288(21.1%) 30.1%
Some college 163 (34.7%) 453(33.2%) 26.7%
Bachelor?s Degree 128 (27.2%) 284(20.8%) 12.3%
Mother?s Education
Graduate school 81 (17.3%) 198(14.5%) 6.3%
a
US Bureau of the Census 1991 birth data.
58
Procedures and Materials
Socioeconomic Status (SES)
SES is a single composite score composed of the following three indicators:
maternal educational level, income-to-needs, and neighborhood safety. SES is defined as
the amount and quality of economic resources available to an individual. While SES
could have been measured through the use of formative indicators because of the overall
instability within the model a single composite score was used (See Kline, 2006).
Maternal education level (1 month). Level of maternal education was assessed
when the child was 1 month old. The current study will measure maternal education level
using the number of years of school attended.
Income-to-needs (6 to 36 months). This was calculated by dividing total family
income by the poverty threshold for the appropriate family size as defined by the U.S.
Census Bureau in 1991. Income-to-needs ratios less than 1 indicate poverty status with
scores of 5 and above indicate relative affluence. The current study will use a composite
score to represent the average income-to-needs ratio based on total income over 1 to 36
months.
Neighborhood Safety (1
st
grade). The current study used the Neighborhood Safety
scale from the Neighborhood Satisfaction and Involvement Questionnaire to assess
maternal perceptions of neighborhood safety. The Neighborhood Safety scale includes 5
questions that are originally asked on a 7-point likert scale and then rescaled by the
NICHD Early Child Care Network to a 10-point scale. The Neighborhood Safety scale
was administered to participant?s mothers at 1
st
grade. The Neighborhood Satisfaction
and Involvement Questionnaire was originally developed for the multi-site FAST Track
59
Intervention Project (Greenberg et al., 1999). Vandell and Piere (1998) examined the
validity of the Neighborhood Questionnaire and reported that maternal perception of the
Neighborhood Safety was negatively correlated (r = -.24 to -.58) with the police records
of property and violent crimes in that neighborhood over the course of 3 years. The
alphas for the overall NICHD study are r=.74 and .71 for the Neighborhood Safety scale
and the Neighborhood Social involvement scale, respectively.
Maternal Sensitivity and Maternal Stimulation
Mother-Child Interaction Task. In the current study maternal sensitivity and
maternal stimulation were measured during the Mother-Child interaction task using three
scoring systems. The Mother-Child interaction task involves three activities during a
semi-structured 15 minute interactions, although only 12 minutes of useable interaction
was required to be validly scored. Activities chosen for the 24, 36, and 54 month
interactions were similar to those used in other studies (Hess & Shipman, 1965). Two of
the three activities were designed and chosen because they were too difficult to be carried
out independently by the child and required maternal assistance and support. The third
activity was selected to encourage pretend play.
At the beginning of the task the mother is presented with three sequentially
numbered boxes and instructed to help their child play with the items in the three boxes
for 15 minutes. The toys were selected based on three criteria. The toys must be
interesting and developmentally appropriate, foster different types of activities, and
general enough to be used differently by children and their mothers (NICHD, 2001).
60
At 24 months, the first activity was a picture book. The second activity was a
model kitchen that included kitchen accessories. The third activity was a discovery
cottage which included toy figurines.
At 36 months, the first activity was to coordinate the use of stencils thus the first
box contained markers, a pad of blank paper, and 4 stencils of animals. The second
activity was developed to stimulate pretend play thus the second box contained dress up
clothes (men?s shoes, women?s shoes, a vest, and a cape) and a cash register. The final
activity required collaborative play thus the third box contained a Duplo Building Set
with a laminated color picture of the model design.
At 54 months, the first activity required completion of a maze using an Etch-A-
Sketch thus the first box contained an Etch-A-Sketch which was altered by attaching a
maze to the screen. The goal of the second activity was to form rectangular cube towers
from wooden blocks thus the second box contained irregularly shaped wooden blocks.
The goal of the third activity was pretend play thus the third box contained six animal
hand puppets (2 parrots, 2 frogs, and 2 alligators).
Task Scoring. From these taped interactions at 24, 36, and 54 months data were
coded using three different methods.
The first method, called the mother-child interaction task scoring at 24 months,
was developed specifically for this study and was modified from Egeland and Hiester
(1993). The mother-child interactions at 24 months were rated by trained coders on a 4-
point global rating scale with 1 = ?Not at all characteristic? to 4 = ?Highly
characteristic?. Inter-rater reliability data from selected tapes from the overall NICHD
data set (n=212) of the 24 month mother-child interaction task scoring indicated Pearson
61
correlation coefficients of .67 for the maternal sensitivity composite at 24 months and .57
for the maternal stimulation scale at 24 months.
The second method, the mother-child interaction task scoring at 36 and 54
months, used a 7-point global rating scale with 1 = ?Very Low? to 7 = ?Very High?.
Inter-rater reliability data from selected tapes from the overall NICHD data set (n=242)
of the 54 month mother-child interaction task scoring indicated Pearson correlation
coefficients of .78 for the maternal sensitivity composite at 54 months and .68 for the
maternal stimulation scale at 54 months. The internal consistency scores of the composite
profiles indicate maternal sensitivity r =.84.
The third method, mothers as mediators of cognitive development scoring,
involved only the 36 month data and examined 28 behaviors (e.g., labeling object,
narrating child?s play, asks question) related to mother?s behavior mediating cognitive
development. This coding system, which was developed by Friedman & Sherman (1985)
for the NICHD study, was applied to 5 minutes of the 15 minutes videotaped session of
mother-child interaction at 36 months. Coding was started when the mother and child
started using the second box. The coding was carried out using a time sampling
technique of observing and then coding each of 20 15-second segments of the videotape.
From this system a behavioral composite was developed (Total Positive Scaffolding).
From the overall data set, reliability data as coded by two trained observers was obtained
for 63 children. Pearson correlation coefficient for the behavior composite (Total
Positive Scaffolding) was reported at 0.96 (NICHD, 1999). Within the overall NICHD
dataset, scores ranged from -2 to 232 with higher scores indicating more positive
62
scaffolding. The 25 items used to create the total positive scaffolding score had low
internal reliability (Cronbach?s alpha = 0.45).
Early Maternal Stimulation (24 and 36 months). The current study will use one
scale score (cognitive stimulation) from the 24 month mother-child interaction task
scoring, one scale score (cognitive stimulation) from the 36 month mother-child
interaction task scoring, and one composite score (positive scaffolding) from the mother
as mediator of cognitive development scoring.
Early Maternal Sensitivity (24 and 36 months). The current study will use one
scale score (positive regard) from the 24 month mother-child interaction task scoring and
two scale scores (respect for autonomy and supportive presence) from the 36 month
mother-child interaction task scoring.
Late Maternal Stimulation (54 months). The current study will use three scale
scores (cognitive stimulation, goal directed partnership, quality of assistance) from the
mother-child interaction task scoring for maternal stimulation at 54 months.
Late Maternal Sensitivity (54 months). The current study will use two scale scores
(supportive presence and respect for autonomy) from the mother-child interaction task
scoring at 54 months.
Maternal Verbal Ability
Peabody Picture Vocabulary Test - Revised (PPVT-R) (36 months). The current
study will use the Peabody Picture Vocabulary Test - Revised (PPVT-R) edition form M
to measure maternal verbal ability (Dunn & Dunn, 1981). The PPVT-R was designed for
individuals aged 2 ? to 90 years of age. The task requires receptive knowledge of
vocabulary and involves items of increasing difficulty. The task was administered to the
63
participants? mothers during the 36 month lab visit. The task takes about 10 to 20
minutes to complete although the task is un-timed. Participants are presented with plates
that include four pictures on each plate and are asked in indicate which picture best tells
the meaning. Reliability for the PPVT-R comes from a nationally representative adult
sample (n=800) based on age, gender, geography, and occupation. Internal spilt-half
consistency ranged from .80 to .83.
Research using the PPVT-R with typically functioning elderly adults (n=91) as
measure of intelligence indicates that it significantly correlated with the short form
WAIS-R FSIQ (r = .61) (Snitz, Bieliauskas, Crossland, Basso, & Roper, 2000). Other
researchers (Altepeter & Johnson, 1989) reported that the PPVT-R correlated positively
and significantly with the WAIS-FSIQ r =.47, the WAIS-R Verbal r = .53, and the
WAIS-R Performance r = .24 thus demonstrating convergent and divergent validity.
Finally Carvajal, Kixmiller, Knapp, Vitt, and Weaver (1991) examined the relationship
between the PPVT-R and the Stanford Binet 4
th
edition for college students (n=31) and
reported a positive correlation (r=.59).
Language
Reynell Developmental Language Scale (RDLS) (36 months). The current study
will use the Reynell Developmental Language Scale (RDLS) (Reynell, 1990). The RDLS
was developed as a test of verbal comprehension and expressive language skills and was
designed to detect changes in language development in children aged 1 to 7. The test was
administered during the 36 month lab visit. The RDLS has two 67-item scales: Verbal
Comprehension and Expressive Language and produces two separate scores. Reliability
for the RDLS comes from a nationally representative sample (age, gender, geography,
64
and parental education) based on data from the 1987 U.S. census with participants
(n=619) aged 1 year to 6 years 11 months. Of the overall RDLS national sample, nearly
100 (n=99) participants were between the age of 36 and 48 months. Internal split-half
reliability for the Verbal Comprehension form (A) was r = .93 and for Expressive
Language was r = .86. Inter-scale correlations between the Verbal Comprehension and
Expressive Language scale was .76.
Validity data comes from several sources. Both the Expressive Language and the
Receptive Language scales highly correlated with the Language Development Survey
(LDS) at .71 and .56 respectively (Rescorla & Alley, 2001). Additionally, Expressive
Language (.78 and .81) and Receptive Language (.56 and .58) were highly correlated
with the LDS over time (23 day test-retest). The Expressive Language scale demonstrated
discriminate validity between at risk toddlers (n=33) and a matched (gender and SES)
control group (n=33). There were significant group differences t(40.18) = 4.97, p<.001
with the at risk participants scoring about 6 months below average. Other studies reported
that the Verbal Comprehension scale and the Predictive Screen Test of Articulation
correlated at r = .43 and the Expressive Language scale and the Predictive Screen Test of
Articulation correlated at r = .56 (Reynell, 1990). The Verbal Comprehension scale and
the Goodenough-Harris Draw-a-Man Test correlated at .51 and the Expressive
Comprehension scale and the Goodenough-Harris Draw-a-Man Test correlated at .46.
Predictive validity was reported between the British version of the RDLS and the WISC.
Predictive validity between the Receptive Language Scale at 7 years of age was .71 for
the WISC Full Scale and .69 for the WISC Verbal. Predictive validity for the Expressive
65
Language Scale at 4 years of age with the WISC at 7 years of age was .63 for the WISC
Full Scale and .67 for the WISC Verbal Comprehension.
Preschool Language Scale?3 (PLS-3) (54 months). The current study will use the
Preschool Language Scale?3 (PLS-3) which was designed to assess vocabulary,
grammar, morphology, and language reasoning (Zimmerman, Steiner, & Pond, 1992).
The test was developed for use with children between 2 weeks and 6 years, 11 months of
age. The test was administered during the 54 month home visit within the current study.
The PLS-3 has 2 scales: Auditory Comprehension and Expressive Communication and
produces two separate scores as well as a Total Language score. The manual reports test-
retest stability coefficients ranging between .82 and .92 for subscales and .91 to .94 for
the Total Language score (Zimmerman et al., 1992). Internal consistency scores range
from .47 to .90. Concurrent validity scores between the PLS-3 with the PLS-R were
reported at .66 for the Auditory Comprehension Scale and .86 for the Expressive
Language Scale.
Other research with the PLS-3 indicates its reliability and validity for use with
African American children. Qi, Kaiser, Milan, Yzquierdo, and Hancock (2003) examined
the performance of low SES African American preschoolers aged 36 to 52 months
(n=701) and a comparison group of European American preschoolers (n=50). The
authors reported PLS-3 internal consistency for Auditory Comprehension = .86,
Expressive Communication =.87, and Total Score = .92. The authors also reported that
the total PLS-3 score correlated with the Expressive Vocabulary Test r=.48 for African
American participants and r=.64 for European American participants. Although the
African American participants performed about 1 SD below the expected means, t tests
66
showed no significant differences between the two groups. Overall, the authors noted
concerns about the use of the PLS-3 as a sole clinical indicator of a language disability
for African American children but reported that the PLS-3 ?produces a range of scores
that are meaningfully related to other measures of language use? (Qi et al., 2003, p. 589).
Woodcock-Johnson Picture Vocabulary Subtest (54 months). Children?s
vocabulary performance was assessed at 54 months using the Picture Vocabulary subtest
of the Woodcock-Johnson Psychoeducational Battery-Revised (WJ-R; Woodcock &
Mather, 1989). The task requires participants to name familiar and unfamiliar pictured
items. The test has a mean score of 100 with a standard deviation of 15. Within the
overall NICHD data sample the internal task reliability was .76.
Attention
Solitary Play. (36 months). Focused attention (i.e., engagement with objects
through exploration) was measured by presenting the participants with different age
appropriate toys. The participants were seated on the floor and video taped for 15
minutes. The interactions were coded for both visual and tactile exploration using a
voice-activated computer coding system (Breznitz & Friedman, 1988). The coder
inputted information about the child?s engagement and the computer automatically
recorded the amount of time the child attended to each object. The current study will use
average time attending to a toy and total time off-task.
Although validity data are not available for the current measure because the
solitary play task was developed specifically for the NICHD study, this methodology
(i.e., direct observation of attention) is a commonly used and accepted method in
developmental research (Barkley, 1994). Other researchers have reported that focused
67
attention is typically consistent across situations and reliable across objects (Ruff &
Rothbart, 1996).
Child Behavior Checklist (CBCL) (54 months). Mothers and an alternative care
giver completed the CBCL (Achenbach, 1991). The parent version contains 118 items
that describe a range of child behaviors and emotions. Mother?s rate how well the item
describes their child in the last six months based on the following scale: 0 = not true, 1 =
somewhat or sometimes true, and 2 = very true or often true. Higher scores indicate
overall more problems. Standard scores range from 50 to 100. The Attention Problems
Subscale contains 11 items and purports to measure impulsivity and inattention (e.g.,
can?t pay attention for long). The CBCL has high 1-week test?retest reliabilities (ranging
from .63 to .97) and high internal consistency (ranging from .66 to .92) (Achenbach,
1991). Convergent validity for the CBCL Attention Problems scale score was reported
with the Child Symptom Inventory?4 (CSI?4) ADHD combined type (r=.65).
Achenbach, Howell, McConaughy, and Stanger (1995) reported that the Attention
Problems scale score of children assessed at age 4 through 12 years via parent reports and
reassessed at 3 and 6 years via parent, teacher, and self-reports and were found to be
developmentally stable.
Delay of Gratification (DOG) (54 months). The Delay of Gratification (DOG)
task is administered in the laboratory during the 54 month visit. The task is designed to
assess a child?s ability to delay gratification and attend to a task by offering the child an
option between waiting for the bigger food prize or obtaining a smaller food prize by not
waiting. The current study uses the amount of time waited. While it is necessary to
68
recognize the DOG task does include aspects of inhibition (see Welsh & Pennington,
1988), the ability to attend to the task is also thought to be a central aspect of the task.
After choosing which of three foods the child most likes, M&Ms, animal
crackers, or pretzels, the child is offered a choice between waiting (for 7 minutes) until
the examiner returns to the room to receive a larger quantity of their chosen food or
ringing a bell and to bring the examiner back into the room and receive a smaller amount
of food. Both larger and smaller amounts of food are placed in front of the child.
A number of studies have examined the relationship between quality of
environment and the capacity to delay food reward. Several studies have reported that
children in lower SES environments or children with less responsive parents are
significantly less able to delay reward (Aber, Rodriguez, Michel, & Shoda, 1995). Other
researchers (Shoda, Mischel, & Peake, 1990; Sethi, Mischel, Aber, Shoda, & Rodriguez,
2000) have reported long term positive effects of preschoolers who can delay
gratification. Specifically they report more self-regulation and greater ability to cope
during adolescents.
Continuous Performance Task (CPT) (54 months). At 54 months children
completed the CPT (Rosvold, Mirsky, Sarason, Bransome, & Beck, 1965). This
individually administered task is presented at the end of the 54 month lab visit. The task
required participants to press a button ?as fast as you can? when a target stimulus (i.e., a
familiar object) was presented on a screen. 220 total stimuli were presented in 22 blocks.
Within each block the target stimulus were randomly presented but appeared twice within
each of the 22 blocks. Each stimulus appeared on the screen for 500 msec. The task
required the participant to not only press the button when he/she saw the target stimulus
69
but to also inhibit pressing the button when other non-target stimuli are presented. The
task takes 7 minutes and 20 seconds to complete.
The current study will use the errors of omission score to measure sustained
attention. Errors of omission occur when the participant fails to press the button when the
target stimulus is presented. It is expected that children with greater ability to sustain
attention would have fewer errors of omission thus lower scores indicate higher levels of
sustained attention. Reliability data from 138 non-referred 7-11 year old ethnically
diverse males reported adequate test-retest reliability (r = .65 - .74) (Halperin, Sharma,
Greenblatt, & Schwartz, 1991). The task has been found to be sensitive to individual
differences and group differences and widely researched with individuals with attention
deficit/hyperactivity disorder (Barkely, 1994). The CPT was reported to be significantly
related to the reading/decoding ability as measured by the WRAT-R but not related to
vocabulary as measured by the PPVT-R (Halperin et al., 1991).
Executive Function
Tower of Hanoi (TOH) (1
st
grade). The TOH is reviewed extensively in Chapter 2
therefore the following is a brief overview with more detail about administration and
scoring. The current study used the TOH as one measure of the child?s executive
functioning ability. The TOH task consists of moving four colored rings of increasing
diameters among three vertical pegs. Participants are presented with an initial tower
configuration with the goal being to move the rings among the pegs to construct a tower
(largest rings on the bottom and smallest on the top) on a specified peg. The following
rules were employed: (a) only one disk is allowed to be moved at a time; (b) a disk
cannot be set on the table or held in the participants hand during the movement of another
70
disk; (c) a larger disk cannot be set on top of a smaller disk. Participants have unlimited
time to complete the task. In order to be successful participants must construct the tower
in the fewest number of moves. Participants were presented with several practice
problems and a set of six test problems.
Several children (n=11) were not able to complete the practice items and were
excluded from the analyses for that reason. The 6 test problems ranged in degree of
difficulty from involving three disks and four moves for completion (least difficult) to
including four disks and eleven moves (most difficult). The problems were presented
from least to most difficult. The child must successfully complete a problem twice in
succession before he or she progressed to a higher level of difficult. Successful
completion of a problem was defined as transforming the initial configuration into the
goal configuration in the fewest possible number of moves while not violating any rules.
A single problem was presented to the child a maximum of 6 times. The child had a
maximum of 20 moves per trial to try to solve the problem.
Participants were given several scores but the current study will use the Total
Efficiency Planning score as it reflects the number of trials needed to successfully
complete a task and provides an overall performance score (Borys, Spitz, & Dorans,
1982). As mentioned above students must get two successive trials correct, thus the
maximum score of 6 is assigned if the problem is successfully completed in the first or
second trials; a score of 5 is assigned if the problem is successfully completed in the
second and third trials; a score of 4 is assigned if the problem is successfully completed
in the third and fourth trials; a score of 3 is assigned if the problem is successfully
completed in the fourth and fifth trials; a score of 2 is assigned if the problem is
71
successfully completed in the fifth and sixth trials; and a 0 is assigned for no consecutive
correct solutions (a score of 1 cannot be obtained).
The scores on the individual items were summed to yield the Total Efficiency
Planning scores. Scores in the NICHD dataset ranged from 0 to 34 with higher scores
indicating greater executive functioning. The items used to create the Total Efficiency
Planning score have moderate internal reliability (Cronbach?s alpha = .70). In the current
NICHD dataset, correlations between the TOH and the Continuous Performance Task
(CPT) ranged from -.155 to .123 (Welsh, 2005). Examining the relationship further,
participants who had a greater number of omission errors on the CPT also had lower
scores on the TOH. Other researchers observed a significant positive relationship, r = .36,
p<.05, between one working memory test (Letter-Number-Sequencing from the WAIS)
and performance on the Tower of Hanoi-Revised. A positive significant relationship (r =
.549, p< .05) between Matrix Reasoning (a measure of nonverbal reasoning/fluid
intelligence similar to Raven?s Matrices) and the Tower of Hanoi-Revised was also
reported (Lock et al., 2002).
Memory for Sentences (MS) (1
st
grade). The MS was administered in the 1
st
grade. Memory for Sentences, which is part of the Woodcock-Johnson Psychoeducational
Battery-Revised (1989), measures short-term memory as well as comprehension
knowledge. The task involves participant?s recalling and repeating simple words, phrases,
and sentences which were presented via tape player. The task also required that the
participant use sentence meaning to aid recall. The normative data for this sub-test are
based upon a nationally standardized sample of 6,359 subjects, aged 24 months to 90
years of age. Split-half reliability for the MS task is reported to be .897 (McGrew, 1986).
72
Memory for Names (MN) (1
st
Grade). The MN was administered in the 1
st
grade.
Memory for Names, which is part of the Woodcock-Johnson Psychoeducational Battery-
Revised (1989), measures long-term retrieval. The task requires participants learn
associations between unfamiliar auditory and visual stimuli (a space creature). The
participant is presented with an unfamiliar stimuli, told the stimuli?s name, and then
asked to recognize and name the stimuli from a sheet of nine space creatures.
Participant?s errors are corrected. The level of difficulty increased as the task continued.
The normative data for this sub-test are based upon a nationally standardized sample of
6,359 subjects, aged 24 months to 90 years of age. Split-half reliability for the MN task is
reported to be .909 (McGrew, 1986).
Continuous Performance Task (CPT) (1
st
Grade). The CPT was administered in
the 1
st
grade. The administration procedure was very similar to the 54 month procedure
expect that the stimuli were presented in 30 blocks with ten stimuli in each block. Each
stimulus was displayed for 200 msec with 1500 msec between stimuli. The target
stimulus (a dot matrix letter X) was randomly presented within each block and appeared
twice in each block. The test took 350 sec. to administer. The current study will use the
errors of commission score to measure inhibition. Commission errors are generally seen
as reflecting disinhibition. Much the research to support the use of commission errors has
come from research related to ADHD (see Riccio et al., 2001 for review).
Data Analysis
The major goal of the current study was to examine the relationship between
maternal behaviors and the future development of executive function. Structural equation
modeling was chosen as it provided more flexibility than multiple regression, controlled
73
for measurement error, and provided an overall model fit (Frazier et al., 2004). As
recommended by Hoyle (1995) confirmatory structural equation modeling, as opposed to
exploratory modeling, was used in the current study. Jorekog and Sorbom (1996 as cited
in MaCallum & Austin, 2000) recognize three strategies in confirmatory modeling. The
first strategy is described as being strictly confirmatory where a single a priori model is
specified and recognized. This strategy as not ideal as it requires the researcher to
evaluate a single model in isolation and rules out few possibilities if that model does not
fit well. The second method is model generation and involves specifying a single model
and then modifying that model as warranted to best fit the data. This method is
discouraged as it is misleading and frequently leads to misinterpretation.
Finally, the alternative model approach, which is the method that was used in the
current study, involved the specification of several a priori models prior to estimation of
the target model so that nested comparisons could be made. Although all three methods
of structural equation modeling were considered for the current study, the literature
suggests that,? the strongest SEM analysis proposes a target model based on careful
consultation of relevant theory and prior research and then compares that model with one
or more previously specified competing model indicated by other theoretical positions,
contradictions in the research literature, or parsimony? (Hoyle & Panter, 1995, p. 171).
MacCallum and Austin (2000) note that the use of multiple a priori models will reduce
the potential for confirmation bias of the targeted model and write that, ?any effort to
examine alternative models can provide some protection against a confirmation bias and
bolster support of a favored model? (2000, p. 213). Within the current study, latent
variables were used as they provide more accurate and stable measures of effects of one
74
variable on another by estimating errors of measurement separately from the model
(MaCallum & Austin, 2000).
Within the current study, several constructs had only a few indicators (e.g., early
language and early attention). Research related to the number of the indicators necessary
to arrive at proper solutions suggests that one is more likely to arrive at a proper solution
with more accurate parameter estimates. Additionally, there is greater reliability with an
increasing number of indicators. But, importantly, researchers have also reported that a
larger N, such as the size of the current study, compensated for a small number of
indicators. While the best practice would have been to have both a large N and a large
number of indicators, the archival nature of the data limited that possibility (Marsh, Hau,
Balla, & Grayson, 1998).
In structural equation modeling there are several fit indices that are considered
when examining a model but there is no ?universally accepted criterion to judge how will
the model fits the data, thus leaving room for subjective opinions, and consequently
disagreements? (Crowley & Fan, 1997, p. 515). Given that there is not a single accepted
criterion in assessing model fit the current study considered several fit indices. First two
parsimonious fit indices were used to determine each of the three competing models
badness of fit Akaike Information Criteria (AIC) and Root Mean Square Error of
Approximation (RSMEA). The AIC was chosen because it takes both fit and estimated
parameters into account and is able to be used to compare two or more models. RSMEA
was chosen because it is relatively insensitive to sample size, adjusts for parsimony, and
provides confidence intervals. Parsimonious fit indices are necessary because the current
study is using a nested series of models and the fit of the model always improves with
75
fewer parameters. For example, if one parameter is freed from each observed variance
then a better fit can be achieved but the model may have little overall value. Thus the
parsimonious fit indices penalize the use of additional parameters and indicate which
model is best fitting and most parsimonious.
After determining the most parsimonious model, the next step will be to examine
incremental fit indices. Incremental fit examines the degree to which the model in
question is superior to an alternative model. Specifically the study will use Chi square
probability (which assesses the model?s fit to the data by assuming all the variable are
exogenous and unrelated), Standardized Root Mean-Square Error (represents the average
value across all standardized residuals), and Bentler?s Comparative Fit Index (alternative
fit index that assess the degree to which the model fails to fit the data).
Current Models
In the current study a series of nested and non-nested models were specified to
represent different hypothesized relationships. Those relationships were postulated based
on the available empirical evidence and theoretical rationales. See Table 2 for a list of
variables identified for inclusion into the study.
Model 1. The first model suggested that early maternal behaviors will indirectly
influence executive function at first grade through a direct influence on attention and
language skills at 36 months and an indirect influence on attention and language skills at
54 months. Socioeconomic status will have both direct and indirect (i.e., via attention,
language, maternal sensitivity, and maternal stimulation) effects on executive function in
the first grade. Maternal behaviors at 54 months will not influence executive function at
first grade.
76
Model 2. The second model suggested that early maternal behaviors will
indirectly influence executive function at first grade through a direct influence on
attention and language skills at 36 months and an indirect influence on attention and
language skills at 54 months. Socioeconomic status will have both direct and indirect
(i.e., via attention, language, maternal sensitivity, and maternal stimulation) effects on
executive function in the first grade. Maternal behaviors at 54 months will indirectly
influence executive function at first grade via attention and language at 54 months.
Model 3. The third model suggested that early maternal behaviors at 36 months
will indirectly influence executive function at first grade through a direct influence on
attention and language skills at 36 months and an indirect influence on attention and
language skills at 54 months. Socioeconomic status will have both direct and indirect
(i.e., via attention, language, maternal sensitivity, and maternal stimulation) effects on
executive function in the first grade. Maternal behaviors at 54 months will directly and
indirectly (via attention and language) influence executive function at first grade.
Ethnicity as a Moderator. Nested models will also be used to examine if the
relationship holds across two ethnic groups (Caucasian and African American) using the
most tenable model in the overall group analyses. These analyses will indicate whether
estimates of model parameters remained invariant across groups, or whether group
membership (Caucasian or African American) moderated the relationships specified in
the model (Frazier et al., 2004). It is hypothesized that ethnicity will have a moderating
effect on executive function at first grade.
Maternal Verbal Ability as a Mediator. Non-nested models will be used to
analyze maternal verbal ability as a mediator between SES and executive function
77
(Frazier et al., 2004). Using the most tenable model in the single group analysis,
comparisons will be made between the overall fit of that model and a model that includes
maternal verbal ability as a mediator. It is anticipated that maternal verbal ability will
serve as a partial mediator between SES and executive which will result in a stronger
predictive relationship between SES and executive function.
78
Table 2
Variables Identified for Inclusion in the Structural Equation Modeling
Latent Construct Indicator
F1 Socio Economic Status
V1 SES Composite Score
Composed of : Mother's Education at 1 Month
Neighborhood Safety at 1
st
grade
Total Income-to-Needs Ratio at 36 Months
F2 Early Maternal Sensitivity
V2 Positive Regard at 24 Months
V3 Supportive Presence at 36 Months
V4 Respect for Autonomy at 36 Months
F3 Early Maternal Stimulation
V5 Cognitive Stimulation at 24 Months
V6 Cognitive Stimulation at 36 Months
V7 Total Positive Scaffolding at 36 Months
F4 Early Language
V8 Reynell: Vocabulary at 36 Months
V9 Reynell: Expressive Language at 36 Months
F5 Early Attention
V10 Average Time Attending at 36 Months
V11 Total Time Off-Task at 36 Months
F6 Late Stimulation
V12 Cognitive Stimulation at 54 Months
V13 Quality of Assistance at 54 Months
V14 Goal Directed Partnership at 54 Months
F7 Late Sensitivity
V15 Supportive Presence at 54 Months
V16 Respect for Autonomy at 54 Months
F8 Late Language
V17 PLS: Auditory Comprehension at 54 Months
V18 PLS: Expressive Language at 54 Months
V19 WJ-R Picture Vocabulary at 54 Months
F9 Late Attention
V20 CBCL-Mother: Attention at 54 Months
V21 CBCL-Other: Attention at 54 Months
V22 Delay of Gratification at 54 Months
V23 CPT: Omission Errors at 54 Months
F10 Executive Functioning
V24 WJ-R: Long Term Memory at 1
st
Grade
V25 WJ-R: Short Term Memory at 1
st
Grade
V26 CPT: Commission Errors at 1
st
Grade
V27 TOH: Total Planning Efficiency at 1
st
Grade
F11 Maternal Verbal Ability
V28 PPVT-R: Mother?s Verbal Ability at 36 months
79
CHAPTER IV
Results
Before addressing the major research questions posed in this investigation,
descriptive statistics (e.g., mean, range, and standard deviation) are provided for all of the
variables used in the analyses. This will be followed by results for each of the three
research questions as well as additional analyses used to examine each research question.
Table A.1 in Appendix A contains the first order correlations among all study variables.
References to specific relationships will be addressed as it relates to each research
question.
Descriptive Statistics
Table 3 compares the means from individuals in the current study to the entire
NICHD study on indicators included in the study to examine if patterns of differences
arise as a result of differences in the sample. Examining these differences there appears to
be a general, stable trend across the two samples in which participants in the current
study have more desirable behavior (e.g., more expressive language ability) although,
more importantly across tasks, the effect sizes were small in size and had little practical
significance. This suggests that the current study sample is similar to the greater NICHD
study population.
80
Table 3
Descriptive Statistics and Effect Sizes for Indicator Variables included in the current
study sample (N=470) and NICHD Study sample (N=1364) (*p<.05)
Current
Study sample
NICHD Study
sample
Mean
Std.
Dev.
Mean
Std.
Dev.
Effect
Size
F1 Socio Economic Status
V1 SES Composite Score 26.77 5.15 26.40 5.43 0.07
Neighborhood Safety at 1
st
grade 8.25 1.30 8.09 1.50 0.11*
Total Income-to-Needs Ratio at 36 Months 3.85 3.22 3.38 2.70 0.17*
Mother's Education at 1 Month 14.67 2.32 14.23 2.51 0.18*
F2 Early Maternal Sensitivity
V2 Positive Regard at 24 Months 2.87 0.66 2.82 .70 0.07
V3 Supportive Presence at 36 Months 5.47 1.04 5.28 1.32 0.14*
V4 Respect for Autonomy at 36 Months 5.44 0.88 5.29 1.10 0.14*
F3 Early Maternal Stimulation
V5 Cognitive Stimulation at 24 Months 2.77 0.71 2.71 .74 0.08
V6 Cognitive Stimulation at 36 Months 4.64 1.35 4.47 1.44 0.12*
V7 Total Positive Scaffolding at 36 Months 119.64 36.90 114.35 40.19 0.13*
F4 Early Language
V8 Reynell: Vocabulary at 36 Months 100.70 14.08 97.85 15.85 0.18*
V9 Reynell: Expressive Language at 36 Months 99.07 13.52 96.88 14.53 0.15*
F5 Early Attention
V10 Average Time Attending at 36 Months 166.66 159.74 166.37 57.04 0.01
V11 Total Time Off-Task at 36 Months 28.59 29.23 28.08 30.34 0.02
F6 Late Stimulation
V12 Cognitive Stimulation at 54 Months 4.50 1.12 4.36 1.29 0.11*
V13 Quality of Assistance at 54 Months 4.79 1.25 4.66 1.39 0.09
V14 Goal Directed Partnership at 54 Months 4.85 1.27 4.72 1.35 0.10
F7 Late Sensitivity
V15 Supportive Presence at 54 Months 5.30 1.10 5.16 1.30 0.11*
V16 Respect for Autonomy at 54 Months 5.33 0.99 5.22 1.11 0.10
F8 Late Language
V17
PLS: Auditory Comprehension at 54
Months
101.48 18.32 98.34 19.92 0.16*
V18 PLS: Expressive Language at 54 Months 103.64 16.61 100.62 19.95 0.15*
V19 WJ-R Picture Vocabulary at 54 Months 102.59 13.25 100.24 15.03 0.16*
F9 Late Attention
V20 CBCL-Mother: Attention at 54 Months 2.55 2.10 2.74 2.40 -0.08
V21 CBCL-Other: Attention at 54 Months 2.57 2.22 2.67 2.36 -0.04
V22 Delay of Gratification at 54 Months 4.89 2.80 4.48 3.00 0.14*
V23 CPT: Omission Errors at 54 Months 8.06 6.77 9.13 7.60 -0.14*
81
Table 3 (continued)
Descriptive Statistics and Effect Sizes for Indicator Variables included in the current
study sample (N=470) and NICHD Study sample (N=1364) (*p<.05)
Current
Study sample
NICHD Study
sample
Mean
Std.
Dev.
Mean
Std.
Dev.
Effect
Size
F10 Executive Function
V24 WJ-R: Long Term Memory at 1
st
Grade 99.84 13.75 98.51 14.09 0.09
V25 WJ-R: Short Term Memory at 1
st
Grade 102.15 14.70 101.54 14.94 0.04
V26 CPT: Commission Errors at 1
st
Grade 5.13 7.96 6.11 10.64 -0.09*
V27 TOH: Planning Efficiency at 1
st
Grade 14.67 6.52 14.38 6.76 0.04
F11 Maternal Verbal Ability
V28 PPVT-R: Mother?s Verbal Ability 101.97 16.23 99.01 18.35 0.16*
In the current study, only participants who have a complete set of data available for
each variable and were able to successfully complete practice items for each task were
included in the data set. After applying these exclusionary criteria and because list-wide
deletion is the recommended approach with structural equation modeling, participants
with missing data were excluded from the modeling. An effort has been made to include
as many subjects as possible in the modeling, but there are obvious limitations. It is
acknowledge that listwise deletion can produce biased estimates of the effects for the full
sample if the people who are excluded from analysis are different in important ways from
those who are included. More specifically, the data in the current study may provide
unbiased results for the subset sample with complete data, but this sample may not be
representative of the population found in the NICHD SECCYD study.
Table 4 compares the means of the indicators from individuals in the current study
to means from individuals whose data was not included in current study as a result of
listwise deletion to examine if patterns of differences arise as a result of selection bias.
82
Examining these differences there appears to be a general, stable trend across the two
samples in which participants in the current study have more desirable behavior.
Examining the effect sizes, while the majority were statistically significant at a p<.05
level, the majority of the effect sizes were small and had little practical significance. This
suggests that there are differences between the current study and data excluded from the
current study but that these differences do not appear to have practical implications.
83
Table 4
Descriptive Statistics and Effect Sizes for Indicator Variables included in the current
study sample (N=470) and data excluded from the current study sample based on
exclusionary criteria (n=297 to 862) (*p<.05)
Current
Study sample
Excluded Data
Mean
Std.
Dev.
Mean
Std.
Dev.
Effect
Size
F1 Socio Economic Status
V1 SES Composite Score 26.77 5.15 25.5 5.54 .23*
Neighborhood Safety at 1
st
grade 8.25 1.30 7.95 1.62 .19*
Total Income-to-Needs Ratio at 36 Months 3.85 3.22 3.3 2.93 .19*
Mother's Education at 1 Month 14.67 2.32 13.9 2.53 .32*
F2 Early Maternal Sensitivity
V2 Positive Regard at 24 Months 2.87 0.66 2.75 .7 .17*
V3 Supportive Presence at 36 Months 5.47 1.04 5.07 1.41 .29*
V4 Respect for Autonomy at 36 Months 5.44 0.88 5.09 1.18 .30*
F3 Early Maternal Stimulation
V5 Cognitive Stimulation at 24 Months 2.77 0.71 2.62 .73 .21*
V6 Cognitive Stimulation at 36 Months 4.64 1.35 4.27 1.43 .26*
V7 Total Positive Scaffolding at 36 Months 119.64 36.90 108 40.9 .29*
F4 Early Language
V8 Reynell: Vocabulary at 36 Months 100.70 14.08 94.3 15.7 .41*
V9 Reynell: Expressive Language at 36 Months 99.07 13.52 94.3 14.7 .32*
F5 Early Attention
V10 Average Time Attending at 36 Months 166.66 159.74 166 146 .00
V11 Total Time Off-Task at 36 Months 28.59 29.23 27.8 29.9 .03
F6 Late Stimulation
V12 Cognitive Stimulation at 54 Months 4.50 1.12 4.11 1.35 .29*
V13 Quality of Assistance at 54 Months 4.79 1.25 4.43 1.44 .25*
V14 Goal Directed Partnership at 54 Months 4.85 1.27 4.54 1.37 .23*
F7 Late Sensitivity
V15 Supportive Presence at 54 Months 5.30 1.10 4.9 1.3 .31*
V16 Respect for Autonomy at 54 Months 5.33 0.99 5.02 1.14 .27*
F8 Late Language
V17
PLS: Auditory Comprehension at 54
Months
101.48 18.32 93.8 19.9 .39*
V18 PLS: Expressive Language at 54 Months 103.64 16.61 101.1 20.1 .15*
V19 WJ-R Picture Vocabulary at 54 Months 102.59 13.25 96.9 15 .38*
F9 Late Attention
V20 CBCL-Mother: Attention at 54 Months 2.55 2.10 2.78 2.2 -.10
V21 CBCL-Other: Attention at 54 Months 2.57 2.22 2.62 2.04 -.02
V22 Delay of Gratification at 54 Months 4.89 2.80 3.9 3.1 .32*
V23 CPT: Omission Errors at 54 Months 8.06 6.77 10.1 8.1 -.25*
84
Table 4 (continued)
Descriptive Statistics and Effect Sizes for Indicator Variables included in the current
study sample (N=470) and data excluded from the current study sample based on
exclusionary criteria (n=297 to 862) (*p<.05)
Current
Study sample
NICHD Study
sample
Mean
Std.
Dev.
Mean
Std.
Dev.
Effect
Size
F10 Executive Function
V24 WJ-R: Long Term Memory at 1
st
Grade 99.84 13.75 96.9 14.3 .23*
V25 WJ-R: Short Term Memory at 1
st
Grade 102.15 14.70 101 14.3 .09
V26 CPT: Commission Errors at 1
st
Grade 5.13 7.96 7.39 12.6 -.18*
V27 TOH: Planning Efficiency at 1
st
Grade 14.67 6.52 13.6 6.83 .16*
F11 Maternal Verbal Ability
V28 PPVT-R: Mother?s Verbal Ability 101.97 16.23 95.5 18.5 .35*
Examining descriptive statistics for the current study (Table 5) it is important to
note that several of the indicators for early attention are non-normal and highly positively
skewed. This suggests that high levels of inattention occurred relatively infrequently
within the sample at 36 months. Also it is necessary to note that scores from standardized
assessments (PPVT-R, WJ-R Memory for Names, WJ-R Memory for Sentences, WJ-R
Picture Vocabulary, PLS Auditory, PLS Expressive, Reynell Expressive, and Reynell
Vocabulary) that were administered across time (at 36 months, 54 months, and 1
st
grade),
across constructs (Executive Function, Early Language, Late Language, and Maternal
Verbal Ability), and across participants? (mother and child), indicate that the participants
mean scores represent average performance (i.e., mean = 100, s.d. = 15) as compared to
the samples by which the standardized assessments were normed.
85
Table 5
Descriptive Statistics for Indicator Variables (n=470)
Mean Min. Max. Std. Dev.
F1 Socio Economic Status
V1 SES Composite Score 21.27 10.11 49.22 4.83
Neighborhood Safety at 1
st
grade 8.25 2.33 10.00 1.42
Total Income-to-Needs Ratio at 36 Months 3.85 0.14 28.50 3.06
Mother's Education at 1 Month 14.67 7.00 21.00 2.36
F2 Early Maternal Sensitivity
V2 Positive Regard at 24 Months 2.87 1.00 4.00 0.69
V3 Supportive Presence at 36 Months 5.47 1.00 7.00 1.18
V4 Respect for Autonomy at 36 Months 5.44 1.00 7.00 1.00
F3 Early Maternal Stimulation
V5 Cognitive Stimulation at 24 Months 2.77 1.00 4.00 0.72
V6 Cognitive Stimulation at 36 Months 4.64 1.00 7.00 1.41
V7 Total Positive Scaffolding at 36 Months 119.64 0.00 223.00 38.78
F4 Early Language
V8 Reynell: Vocabulary at 36 Months 100.70 62.00 134.00 15.24
V9 Reynell: Expressive Language at 36 Months 99.07 62.00 138.00 13.90
F5 Early Attention
V10 Average Time Attending at 36 Months 166.66 0.00 900.00 150.84
V11 Total Time Off-Task at 36 Months 28.59 0.00 131.00 29.51
F6 Late Stimulation
V12 Cognitive Stimulation at 54 Months 4.50 1.00 7.00 1.23
V13 Quality of Assistance at 54 Months 4.79 1.00 7.00 1.34
V14 Goal Directed Partnership at 54 Months 4.85 1.00 7.00 1.31
F7 Late Sensitivity
V15 Supportive Presence at 54 Months 5.30 1.00 7.00 1.21
V16 Respect for Autonomy at 54 Months 5.33 1.00 7.00 1.08
F8 Late Language
V17 PLS: Auditory Comp. at 54 Months 101.48 53.00 139.00 19.15
V18 PLS: Expressive Language at 54 Months 103.64 50.00 127.00 18.44
V19 WJ-R: Picture Vocabulary at 54 Months 102.59 10.00 143.00 14.15
F9 Late Attention
V20 CBCL-Mother: Attention at 54 Months 2.55 0.00 12.00 2.19
V21 CBCL-Other: Attention at 54 Months 2.57 0.00 14.00 2.23
V22 Delay of Gratification at 54 Months 4.89 0.00 7.00 2.86
V23 CPT: Omission Errors at 54 Months 8.06 0.00 37.00 6.80
F10 Executive Function
V24 WJ-R: Long Term Memory at 1
st
Grade 99.84 2.00 154.00 14.68
V25 WJ-R: Short Term Memory At 1
st
Grade 102.15 43.000 151.00 13.87
V26 CPT: Commission Errors at 1
st
Grade 5.13 0.00 58.00 8.23
V27 TOH: Total Planning Efficiency at 1
st
Grade 14.67 0.00 31.00 6.56
F11 Maternal Verbal Ability
V28 PPVT-R: Mother?s Verbal Ability 101.97 44.00 159.00 17.69
86
Research Questions
Question 1. Does maternal stimulation and sensitivity at 54 months during a joint play
task influence performance on a later executive function task? If so, is that influence
direct or indirect through the effect on children?s early language and attention skills?
In order to answer this question, structural equation modeling was used. Modeling
proceeded in the following manner: A measurement model was first developed in which
all factors were allowed to correlate (i.e., no restrictions were placed on the relation
between constructs) and a priori error terms were allowed to covary. The a priori decision
to allow the residuals to covary was largely based on the effects of shared method or
shared respondents. After the measurement model, the a priori structural paths were
modeled and their fit indices were compared. Because several variables within the model,
particularly related to the construct of attention, were non-normal, the Sattora Bentler (S-
B) robust procedure was used (Satorra, & Bentler, 1988). The S-B scaling method is
recommended with continuous non-normal data and has been shown to outperform the
maximum likelihood method (Curran, West, & Finch, 1996; Yu & Muthen, 2002). The S-
B scaling method adjusts the x
2
, fit indices, and standard error by a factor based on the
amount of non-normality in the data. The S-B procedure adjusts the maximum likelihood
x
2
in order to better approximate the distribution and incorporates the kurtosis of the
variables into the adjustment. Using the S-B x
2
does not permit direct comparison of the
x
2
for nested models but instead requires the following adjustments:
87
__ __
1. k
0
=T
0
/ T
0 ,
where T
0
= MLx
2
and T
0
= S-B x
2
__ __
2. k
1
=T
1
/ T
1 ,
where T
1
= MLx
2
and T
1
= S-B x
2
3. x
2
= T
0
- T
1
4. S-B scaling coefficient = k = (d
0
k
0
? d
1
k
1
) /d, where d = d
0
- d
1
, where d =
degrees of freedom
_
5. S-B x
2
= D = D/k
Measurement Model. The confirmatory factor analysis (CFA) model is shown in
Figure 1. The CFA analysis model allowed all latent constructs to correlate as well as the
a priori residual error terms. One factor loading for each latent factor was fixed to one.
Factors were fixed to one based on hypothesized factor loadings. Non-hypothesized
factor loadings were constrained to zero. Initial problems with model convergence led to
the addition of several indicators from earlier time periods to the constructs of Early
Maternal Sensitivity (Sensitivity to Distress at 24 months), Early Maternal Stimulation
(Cognitive Stimulation at 24 months), and Late Attention (Delay of Gratification). As
expected, the addition of these indicators resulted in an adequately fitting model (x
2
(274)
= 579.772, p <.001 (S-B x
2
(274) = 570.5004, p < .001), CFI =.931 (.926), SRMR (.053),
RMSEA = .049 (.048), RMSEA 90% confidence interval: .043-.054 (.042-.054) (See
Table 6).
88
Table 6
Summary of Maximum Likelihood (and) Robust Model-Fit Statistics
Model x
2
df p-value CFI SRMR RMSEA
90%
Confidence
Interval
RMSEA
S-B

x
2
AIC
Measurement Model
579.772
(570.5004)
274 .000
.931
(.926)
.053
.049
(.048)
.043 - .054
(.042 - .054)
1. Structural Model: No
Relationship
1340.817
(1323.4924)
304 .000
.766
(.746)
.116
.085
(.085)
.081 - .090
(.080 - .090)
2. Structural Model:
Indirect Relationship
1325.092
(1311.0427)
300 .000
.769
(.748)
.117
.085
(.085)
.081 - .090
(.080 - .089)
Difference between
Model 2 and Model 1
13.521*
3. Structural Model:
Direct and Indirect
Relationship
1323.769
(1308.5512)
298 .000
.768
(.748)
.117
.086
(.085)
.081 - .090
(.080 - .090)
Difference between
Model 3 and Model 2
1.85
4. Mediator Model
1424.940
(1413.7154)
325 .000
.763
(.741)
.118
.085
(.085)
.080-.089
(.080-.089)
AIC Difference
between Mediator
Model and Model 2
49.849
(52.673)
*p<.05
Note. CFI = comparative fit index; SRMR = standardized root mean-square error of approximation; RMSEA = root mean-
square error of approximation; AIC = Akaike Information Criterion. Robust Fit Statistics are in ( )
89
As expected based on the different measurement scales entered into a single
model, the standardized and unstandardized parameter estimates for the measurement
model (See Table 7) varied greatly. The standardized estimates will be examined as
they allow you to evaluate the relative value of each predictor variable. Not all of the
statistics are significant, particularly measures of executive function and late
attention. Examining the relative contribution of indicators within a construct, the
most notable differences are within the construct of early attention (i.e., average time
attending = .031 and total time off-task =.979). While it may have been beneficial to
remove indicators that are redundant or contribute little to overall construct, the
current study was limited by an archival data set. Thus having achieved a
measurement model that adequately reflected the data, the next step was to impose
the a priori structural paths and compare the fit indices between three nested models.
Table 7
Unstandardized, Standardized, and Significance Levels for Measurement Model in
Figure 2 (Standard Errors in Parentheses; N = 470)
Parameter Estimate Error Unstandardized Standardized
Measurement Model Estimates
SES  SES composite (V1)
.000 1.00 1.00
Early Maternal Sensitivity  Positive
Regard for Child at 24 months (V2)
.948 5.59(.044)* .318
Early Maternal Sensitivity 
Supportive Presence at 36 months (V3)
.603 1.00 .798
Early Maternal Sensitivity  Respect
for Child?s Identity at 36 months (V4)
.625 9.536 (.085)* .781
Early Maternal Stimulation 
Cognitive Stimulation at 24 months
(V5)
.877 8.420 (.002)* .480
Early Maternal Stimulation 
Cognitive Stimulation at 36 months
(V6)
.641 10.990 (.004)* .768
90
Table 7 (continued)
Unstandardized, Standardized, and Significance Levels for Measurement Model in
Figure 2 (Standard Errors in Parentheses; N = 470)
Parameter Estimate Error Unstandardized Standardized
Early Maternal Stimulation  Total
Positive Scaffolding Score at 36
months (V7)
.757 1.00 .654
Early Language  Reynell Vocab. at
36 months (V8)
.398 1.00 .917
Early Language  Reynell Expressive
at 36 months (V9)
.824 11.725(.049)* .567
Early Attention  Average time
attending at 36 months (V10)
1.00 1.00 .031
Early Attention  Total time off-task
at 36 months (V11)
.203 2.865(.423) .979
Late Maternal Stimulation Cognitive
Stimulation at 54 months (V12)
.637 1.00 .771
Late Maternal Stimulation Quality of
Assistance at 54 months (V13) .409 20.039 (.066)* .913
Late Maternal Stimulation Goal
Directed Partnership at 54 months
(V14)
.605 17.223 (.067)* .796
Late Maternal Sensitivity Supportive
Presence at 54 months (V15)
.630 1.00 .777
Late Maternal Sensitivity Respect for
Autonomy at 54 months (V16)
.429 8.237 (.125)* .903
Late Language Preschool Language
Scale Auditory at 54 months (V17)
.616 1.00 .788
Late Language  Preschool Language
Scale Expressive at 54 months (V18)
.673 18.972(.046)* .739
Late Language WJ-R Picture Vocab.
at 54 months (V19)
.702 15.442(.042)* .713
Late Attention  CBCL Mother
Report at 54 months (V20)
.994 1.00 .107
Late Attention  CBCL Other
Caregiver Report at 54 months (V21)
.966 1.586(1.601) .260
Late Attention  Delay of
Gratification at 54 months (V22)
.917 1.451(.874) .398
91
Table 7 (continued)
Unstandardized, Standardized, and Significance Levels for Measurement Model in
Figure 2 (Standard Errors in Parentheses; N = 470)
Parameter Estimate Error Unstandardized Standardized
Late Attention  CPT Omission Errors
at 54 months (V23)
.888 1.396(9.738) .460
Executive Function  WJ-R Memory
for Sentences at 1
st
Grade (V24)
.707 3.702(2.086)* .707
Executive Function  WJ-R Memory
for Names at 1
st
Grade (V25)
.934 3.371(1.076)* .358
Executive Function  CPT
Commission Errors at 1
st
Grade (V26)
.980 -2.709(.433) -.199
Executive Function  Tower of Hanoi
Planning Efficiency at 1
st
Grade (V27)
.979 1.00 .204
*p<.05
92
Early
(24 and 36 Months)
1st grade
Attention
Language
Executive
Functioning
Late
(54 Months)
SES
Stimulation
Sensitivity
Figure 1. Measurement Relationship between Maternal Behaviors, Executive
Function, Attention, and Language
V2
V11
V25
V10
V24
V1
Attention
V23
V22
V20
V21
V9V8
Language
V17
V18
V19
V5
V12
V13
V14
V26
V27
Sensitivity
V3
V4
V15 V16
Stimulation
V7
V6
93
Model 1. The first structural model (see Figure 2) suggests that early maternal
stimulation and early maternal sensitivity will indirectly influence executive function
at first grade through a direct influence on early attention and early language skills
and an indirect influence on late attention and late language skills. Socioeconomic
status will have both direct and indirect (i.e., via attention, language, maternal
sensitivity, and maternal stimulation) effects on executive function in the first grade.
Maternal sensitivity and maternal stimulation at 54 months will not influence
executive function at first grade. The model fit was less than ideal (x
2
(304) =
1340.817, p< .000 (S-B x
2
(304) = 1323.4924, p<.000), CFI =.766 (.746), SRMR
(.116), RMSEA = .085 (.085), RMSEA 90% confidence interval: .081-.090 (.080-
.090) (See Table 6). This suggested that the structural paths did not reflect the data.
The squared multiple correlations for Model 1 indicate that the model
accounted for 14 percent of the variance in early maternal sensitivity, 14 percent of
the variance in early maternal stimulation, 40 percent of the variance in early
language, 63 percent of the variance in early attention, 29 percent of the variance in
late maternal sensitivity, 24 percent of the variance in late maternal stimulation, 90
percent of the variance late language, 50 percent of the variance in late attention, and
70 percent of the variance in executive function. Because the model did not fit the
data well it is difficult to appropriately interpret the percent of variance accounted for
by the model. The large amount of variance explained does suggest that the indicators
are indeed measuring the underlying construct despite the poor overall fit of the
model. The next step was to examine if the addition of indirect paths better reflected
the data.
94
Stimulation
Early
(24 and 36 Months)
1st grade
Attention
Attention
Language
Language
Executive
Functioning
Sensitivity
Late
(54 Months)
SES
Stimulation
Sensitivity
Figure 2. Structural Relationship between Maternal Behaviors, Executive Function,
Attention, and Language: No Relationship between Maternal Behaviors at 54 months
and Executive Function at 1st grade
D .925
D .928
D .776
D .604
D .872
D .319
D .844
D .707
.067
D .545
.701
.292
.048
.355
.197
.372
.380 .708
.489
.536
.243
.948
.810
.025
95
Model 2. Model 2 is identical to Model 1 (see Figure 3) except that late
maternal sensitivity and late maternal stimulation indirectly influence executive
function at first grade via late attention and late language. Once again the model fit
was poor (x
2
(300) = 1325.092, p<.000 (S-B x
2
(300) = 1311.0427, p< .000), CFI =.769
(.748), SRMR (.117), RMSEA = .085 (.085), RMSEA 90% confidence interval: .081-
.090 (.080-.089) (See Table 6). Importantly there were significant differences when
comparing model 2 and model 1 (
 S-B x
2
(4) = 13.521, p<.05) which suggests the
addition of the indirect path improves the model.
The squared multiple correlations for Model 2 were very similar to Model 1
and show that the model accounted for 14 percent of the variance in early maternal
sensitivity, 13 percent of the variance in early maternal stimulation, 38 percent of the
variance in early language, 48 percent of the variance in early attention, 28 percent of
the variance in late maternal sensitivity, 22 percent of the variance in late maternal
stimulation, 89 percent of the variance late language, 46 percent of the variance in
late attention, and 70 percent of the variance in executive function.
Because Model 2 reflects the most parsimonious model as compared to model
1 and model 3, the standardized total effects for the paths directly and indirectly
linking the model will be further examined (See Table 8). The effects indicate similar
strength of influence between SES and early maternal behaviors. SES strongly
directly and indirectly influenced early language ability but had a smaller effect on
early attention. Both late maternal stimulation and maternal sensitivity were similarly
influenced by SES. Both early language ability and SES had large effects on late
langue ability. Late attention was strongly indirectly influenced by early maternal
96
stimulation and directly by early attention. Finally, executive function was
moderately influenced directly and indirectly by SES and, surprisingly, strongly
influenced directly and indirectly by early and late language ability.
Table 8
Decomposition of Standardized Effects for Structural Paths of Indirect Model
Paths
Direct
Effect
Indirect
Effect
Total
Effect
DV: Early Maternal
Sensitivity
SES .379 -- .379
DV: Early Maternal
Stimulation
SES .366 -- .366
DV: Early Language
SES .362 .172 .534
Early Maternal
Sensitivity
.190 -- .190
Early Maternal
Stimulation
.273 -- .273
DV: Early Attention
SES .008 .304 .312
Early Maternal
Sensitivity
.180 -- .180
Early Maternal
Stimulation
.644 -- .644
DV: Late Maternal
Sensitivity
SES -- .200 .200
Early Maternal
Sensitivity
.530 -- .530
DV: Late Maternal
Stimulation
SES -- .170 .170
Early Maternal
Stimulation
.464 -- .464
97
Table 8 (Continued)
Decomposition of Standardized Effects for Structural Paths of Indirect Model
Paths
Direct
Effect
Indirect
Effect
Total
Effect
DV: Late Language
SES -- .553 .553
Early Maternal
Sensitivity
-- .213 .213
Early Maternal
Stimulation
-- .263 .263
Early Language .930 -- .930
Late Maternal
Sensitivity
.160 -- .160
Late Maternal
Stimulation
.258 -- .258
DV: Late Attention
SES -- .235 .235
Early Maternal
Sensitivity
-- .176 .176
Early Maternal
Stimulation
-- .447 .447
Early Attention .508 -- .508
Late Maternal
Sensitivity
.067 -- .067
Late Maternal
Stimulation
.019 -- .019
DV: Executive
Function
SES .066 .416 .482
Early Maternal
Sensitivity
-- .173 .173
Early Maternal
Stimulation
-- .216 .216
Early Language -- .751 .751
Early Attention -- .005 .005
Late Maternal
Sensitivity
-- .055 .055
Late Maternal
Stimulation
-- .0176 .0176
Late Language .807 -- .807
Late Attention .009 -- .009
98
Stimulation
Early
(24 and 36 Months)
1st grade
AttentionAttention
Language
Language
Executive
Functioning
Sensitivity
Late
(54 Months)
SES
Stimulation
Sensitivity
Figure 3. Structural Relationship between Maternal Behaviors, Executive Function,
Attention, and Language: Indirect Relationship between Maternal Behaviors at 54
months and Executive Function at 1st grade
D .925
D .931
D .787
D .725
D .886
D .338
D .848
D .737
.066
D .542
.644
.019
.273
.008
.362
.190
.366
.258
.379 .508
.464
.530
.180
.067
.160
.930
.807
.009
99
Model 3. The third model is identical to the model 1 and 2 (See figure 4)
except that late maternal sensitivity and late maternal stimulation both directly and
indirectly influence executive function at first grade. The model fit remained poor
(x
2
(298) = 1323.769, p<.000 (S-B x
2
(298) = 1308.5512, p<.000), CFI =.768 (.748),
SRMR (.117), RMSEA = .086 (.085), RMSEA 90% confidence interval: .081-.090
(.080-.090) (See Table 6). There was not a significant difference between model 3
and model 2 (
 S-B x
2
(2) = 1.858, p >.05) which suggests the addition of the direct
paths between late maternal sensitivity and late maternal stimulation to executive
function is redundant. The squared multiple correlations for Model 3 were nearly
identical to Model 1 and 2 thus will not be presented.
100
Stimulation
Early
(24 and 36 Months)
1st grade
AttentionAttention
Language
Language
Executive
Functioning
Sensitivity
Late
(54 Months)
SES
Stimulation
Sensitivity
Figure 4. Structural Relationship between Maternal Behaviors, Executive Function,
Attention, and Language: Indirect and Direct Relationship between Maternal
Behaviors at 54 months and Executive Function at 1st grade
D .925
D .931
D .787
D .729
D .886
D .337
D .848
D .729
.058
D .535
.805
.028
.092.642
.012
.273
.008
.362
.190
.366
.257
.379 .518
.464
.531
.174
.053
.161
.932
.052
101
Supplementary Analyses for Question 1
In order to further examine the findings of structural equation models which
suggested an overall poor fit, the constructs were further examined. The purpose for
these additional analyses was to determine possible explanations for the poor fit of
the model, specifically examining the data for support of the theorized constructs. In
order to examine the constructs, the intercorrelations among the constructs and factor
analysis of the theorized constructs were examined.
Maternal Sensitivity. First, the intercorrelations among measures of maternal
sensitivity were examined in Table 9. The correlational analyses of maternal
sensitivity measures indicated weak to moderate relationships among measures of
maternal sensitivity with the strongest relationships occurring within the same time
period (i.e., measures at 36 months most highly correlate with other measures at 36
months and so forth). Because similar tasks and scoring were used across tasks, it is
expected that the correlations would be increased based on methodological
similarities. Factor analysis for the indicators of the early sensitivity construct
indicated that 60% of the variance was accounted by a single factor. Two factors
accounted for 88% of the variance with the 36 month tasks loading on factor one and
the 24 month task loading on factor two. Factor analysis for the indicators of the late
sensitivity indicated that 84% of the variance was accounted for by a single factor. As
expected this provides support for convergent validity of the construct of maternal
sensitivity.
102
Table 9
Intercorrelations Among measures of Maternal Sensitivity
Positive
regard at 24
Months
Supportive
Presence at
36 Months
Respect for
Autonomy at
36 Months
Supportive
Presence at
54 Months
Supportive
Presence at
36 Months
0.256***
Respect for
Autonomy at
36 Months
0.206*** 0.637***
Supportive
Presence at
54 Months
0.285*** 0.369*** 0.304***
Respect for
Autonomy at
54 Months
0.277*** 0.352*** 0.378*** 0.699***
*p<.05, **p<.01, ***p<.001
Maternal Stimulation. Correlations among measures of maternal stimulation
indicated weak to moderate relationships, of similar strength to the analogous
correlations among measures of maternal sensitivity (See Table 10). Once again the
strongest relationships occurred within the same time period (i.e., measures at 36
months most highly correlate with other measures at 36 months and so forth). Again
the method of administration and scoring was similar across tasks. Factor analysis for
the indicators of the early stimulation construct indicated that 60% of the variance
was accounted by a single factor. Two factors accounted for 85% of the variance
with the 36 month tasks loading on factor one and the 24 month task loading on factor
two. Factor analysis for indicators of the late stimulation construct indicated that 78%
of the variance was accounted for by a single factor. Once again this provided
convergent validity for maternal stimulation as a construct.
103
Table 10
Intercorrelations Among measures of Maternal Stimulation
Cognitive
Stimulation
at 24 Months
Cognitive
Stimulation
at 36 months
Total
Positive
Scaffolding
at 36 Months
Cognitive
Stimulation
at 54 Months
Quality of
Assistance
at 54
Months
Cognitive
Stimulation at
36 Months
0.338***
Total Positive
Scaffolding at
36 Months
0.303*** 0.552***
Cognitive
Stimulation at
54 Months
0.259*** 0.347*** 0.321***
Quality of
Assistance at
54 Months.
0.229*** 0.285*** 0.238*** 0.701***
Goal Directed
Partnership at
54 Months
0.201*** 0.238*** 0.208*** 0.595*** 0.733***
*p<.05, **p<.01, ***p<.001
Maternal Sensitivity and Maternal Stimulation. These findings are remarkably
similar to the findings for maternal stimulation. Thus factor analysis was run to
examine if maternal sensitivity and maternal stimulation were indeed two separate
constructs. Although sensitivity and stimulation were originally hypothesized as
unidimensional, there is some literature to suggest otherwise. Factor analyses of the
variables that comprised these constructs (maternal sensitivity and maternal
stimulation) indicates that there are three distinct constructs but the factors are based
on the longitudinal nature of the data (e.g., all indicators at 24 months loaded on
single factor, all indicators at 36 months loaded on a single factor, and all indicators
at 54 months loaded on a single factor). Further factor analysis across indicators of
104
maternal sensitivity and maternal stimulation (all measures of maternal sensitivity and
maternal stimulation at 54 months) indicated in a single factor was underlying the
constructs of maternal sensitivity and maternal stimulation and explained 70% of the
variance. This lack of divergent validity will be further examined in the discussion
section.
Language Measures. The analyses generally indicated moderate relationships
across language measures (See Table 11). While the intercorrelation of .76 is reported
between the Reynell Expressive and Reynell Vocabulary by the test publishers, they
intercorrelate at r = .536 in the current study. While this is notable, the descriptive
statistics indicate means and standard deviations in line with the greater population.
Factor analysis for the early language construct indicated that 76% of variance was
accounted by a single factor. Factor analysis for the late language construct indicated
that 70% of the variance was accounted by one factor and two factors accounted for
87% of the variance.
105
Table 11
Intercorrelations Among Language Measures
Reynell:
Vocabulary
at 36 Months
Reynell:
Expressive
Language at
36 Months
PLS:
Auditory
Comp. at 54
Months
PLS:
Expressive
Language
at 54 Months
Reynell:
Expressive
Language at
36 Months
0.536***
PLS:
Auditory
Comp. at 54
Months
0.698*** 0.426***
PLS:
Expressive
Language
at 54 Months
0.640*** 0.428*** 0.618***
WJ-R: Picture
Vocabulary at
54 Months
0.542*** 0.367*** 0.528*** 0.526***
*p<.05, **p<.01, ***p<.001
Attention Measures. The analyses generally indicated weak or no relationship
across attention measures with the CBCL mother and CBCL alternative caregiver
relationship being the notable exception (See Table 12). Based on the principles of
structural equation modeling, it is generally desirable to have moderate correlations.
Weak correlations may suggest that several different constructs are being measured.
As expected the child behavior questionnaire negatively correlated with the CBCL
and the CPT, as higher scores indicated an increased tendency to maintain attentional
focus upon task-related channels. Factor analysis of the measures of the early
attention indicated that 49% of variance was accounted by a single factor. Factor
analysis for the indicators of the late attention construct indicated that 67% of
variance was accounted by two factors. While the indicators for attention were
106
specifically designed to be heterotypic in nature and do represent measurement using
multiple methods and multiple sources, the low correlations between the measures
indicate that a weak relationship which makes the measurement of the construct of
attention suspect in this case.
Table 12
Intercorrelations Among measures of Attention
Average
Time
Attending
at 36 Months
Total Time
Off-Task
at 36
Months
CBCL-
Mother:
Attention
at 54 Months
CBCL-Other:
Attention at
54 Months
Delay of
Gratification
at 54 Months
Total Time
Off-Task
at 36 Months
0.003
CBCL-
Mother:
Attention
at 54 Months
0.025 -0.062
CBCL-Other:
Attention at 54
Months
0.009 -0.067 0.463***
Delay of
Gratification
at 54 Months
-0.030 -0.016 0.047 0.109*
CPT:
Omission
Errors at 54
Months
-0.005 -0.079 0.042 .112* .212***
*p<.05, **p<.01, ***p<.001
Executive Function. The analyses generally indicated weak or no relationship
across executive function measures with short term (WJ-R Memory for Sentences)
and long term memory (WJ-R Memory for Names) having a moderate relationship
(See Table 13). As expected the relationships between the CPT: Commission Errors
107
and the other measures of executive function were negative. Higher scores on the
CPT: Commission task indicated decreased ability to inhibit a prepotent response and
decreased sustained attention where as higher scores on the other measures of
executive function indicated greater planning and memory abilities. Factor analysis
for measures of executive function indicated that 35% of the variance was accounted
for by a single factor and that 61% of the variance was accounted for by two factors,
and 82% of the variance was accounted for by three factors. While the third factor in
the three factor solution has an Eigenvalue of .845 it is important to note the three
factors are similar to a proposed three factor model of executive function that
includes aspects of planning, memory and inhibition. It should be noted that
traditional models of executive function includes measures of working memory,
while the current study defines memory through short-term memory and long-term
memory. These results will be further examined in the discussion section.
Table 13
Intercorrelations Among Executive Function Measures at 1
st
grade
CPT: Comission
Errors at 1
st
Grade
WJ-R: Memory
for Names at 1
st
Grade
WJ-R: Memory
for Sentences at
1
st
Grade
WJ-R: Memory
for Names at 1
st
Grade
-0.065
WJ-R: Memory
for Sentences at
1
st
Grade
-0.074 0.304***
TOH: Total
Planning
Efficiency Score
at 1
st
Grade
-0.153*** 0.116* 0.099*
*p<.05, **p<.01, ***p<.001
108
Summary
In summary, there were significant differences between the three models with
the model that included an indirect path (Model 2) between maternal sensitivity and
maternal stimulation and executive function providing the most parsimonious
representation of the data. The significant differences aside, the overall fit of the
model was poor which suggests that the current model does not adequately reflect the
data. Further, as the intercorrelations among the indicators suggest, there is little
convergent or divergent validity among the indicators of several constructs (i.e.,
attention, maternal sensitivity, and maternal stimulation) which in turn makes
accurate interpretation of the model difficult.
Question 2: Does ethnicity, African American and Caucasian group membership,
moderate the strength of the relationship between maternal stimulation and maternal
sensitivity and executive function?
When the current study was designed, it was anticipated that there would be
large enough subsamples to allow for analysis by ethnicity. This was not the case.
Due to the sample size of each group (African America n= 26, Caucasian n = 413) the
nested multiple group models did not converge on a solution. This is not surprising
based on the low power of the African American group. Therefore the current study
was unable to examine whether ethnic group membership moderates the relationship
between maternal behaviors and executive functioning. The literature does suggest
that, even if the model had converged, differences would be difficult to detect
because of the large differences in the sample sizes (Hancock, 2006).
109
Supplementary analysis for question 2
Although it is not possible to examine multi-group differences using structural
equation modeling, direct comparison of mean differences suggests significant
differences between both groups with the Caucasian participants demonstrating
higher performance across the majority of the measures (See Table 14). Examining
the effect sizes, 10 were strong (above .8) and 9 were moderate (.5 to .8) suggesting
important differences between the two populations.
110
Table 14
Descriptive Statistics for Caucasian (N=427) and African American (N=26) (*p<.05)
Caucasian African American
Mean
Std.
Dev.
Mean
Std.
Dev.
Effect
Size
F1 Socio Economic Status
V1 SES Composite Score 27.45 5.15 22.53 2.01 2.45*
Neighborhood Safety at 1
st
grade 8.38 1.30 7.15 1.59 0.78*
Total Income-to-Needs Ratio at 36 Months 4.14 3.22 2.00 1.37 1.56*
Mother's Education at 1 Month 14.93 2.32 13.38 1.10 0.66*
F2 Early Maternal Sensitivity
V2 Positive Regard at 24 Months 2.97 0.66 2.50 0.76 0.61*
V3 Supportive Presence at 36 Months 5.66 1.04 4.38 1.53 0.83*
V4 Respect for Autonomy at 36 Months 5.61 0.88 4.73 1.22 0.72*
F3 Early Maternal Stimulation
V5 Cognitive Stimulation at 24 Months 2.87 0.71 2.23 0.76 0.83*
V6 Cognitive Stimulation at 36 Months 4.80 1.35 3.69 1.72 0.65*
V7 Total Positive Scaffolding at 36 Months 124.02 36.90 100.59 38.83 0.60*
F4 Early Language
V8 Reynell: Vocabulary at 36 Months 103.47 14.08 86.19 14.02 1.23*
V9 Reynell: Expressive Language at 36 Months 100.54 13.52 94.92 12.73 0.44*
F5 Early Attention
V10 Average Time Attending at 36 Months 166.73 159.74 167.21 103.66 0.00
V11 Total Time Off-Task at 36 Months 28.99 29.23 23.00 32.47 0.18*
F6 Late Stimulation
V12 Cognitive Stimulation at 54 Months 4.67 1.12 4.04 1.25 0.51*
V13 Quality of Assistance at 54 Months 4.96 1.25 4.00 1.47 0.65*
V14 Goal Directed Partnership at 54 Months 4.94 1.27 4.31 1.54 0.41*
F7 Late Sensitivity
V15 Supportive Presence at 54 Months 5.47 1.10 4.54 1.70 0.55*
V16 Respect for Autonomy at 54 Months 5.51 0.99 4.69 1.26 0.65*
F8 Late Language
V17 PLS: Auditory Comprehension at 54 Months 104.91 18.32 86.00 15.34 1.23*
V18 PLS: Expressive Language at 54 Months 106.60 16.61 88.85 21.14 0.84*
V19 WJ-R Picture Vocabulary at 54 Months 104.82 13.25 88.81 10.32 1.55*
111
Table 14, continued
Descriptive Statistics for Caucasian (N=427) and African American (N=26) (*p<.05)
Summary
In conclusion, the original proposed multi-group analysis (Caucasian vs.
African American) was not possible due to the small sample size. Examination of the
effect sizes for the study indicators does indicate a number of large and moderate
differences which would be expected to have important implications. While it is not
possible to understand this influence in relation to the proposed model, in the next
chapter these effect sizes will be discussed in relation to current theory and literature
relating to ethnicity, SES, and the development of executive functions.
Caucasian African American
Mean
Std.
Dev.
Mean
Std.
Dev.
Effect
Size
F9 Late Attention
V20 CBCL-Mother: Attention at 54 Months 2.40 2.10 2.42
2.72
-0.01
V21 CBCL-Other: Attention at 54 Months 2.51 2.22 2.77 2.61 -0.10
V22 Delay of Gratification at 54 Months 5.05 2.80 2.65 2.83 0.85*
V23 CPT: Omission Errors at 54 Months 8.03 6.77 9.33 5.08 -0.25
F10 Executive Function
V24 WJ-R: Long Term Memory at 1
st
Grade 102.63 13.75 98.19 11.92 0.37*
V25 WJ-R: Short Term Memory at 1
st
Grade 101.14 14.70 91.88 10.66 0.87*
V26 CPT: Commission Errors at 1
st
Grade 4.78 7.96 4.27 6.46 0.08
V27 TOH: Total Planning Efficiency at 1
st
Grade 15.21 6.52 11.46 6.87 0.55*
F11 Maternal Verbal Ability
V28 PPVT-R: Mother?s Verbal Ability 105.27 16.23 82.96 16.48 1.35*
112
Question 3: Does maternal verbal ability mediate the relationship between SES and
executive functioning?
In order to answer this question, structural equation modeling was used. Using
model 2, which includes an indirect relationship between maternal behaviors at 54
months and executive function at 1
st
grade as a base model, maternal verbal ability
was added as a mediator between SES and executive functioning while retaining a
direct path between SES and executive functioning (See Figure 6). The overall model
fit remained poor (x
2
(325) = 1424.940, p< .000 (S-B x
2
(325) = 1413.7154, p< .000),
CFI =.763 (.741), SRMR (.118), RMSEA = .085 (.085), RMSEA 90% confidence
interval: .080-.089 (.080-.089) (See Table 6). Comparisons were next made between
the AIC fit index of the current model and model 2. Model 2?s AIC fit index was
smaller than the mediator model. This suggests that the fit of Model 2 fit is more
parsimonious and more desirable and that the addition of maternal verbal ability in
the mediator model is not making a useful contribution.
113
.763
.014
.028
.774
.144
.260
.157
.512
.019
.063
.931
.007
.530
.464
.643
.177
.379
.362
.274
.189
.366
Stimulation
Early
(24 and 36 Months)
1st grade
Attention
Attention
Language
Language
Executive
Functioning
Sensitivity
Figure 5. Structural Relationship between SES, Maternal Verbal Ability, Maternal
Behaviors, Executive Function, Attention, and Language: Direct and Indirect
Relationships via Maternal Verbal Ability between SES and Executive Function.
Late
(54 Months)
SES
Stimulation
Sensitivity
D .931
D .886
D .787
D .335
D .848
D .952
D .727 D .734
Verbal
Ability
114
Supplementary analysis for question 3
Because the theoretical relationship between maternal verbal ability, SES, and
executive functioning is unclear, two other models were examined. In the first
alternative model (See Figure 6) in which maternal verbal ability replaced SES, the
model did not converge on a solution. In a second model SES directly influenced
maternal verbal ability which in turn directly and indirectly influenced maternal
behaviors, attention, language, and executive functioning. This model (See Figure 7)
failed to result in a solution as there was not enough information in the model.
Summary
In conclusion, the addition of maternal verbal ability did not produce
significantly different results. This suggests that within the current model maternal
verbal ability does not mediate the relationship between SES and executive function.
The findings of the supplemental analysis continue to support that the model does not
accurately reflect the data and is generally unstable.
115
Verbal
Ability
Stimulation
Early
(24 and 36 Months)
1st grade
AttentionAttention
Language
Language
Executive
Functioning
Sensitivity
Figure 6. Proposed Structural Relationship between SES, Maternal Verbal Ability,
Maternal Behaviors, Executive Function, Attention, and Language: Maternal Verbal
Ability in place of SES
Late
(54 Months)
Stimulation
Sensitivity
116
Verbal
Ability
Stimulation
Early
(24 and 36 Months)
1st grade
AttentionAttention
Language
Language
Executive
Functioning
Sensitivity
Figure 7. Proposed Structural Relationship between SES, Maternal Verbal Ability,
Maternal Behaviors, Executive Function, Attention, and Language: Indirect Relationship
via Maternal Verbal ability between SES and Executive Function
Late
(54 Months)
Stimulation
Sensitivity
SES
117
CHAPTER V
Discussion
The present study examined the relationship between maternal behaviors,
specifically maternal sensitivity and maternal stimulation, and later development of
executive functioning. In order to do this, data from the NICHD SECCYD were
analyzed. The relationships between children?s performance on measures of executive
function at 1
st
grade and early and late maternal behaviors as well as early and late
child attention and child language abilities were examined. The study had several
strengths. First, the data set was relatively large which allowed for the detection of
fairly small effects. Additionally, constructs typically included several, well-known
measures. Another positive aspect of the current study is the strength of the structural
equation methodology which allowed for the examination of both direct and indirect
paths as well as the examination of mediator variables. Finally, through
supplementary analyses additional important insights have been gained about the
constructs of maternal sensitivity, maternal stimulation, and executive functioning.
Overall, there were significant differences between the three proposed models but the
poor fit of the models suggest that they do not accurately represent the data. Below,
each of the research questions is discussed including the hypotheses and the
supplementary analyses. Finally, limitations and recommendations for future research
are presented.
118
Question 1. Does maternal stimulation and maternal sensitivity at 54 months
influence performance on later executive functioning tasks? If so, is that influence
direct or indirect through the effect on children?s early language and attention skills?
The main goal of the current study was to examine the influence of maternal
stimulation and maternal sensitivity on later executive functioning. It was expected
that later maternal sensitivity and maternal stimulation would have less influence on
development of executive functioning as compared to the influence of early maternal
stimulation and sensitivity. This hypothesis was based on the work of Landry and
others (2002) which found that early maternal behaviors had a greater impact on the
development of executive functioning. Conversely, however, an NICHD Early Child
Care Research Network study (2005a) that used the NICHD SECCYD dataset found
that a child?s early environment as well as later environment, including maternal
behaviors, influenced the later development of attention and memory but not
planning.
The findings from the current study support both of these studies in that there
were significant differences between the models. Specifically, Model 2, which
included an indirect path between late maternal behaviors and executive functioning
via late attention and late language abilities, represented the best fitting model. This
finding supports Landry?s (2002) research which suggests that there is not a direct
effect between later maternal behaviors and the development of executive functioning
but instead a indirect path between early maternal behaviors (verbal scaffolding at 3
years) and later executive processes (search retrieval task at 6 years) via language
skills at 4 years. Notably, Landry?s (2002) research was limited in that it only
119
examined direct paths between later maternal behaviors and the development of
executive function. Thus findings from the current study build on Landry?s work and
suggest that later maternal behaviors do indirectly influence the development of
executive functioning by directly influencing language and attention. The findings
also support the results of the NICHD Early Child Care Research Network study
(2005a) which found, using hierarchal regression, that statistically both early and late
maternal behaviors play important roles in the development of executive function.
While hierarchal regression did not allow the researchers to examine if that
relationship was direct or indirect, it does provide validity for the current study.
Overall these findings suggest that later maternal behaviors do uniquely contribute
over and above early maternal behaviors. The current study does not discount the
importance of early maternal behaviors but suggests that later maternal behaviors are
also important by supporting the basic cognitive and language processes necessary for
children to develop executive functions.
While the proposed model did not fit the data well, these results do suggest
that later maternal sensitivity and maternal stimulation do not directly influence
executive functioning in the 1
st
grade but instead influence the development through
assisting the child in development of attention and language skills. Importantly,
language skills had a particularly strong influence on the development of executive
functioning. A number of other executive function researchers have supported the
notion that language ability is related to the development of executive function (see
Barkely, 1997; Saxton, 1997; Stuss & Benson, 1990). For example, within Barkley?s
model of executive function, language plays a prominent role in the development of
120
verbal self-regulation and verbal heuristics that are used as reference points during
problem-solving. Benson (1997) writes, ?Language is a tool that supports higher-
order mental processes that are key for internalizing what children initially learn
through their action and for imposing order and structure on action? (p. 53). Benson
hypothesizes that through verbalizing their temporal sequence, children are able
discuss the future which then influences their ability to plan what is a recognized
component of executive function.
Other notable researchers, namely Vygotsky (1934/1962) and Luria (1961),
regarded internal speech as having a necessary role in self-regulation. Specifically,
Vygotsky presented the notion that internal speech impacts an individual?s ability to
attend to their environment and master multi-step actions. Vygotsky theorized that
internal speech developed in three stages. The first stage of internal speech begins
with externalized utterances that resemble social speech. The second stage involves
verbalizations with self-regulatory function. In the final stage, verbalizations become
internal, ending with silent thoughts. Bivens and Berk (1990) reported evidence in
support of Vygotsky?s theory of the relationship between internal speech and self
regulation. These researchers found support for a developmental progression from
external self-stimulating and task-irrelevant verbalizations to task relevant external
manifestations of inner speech (i.e., inaudible muttering and lip and tongue
movement). While the current study suggests that there may be a relationship
between language (specifically self-directed speech) and executive function, there is
limited emerging empirical support for this claim. In a related study, Kopecky,
Chang, Klorman, Thatcher, and Borgstedt (2005) examined executive function
121
performance on the TOH and self-regulatory verbalizations with children with and
without ADHD. They reported that children in all groups exhibited higher rates of
self regulatory verbalizations compared to non-self-regulatory verbalizations during
the TOH task.
Another important avenue to consider related to both language development
and executive function is the role of theory of mind. Notably, the development of
executive functioning is thought to influence the developmental unfolding of the
theory of mind. Several studies have shown that performance on false-belief and
other theory of mind tasks can be predicted from response inhibition, cognitive
conflict resolution, and working memory (all executive functioning skills) even after
controlling for verbal ability (Frye, Zelazo, & Palfai, 1995; Hughes, 1998; Perner &
Lang, 2000). Researchers theorize that children who fail theory of mind tasks may do
so ?not because they lack an understanding of false belief, but because of the
peripheral executive demands that these tasks pose? (Sabbagh, Xu, Carlson, Moses,
& Lee, 2006, p.75). Therefore, as a child?s executive abilities increase so should their
ability to perform theory of mind tasks.
While the role of language in the development of executive function and
theory of mind is not clear, there is data to suggest that there is a significant positive
relationship between impairments in language and executive function, as evidenced
by commonly known neurodevelopmental disorders such as autism. Russell (1997)
theorized that individuals with autism are less able to use verbal self-regulation,
which makes them more vulnerable to errors that require an individual to inhibit a
prepotent response, such as the TOH task. The most robust evidence of executive
122
dysfunction in autism has come from the Tower of Hanoi (Ozonoff, Pennington, &
Rogers, 1991). In a study comparing performance of autistic children with non-
autistic children on measures of executive function, Joseph and colleagues (2005)
found that children with autism performed significantly worse than the comparison
group on the Tower of Hanoi task. The authors reported that children with autism did
not use their language skills during the TOH task to verbally encode and rehearse the
sequence of correct moves. The finding that individuals with autism have
consistently worse performance on the TOH task has been replicated in a number of
laboratories and with participants with a range of IQ?s.
Recent research has supported the hypotheses that language ability accounts
for a significant amount of variance in performance on executive function tasks and
that language ability predicts performance on measures of executive function. This
may be due to language?s shared variance with intelligence. Kalback (2004)
examined the developmental language differences and executive functioning in deaf
children. He reported that ? not only was language ability predictive of overall
executive function performance, but when examining individual components,
language ability also predicted performance in a working memory task and ratings of
inhibitory behavior? (p. 147). This suggests that language development is uniquely
tied to executive function development.
As expected, the executive function tasks chosen for the current study were
not clearly linear (i.e., they typically taped more than one skill or ability). The most
prominent example being the TOH task, which is recognized as requiring higher-
order cognitive abilities such as advanced planning but also includes several single
123
component skills (i.e., memory) (Scholnick & Friedman, 1993). To this end,
conceptualization of executive functioning is further complicated by the fact that it
requires the application of basic processes. Therefore it is recognized that poor
performance on the TOH may be the result of deficiency in a single competent skill
or a deficiency in the coordination of those skills for higher-order tasks.
Based on previous empirical and theoretical work, the executive functioning
tasks were expected to be interdependent and to some extent predictive of
performance of one another (NICHD Early Child Care Network, 2005a). The
extremely low intercorrelations (-.065 - .304) among the measures of executive
function in the current study suggest that they are largely independent in the current
study. This finding is in contrast to recent work (Friedman, Miyake, Corley, Young,
DeFries, & Hewitt, 2006) that reports intercorrelations of .10 to .58 among measures
of executive function (inhibiting, updating, and shifting) with the majority of the
intercorrelations being between .21 to .25. They further reported that the interfactor
correlations among inhibiting, updating, and shifting ranged from .64 to .39. This lack
of relationship between measures of executive function, particularly the TOH,
suggests that the tasks were not functioning as originally expected and as they have in
previous studies.
This lack of replication is difficult to explain as the measures used have been
empirically validated and traditionally administered. A notable limitation of the
current study was the lack of information about the stability of the TOH measure in
the NICHD study. Within the NICHD SECCYD data set the TOH was administered
at 1
st
, 3
rd
, and 5
th
grades, unfortunately only the data set through 1
st
grade is currently
124
available to qualified researchers. Therefore at the current time is not possible to get
an accurate estimate about the predictive validity of the TOH as well as the
correlation between the administrations of the TOH across time. Some researchers
(e.g., Borys et al., 1982) have reported a number of practice effects with the TOH on
both normative and mentally retarded populations while others (Bishop, Aamodt-
Leeper, Creswell, McGurk, & Skuse, 2001) have found little or no practice effects
with a normative childhood population. In the future, this will be important to
examine as the TOH data in the current study had lower correlations in the current
study than expected.
It is recognized that a myriad of factors can differentially influence
performance on measures of executive function. For example, planning is influenced
by a number of factors, namely motivation, emotion, and social context (Scholnick &
Friedman, 1993). Of interest to the current study, Ellis and Seigler (1997) noted
reasons why individuals fail to plan. They noted that children may fail to plan with
the expectation that someone (i.e., a parent) else will take care of it for them. This
type of inappropriate scaffolding or intrusiveness over time may result in a decrease
the ability to solve tasks, such as the TOH, that requires coordination of multiple
components but would less likely influence the development of short-term memory.
Unfortunately the current study was unable to tease out those effects due, in part, to
the archival nature of the data.
Another important factor when considering executive function was the
measures of inhibition and executive function. Errors of omission occur when the
participant fails to respond to the target stimulus while errors of commission occur
125
when the target responds to a non target stimulus (Conners, Epstein, Angold, &
Klaric, 2003). Based on these definitions, researchers and practitioners have used
measures of errors of commission to represent inhibition and measures of errors of
omission to represent attention, or more specifically inattention. Unfortunately,
attempts to empirically support these distinctions have produced inconclusive results
(Barkley, 1991). Within the current study both measures of commission and measures
of omission were used to measure different constructs at different points in time.
Specifically, at 54 months errors of omission loaded on the attention factor and at 1
st
grade errors of commission loaded on the executive function factor.
Conners, Epstein, Angold, and Klaric (2003) analyzed data from a normative
sample of the CPT test, a measure similar to that used in the current study, and
discovered several notable findings. First, they reported that males tended to have
quicker response times and make more commission errors than their female
counterparts. In line with the current study, the authors did not find ethnic differences
within the sample. Second, the measures associated with the CPT were strongly
affected by age. However, for the different measures associated with the CPT,
namely attention and inhibition, the developmental trajectory varied. Conners and
associates (2003) note limitations related to measurement of commission errors,
namely difficulty truly understanding what is being measured. They report that some
commission errors are the result of ?sluggishness in response? while others are the
result of true impulsive behavior. In relation to the current study, the distinction
between each type of error representing a behavioral topography may have been
126
artificial and may not have accurately represented the underlying data thus
contributing to a poor fit within the model.
Another important implication of the current study is the relationship between
maternal sensitivity and maternal stimulation. Exploratory factor analysis suggests
that these variables in this data set form a single construct. This may be due to the
shared methodology or due to lack of differentiation in the operationalization of the
constructs. While available literature does appear to recognize theoretical differences
between maternal sensitivity and maternal stimulation, upon further examination of
the grain size of the construct used in the current study it would seem that the
differences may not have been clearly operationalized. This lack of differentiation
between scores for maternal sensitivity and maternal stimulation is especially
prominent in the lower scores (i.e., scores of 1, 2, or 3). There appeared to be
measurable and conceptual differences in the operationalization of higher scores.
Unfortunately, there was much less differentiation across the lower scores such that if
a mother was not attentive or if they were inappropriately over attentive they would
receive low scores across both measures of sensitivity and stimulation. This appears
to be the result of both theoretical and measurement limitations of the construct such
that positive parenting behaviors are more uniquely measured as compared to
negative parenting behaviors which are typically characterized by either ignoring or
intrusiveness.
Conclusion
While the overall fit of the model suggested it did not accurately reflect the
data, several tentative findings did emerge. First, the results of the current study are in
127
line with other recent research and suggest that both early and later maternal
behaviors play important roles in the development of executive function. Secondly,
language ability played a more prominent role in the development of executive
function than previously thought and should be further investigated. Third, in the
current study the measures of executive function were less interdependent than
previously theorized making it difficult to truly conceptualize executive function.
Finally, the results from the current study should act as a cautionary tale for other
studies using the constructs of maternal sensitivity and maternal stimulation. The
results from the current study would suggest that maternal sensitivity and maternal
stimulation actually represent a single construct, possibly due to the lack of
specificity in the operational definition of the construct.
Question 2. Does ethnicity, African American and Caucasian group membership,
moderate the strength of the relationship between maternal stimulation and maternal
sensitivity and executive function?
The relationship between maternal behaviors on the development of executive
function across two ethnicities (Caucasian and African American) is difficult to
ascertain. As noted in the results section, multi-group comparison using structural
equation modeling was not possible due to the small sample size of African American
participants. Using other methods, the current study found surprisingly large effect
size differences between Caucasian and African American participants? scores across
several variables. Notably, the biggest effect size difference was the SES overall
composite score. This suggests that diverging childhood outcomes with families of
different ethnicities may be related to membership in a lower socioeconomic group as
opposed to ethnic group membership. To this end, the next section will examine the
128
influence of SES, specifically parental educational level and affluence, on the
development of executive functioning.
Socioeconomic Status and Executive Function
Affluence. According to the United States Census Bureau (2003) among all
children under age 18, African American children are three times as likely to live in
poverty (30 percent) as compared to their non-Hispanic White counterparts (10
percent). The researchers also reported that 23 percent of African Americans lived
below the poverty level as compared to 8 percent of non-Hispanic Whites. Duncan,
Yeung, Brooks-Gunn, and Smith (1998), using data from the National Longitudinal
Survey of Youth-Child Supplement (NLSY-CS), reported that income effects are
most influential during early childhood.
Other studies related to the current data set have reported on the effect of
affluence. Dearing, McCartney, and Taylor (2001) reported using the NICHD
SECCYD database that changes in income-to-needs ratio for non-poor individuals
had little effect, as compared to changes for poor individuals. Poverty level had the
greatest effects on measures of school readiness, receptive language, expressive
language, positive social behavior, and behavior problems. Thus change in income-
to-needs status matters more for children with less. In another study, the NICHD
Early Child Care Research Network (2005b) examined the influence of timing and
duration of poverty on maternal behaviors and subsequent child outcomes. They
reported that mothers in chronically low income families displayed steady decreases
in maternal sensitivity over time. Importantly, they noted that income gain was
129
associated with increases in quality of maternal behaviors, including maternal
sensitivity, as well as improved child cognitive and language outcomes.
Other researchers have noted that low-income mothers, compared to their
upper income counterparts, are less likely to engage in maternal behaviors found to be
associated with future reading and language abilities such as use of prompts, cues,
positive reinforcement, and use of open-ended questions (Portes, 1991). Mistry,
Biesanz, Taylor, Burchinal, and Cox (2004) examined the impact of the quality of
maternal-child interactions as one potential mediator between income and child
development. Using the NICHD data measures of maternal stimulation at 6, 15, 24,
and 36 month, the authors reported that (a) income and family processes, namely
maternal sensitivity, influenced both social and cognitive development (b) income
influenced cognitive development above and beyond family processes and (c)
behavioral outcomes were solely mediated by family processes. They also reported
that income had a greater impact on individuals living in poverty than those not living
in poverty. The authors write that, ?On average, low-income parents (or those facing
economic loss), compared with more affluent parents, are less child-centered and
nurturing in interactions with their children and are more parent-centered, rejecting,
and inconsistent when disciplining their children? (p. 728).
As expected in the current study, the largest effect size is the SES composite
score, while the income-to-needs ratio effect is also very large and nearly twice the
effect size for maternal education level and neighborhood safety. Welsh (2005)
examined the relationship between affluence (i.e., income-to-needs ratio) and 1
st
grade student?s performance on the TOH (i.e., planning efficiency score) and reported
130
a statistically significant relationship (p<.05) between affluence and performance on
the TOH with the correlation of .14 which accounting for 2% of the total variability in
the TOH scores. The current study found similar results for the TOH but higher
relationships between other measures of executive function and affluence. This
suggests that affluence as a single indicator does play a small but significant role in
performance on measures of executive function.
Parental Educational Level. Ardila, Rosseli, Matute, and Guajardo (2005)
examined the influence of educational variables on the development of executive
functions in children from Mexico and Colombia. The researchers administered eight
measures of executive function to participants (n=276) aged 5 to 14 who were
enrolled in either public or private school. It is important to note that years of
education is at best an imprecise measure for determining education. As Helms
(1997) reports, years of education fails to account for the quality of education which
is more likely to be inferior for African Americans. While the authors reported that
there were significant effects for age and type of school attended, they also noted that
parents? educational level contributed significantly over and above individual student
factors, particularly on measures of executive function that required more verbal
ability.
This finding was supported by the current study. The only measure of executive
function (WJ-R short term memory) that included a crystallized knowledge base had
the largest effect size. It is suspected, based on the large effect size differences on
measures of language, that differences between measures of executive function that
included crystallized knowledge will be confounded with maternal educational level
131
and maternal verbal ability. This is supported by the maternal scores on the PPVT-R
which were also quite large. While there is a more clear connection (i.e., verbal
ability) between measures of executive function that require a crystallized knowledge
base and SES, what is less clear is the relationship for measures of executive function
that require process-based knowledge.
Process-Based Measures of Executive Function
The next step is to examine effect sizes on other measures of executive function
that did not include crystallized knowledge bases which should then be theoretically
less driven by maternal education. Differences in student?s performance on the WJ-R
measure of long-term memory remained moderate, although much smaller than the
WJ-R short term memory. Conversely performance on the CPT commission errors
indicated no practical difference. This finding is inline with Mezzacappa?s findings
(2004). He examined executive attention, which is theorized to be a correlate of
executive function, and reported that, as expected, older children performed better on
the tasks than younger children. However, they also reported that African American
and Hispanic children did better at resisting interference than their peers. Finally,
examining the performance of participants in the current study on the TOH task,
which as a measure relies more on process than crystallized knowledge, there was a
moderate effect size difference.
Ethnic Differences
While the current data suggests between-group differences are difficult to tease
apart, Welsh (2005) more closely examined errors of participant?s (white/non-
Hispanic and Other) on TOH performance using the SECCYD dataset. Welsh
132
reported that white/non-Hispanic children performed significantly better on certain
TOH problems (i.e., items 2, 3, and 5) and had significantly higher total efficiency
planning scores. The study also reported that there were significant differences in the
type of errors committed by participants. White/non-Hispanic participants were more
likely to build towers with more than the required number of moves and the group
consisting of all other ethnicities in the sample was more likely to build towers on the
wrong peg. This suggests while/non-Hispanic children may have more difficulty
inhibiting the prepotent response while the the non-white ethnic group had more
difficulty with the basic visual spatial aspect of the task. While this provides some
insight into the between group performance differences on the TOH, it does not allow
for a more fine grain analyses of ethnic group differences.
Conclusion
The current study echo the findings of Noble, Norman, and Farah (2005) who
examined the neurocogntive correlates of 30 low SES and 30 middle SES
kindergarten children. The authors reported that, ?SES differences were associated
with disparities in performance in both the language and executive function systems?
(p. 83). The disparities appeared be greatest in knowledge-based tasks as compared to
process-based tasks (Campbell, Dollaghan, Needleman, & Janosky, 1997). While the
current study originally theorized that cultural differences in the amount and type of
scaffolding support provided would significantly contribute to differences in the
development of executive functioning, it appears that it not possible to separate
ethnicity and SES. Specifically it appears that SES influences the development of
language which in turn influences the development of executive functioning. This
133
topic should be further examined in order to provide culturally sensitive, educational
interventions so that all parents can best assist their children in optimal development.
Question 3. Does maternal verbal ability mediate the relationship between SES and
executive functioning?
This question was developed to better understand the relationship between
maternal verbal ability, SES, and executive functioning. As expected there was a
strong moderate correlation between the SES composite score and maternal PPVT-R
(r=.530) but conceptually it was expected that the addition of a measure of maternal
verbal ability would improve the fit of the overall model. It was theorized that
mothers who have higher verbal ability would provide greater verbal variety and
sophistication in their interactions with their children which over time may influence
the development of executive function. While SES refers to the amount and quality of
economic resources available to an individual, it does not specifically consider the
role of language in the development of executive function.
While several different models were examined, only one model appropriately
converged on a solution. Comparing the AIC index, the inclusion of maternal verbal
ability as a mediator did not improve the model fit. This may be because of the strong
relationship between SES and maternal verbal ability. Thus, maternal verbal ability
may not have provided additional non-redundant information.
The other proposed models either did not converge or resulted in a Heywood
case in which the estimated correlation between the indicator and its factor is greater
than 1. Heywood cases are caused by misspecification in the model such as outliers,
small samples sizes with only two indicators on a latent variable, or very high or low
population correlations. While the instability within the current model is not
134
surprising given what is now known about the proposed constructs, it not possible to
conceptually understand these findings.
Conclusion
The results of the current study suggest that the addition of maternal verbal
ability to model as a mediator between SES and executive function does not
significantly improve the fit of the model. While this finding is important, the overall
lack stability in the model should be considered when interpreting this finding.
Limitations
While the current study had a number of strengths, it is necessary to
understand the study?s limitations as well. First the study was limited by the
constructs that underlie the model. Obviously if the constructs are not valid or reliable
then the overall model is not representing what it was intended to represent.
Specifically it appeared that the constructs of early and late attention, early and late
maternal sensitivity, and early and late stimulation are suspect. Another notable
limitation is the lack of agreed-upon definition and measurement of executive
function. While the current model has utilized a commonly accepted definition of
executive function, the operationalization of executive function is limited by the
measures available in the archival data set. While the measures were chosen based on
theoretical and empirical literature, it is recognized that they, particularly TOH,
mostly likely were not solely measures of the process of interest. Additionally, there
are concerns about the measures of memory not best representing the type of memory
that theoretically underlies executive function. While there is some limited literature
to support the use of short-term and long-term memory, as related to both working
135
memory and executive functioning, it is important to note that they don?t represent
the traditional working memory tasks that are most commonly associated with
executive functioning. This is an important confounding factor causing the models
reviewed in the current study to be limited in their ability to define and operationalize
constructs.
Future Research
The current study attempted to measure the development of executive function
but executive function needs to be more clearly defined before developmental
mechanisms can be examined. The current study has resulted in few tentative findings
which suggest that language is a possible mechanism by which executive function
develops. This relationship between language and executive function is not well
established and frequently confounded by SES but should be a point of future
research possibly using the SECCYD data, specifically examining the various areas
of language ability (expressive vs. receptive).
Secondly the underlying model of executive function needs to be more clearly
understood. While the current model suggested a single construct of executive
function, in fact the underlying structure may be more similar to that depicted in
Figure 8. Before empirical research can be applied, these basic theoretical questions
need to be answered.
136
Figure 8
Possible Model of Executive Function
Finally, in order to understand important differences in how ethnicity and
culture influences the development of executive function, analysis may need to go
beyond mean differences in performance (Ellis & Seigler, 1997). For example, it may
be necessary to instead examine error type and planning time and to look for
similarities in response patterns within and across constructs of executive function.
This may provide additional information about the mechanism by which ethnicity
influences the development of executive function. It is also necessary that future
research attempts to disentangle the confounding relationship between SES and
Planning
Memory
Inhibition
Executive
Function
137
ethnicity by either examining the development of executive function within a single
SES strata or matching participants based on SES or suitable proxy.
138
Appendix A
Correlations Among All Study Variables n=470 (*p<.05, **p<.01)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
2 Neighborhood
Safety
at 1
st
grade
.139
**
3 Mother's
Education
at 1 Month
.477
**
.183
**
4 Positive
Regard
at 24 Months
0.09 .100
*
.202
**
5 Supportive
Presence
at 36 Months
.239
**
.153
**
.271
**
.256
**
6 Respect for
Autonomy
at 36 Months
.167
**
.178
**
.280
**
.206
**
.637
**
7 Cognitive
Stimulation
at 24 Months
.260
**
.122
**
.217
**
.223
**
.694
**
.420
**
8
Total Positive
Scaffolding
at 36 Months
.133
**
.082
.147
**
.178
**
.427
**
.225
**
.552
**
9 Cognitive
Stimulation
at 36 Months
.172
**
.181
**
.248
**
.546
**
.317
**
.227
**
.338
**
.303
**
10 Reynell:
Vocabulary
at 36 Months
.371
**
.248
**
.429
**
.153
**
.357
**
.325
**
.347
**
.194
**
.332
**
11 Reynell:
Expressive
Language at
36 Months
.197
**
.093
*
.286
**
.110
*
.197
**
.206
**
.212
**
.150
**
.177
**
.536
**
139
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
23 CBCL-Other:
Attention
at 54 Months
-.041 -.042
-.126
**
-.076
-.097
*
-.117
*
-.124
**
-.062
-.039
-.175
**
-.103
*
.009 -.067
-.112
*
-.127
**
-.131
**
-.175
**
-.122
**
-.218
**
24 CPT:
Omission
Errors
at 54 Months
-.068
-.104
*
-.123
**
-.068
-.105
*
-.111
*
-.163
**
-.131
**
-.177
**
-.264
**
-.154
**
-.005 -.079
-.205
**
-.210
**
-.263
**
-.209
**
-.201
**
-.227
**
25 Delay of
Gratification
at 54 Months
-.137
**
-.078
-.146
**
-.087
-.158
**
-.175
**
-.147
**
-.127
**
-.109
*
-.257
**
-.068 -0.03 -.016
-.175
**
-.174
**
-.154
**
-.153
**
-.155
**
-.273
**
26 CPT:
Commission
Errors
at 1
st
Grade
-.105
*
-.119
**
-.109
*
-0.03
2
-.091
*
-.165
**
-.061 -0.05
-.128
**
-.202
**
-.113
*
.031 -.012 -.054 .011 -0.07 -.066 -.064
-.152
**
27 WJ-R: Short
Term Memory
at 1
st
Grade
.226
**
.059
.192
**
.083 .089 .077
.153
**
.045 .058
.192
**
.098
*
.125
**
.024
.121
**
.085 .073
.106
*
.107
*
.185
**
28 WJ-R: Long
Term Memory
at 1
st
Grade
.224
**
.089
.302
**
.064
.250
**
.227
**
.234
**
.127
**
.135
**
.517
**
.346
**
.018 0.02
.197
**
.166
**
.138
**
.187
**
.204
**
.468
**
29 TOH: Total
Planning
Efficiency at
1
st
Grade
.179
**
.00
.113
*
-.046 .002 .065 .007 -.061 .013
.156
**
.015 .049 .068 -.014
.108
*
.098
*
0.09
.130
**
.161
**
30 PPVT-R:
Mother?s
Verbal Ability
.35
**
.134
**
.553
**
.144
**
.269
**
.274
**
.185
**
.165
**
.183
**
.413
**
.262
**
-.027 .043
.318
**
.260
**
.187
**
.291
**
.311
**
.396
**
140
20 21 22 23 24 25 26 27 28 29
21 WJ-R: Picture
Vocabulary
at 54 Months
.526
**
22 CBCL-
Mother:
Attention
at 54 Months
-.042 -.005
23 CBCL-Other:
Attention
at 54 Months
-.133
**
-.039
.463
**
24 CPT:
Omission
Errors
at 54 Months
-.254
**
-.172
**
.042
.112
*
25 Delay of
Gratification
at 54 Months
-.26
**
-.291
**
.047
.109
*
.212
**
26 CPT:
Commission
Errors
at 1
st
Grade
-.188
**
-.074
.176
**
.125
**
.099
*
.062
27 WJ-R: Short
Term Memory
at 1
st
Grade
.183
**
.277
**
-.021 -.022
-.091
*
-.135
**
-.065
28 WJ-R: Long
Term Memory
at 1
st
Grade
.440
**
.473
**
-.061
-.107
*
-.132
**
-.154
**
-.074
.304
**
29 TOH: Total
Planning
Efficiency at
1
st
Grade
.134
**
.119
**
-.135
**
-.083
-.160
**
-.057
-.153
**
.116
*
.099
*
30 PPVT-R:
Mother?s
Verbal Ability
.325
**
.447
**
-.112
*
-.105
*
-.110
*
-.203
**
-.059
.236
**
.355
**
.037
141
References
Aber, L., Rodriguez, M., Michel, W., & Shoda, Y., (1995). Mother-Toddler
Interactive Experiences as Precursors of Preschool Self-Regulation. Poster
presented at the biennial meeting of the Society for Research. Indianapolis,
IN.
Achenbach, T.M. (1992). Manual for the Child Behavior Checklist / 2-3 and 1992
Profile. Burlington, VT: University of Vermont Department of Psychiatry.
Achenbach, T.M., Howell, C.T., McConaughy, S.H., & Stanger, C. (1995). Six-Year
Predictors of Problems in a National Sample: IV. Young Adult Signs of
Disturbance. Journal of the American Academy of Child & Adolescent
Psychiatry. 37(7), 718-727.
Altepeter, T. S., & Johnson, K. A. (1989). Use of the PPVT-R for intellectual
Screening with adults: A caution. Journal of Psychoeducational Assessement,
7, 39-45.
Anastasi, A. (1982). Psychological testing (5th ed.). New York, NY: Macmillan
Publishing.
Ardila, A., Rosselli, M., Matute, E., & Guajardo, S. (2005). The influence of the
parents' educational level on the development of executive functions.
Developmental Neuropsychology, 28(1), 539-560.
Aspland, H. & Gardner, F. (2003). Observational measures of parent child
interaction. Child and Adolescent Mental Health, 8, 136-144.
Assel, M.A., Landry, S.H., Swank, P., Smith, K., & Steelman, L.M. (2003).
Precursors to mathematical skills: Examining the roles of visual-spatial skills,
142
executive processes, and parenting factors. Applied Developmental Science, 7,
27-38.
Atkinson, R.C. & Schiffrin, R.M. (1968). Human memory: a proposed system and its
control processes. In K.W. Spence & J.T. Spence (eds.) The Psychology of
Learning and Motivation Volume 2, Academic Press, London.
Bagely, A., Welsh, M.C., Retzlaff, P. Wolf, C., & Bryan, E. (2002). Towers of Hanoi
and London: Contribution of procedural learning and inhibition. Journal of
the International Neuropsychological Society, 8(2), p. 229.
Baird, G., Charman, T., Baron-Cohen, S., Cox, A., Swettenham, J., Wheelwright, S.,
& Drew, A. (2000). A screening instrument for autism at 18 month of age: A
six-year follow-up study. Journal of the American Academy of Child and
Adolescent Psychiatry, 39, 694-702.
Baldwin, A. L., Baldwin, C., & Cole, R. (1990). Stress-resistant families and stress-
resistant children. In J. Rolf, A. S. Masten, D. Chichetti, K. H. Neuchterlein,
and S. Weintraub. (Eds.). Risk and protective factors in the development of
psychopathology. New York: Cambridge University Press, pp. 257-280.
Banfield, J.F., Wyland, C.L., & Macrae, C. (2004). The cognitive neuroscience of
self-regulation. In Baumeister, R.F., & Vohs, K.D. (Eds.), Handbook of self-
regulation: Research, theory, and applications. (pp. 62-83). New York, NY,
US: Guilford Press.
Barkley, R. A. (1994). The assessment of attention in children. In G. R. Lyon (Ed.),
Frames of Reference for the Assessment of Learning Disabilities: New views
on measurement issues (pp. 69-105). Baltimore: Paul H. Brooks Publishing
143
Company.
Barkley, R. A. (1997). ADHD and the nature of self-control. New York: Guilford
Becker, M. G., Isaac, W., & Hynd, G. W. (1987). Neuropsychological development
of nonverbal behaviors attributed to ?frontal lobe? functioning.
Developmental Neuropsychology, 3, 275-298.
Befera, M., & Barkley, R. (1985). Hyperactive and normal boys and girls : Mother-
child interaction, parent psychiatric status, and child psychopathology.
Journal of Child Psychology and Psychiatry, 26, 439-452.
Benson, J. B. (1997). The development of planning: Its About Time. In S. Friedman
and E. Skolnick (Eds.), The developmental psychology of planning. Erlbaum.
Bishop, D.V.M., Aamodt-Leeper, G., Creswell, C., McGurk, R., & Skuse, D.H.
(2001). Individual differences in cognitive planning on the Tower of Hanoi
Task: Neuropsychological maturity or measurement error? Journal of Child
Psychology Psychiatry, 42, 551-556.
Bivens, J. & Berk, L. (1990). A Longitudinal Study of the Development of
Elementary School Children?s Private Speech, Merrill-Palmer Quarterly, 36
(4), 443-463.
Breznitz, Z., & Friedman, S.L. (1988). Toddlers' ability to concentrate: The influence
of maternal depression. Journal of Child Psychology and Psychiatry, 29, 267-
279.
Bornstein, M. H., Haynes, O. M., O'Reilly, A. W., & Painter, K. (1996). Solitary and
collaborative pretense play in early childhood: Sources of individual variation
in the development of representational competence. Child Development, 67,
144
2910-2929.
Bornstein, M. H., Haynes, O. M., Pascual, L., Painter, K. M., & Galper?n, C. (1999).
Play in two societies: Pervasiveness of process, specificity of structure. Child
Development, 70, 317-331.
Borys, S. V., Spitz, H. H., & Dorans, B. A. (1982). Tower of Hanoi performance of
retarded young adults and nonretarded children as a function of solution
length and goal state. Journal of Experimental Child Psychology, 33, 87-110.
Breznitz, Z., & Friedman, S.L. (1988). Toddlers' concentration: Does maternal
depression make a difference? Journal of Child Psychology & Psychiatry
(29), 267-79.
Brocki, C., & Bohlin, G. (2004). Executive functions in children aged 6 to 13: A
dimensional and developmental study. Developmental Neuropsychology, 26,
571?593.
Bronfenbrenner, U., & Ceci, S. (1994). Nature/Nuture reconceptualized in
developmental perspective: A bioecological model. Psychological Review, 4,
568-586.
Brown, A. L., & Campione, J. C. (1990). Communities of learning and thinking, or a
context by any other name. Contributions to Human Development, 21, 108-
126.
Burchinal, M. R., Campbell, F. A., Bryant, D. B., Wasik, B. H., & Ramey, C. T.
(1997). Early intervention and mediating processes in cognitive performance
of children of low-income African-American families. Child Development,
68, 935-954.
145
Campbell, T., Dollaghan, C., Needleman H, & Janosky, J. (1997). Reducing bias in
language assessment: Processing-dependent measures. Journal of Speech,
Language and Hearing Research, 40, 519-525.
Carvajal, H., Kixmiller, J., Knapp, M., Vitt, J., & Weaver, K. A. (1991). The use of
the PPST and intelligence tests in teacher education programs. Bulletin of the
Psychonomic Society, 29, 192-194.
Connell, C.M., & Prinz, R.J. (2002). The impact of childcare and parent-child
interactions on school readiness and social skills development for low-income
African American children. Journal of School Psychology, 40, 177-193.
Conners, C.K., Epstein, J.N., Angold, A., & Klaric, J. (2003). Continuous
Performance Test Performance in a Normative Epidemiological Sample.
Journal of Abnormal Child Psychology, 31, 555-562.
Crowley, S.L., & Fan, X. (1997). Structural equation modeling: basic concepts and
applications in personality assessment research. Journal of Personality
Assessment, 68 (3), 508-531.
Curran, P.J., West, S.G., & Finch, J. (1996). The robustness of test statistics to non-
normality and specification error in confirmatory factor analysis.
Psychological Methods, 1, 16-29
Dallaire, D. H., & Weinraub, M. (2005). The stability of parenting behaviors over the
first six years of life. Early Childhood Research Quarterly, 20, 201-219.
de la Ossa, J., & Gauvain, M. (2001). Joint attention by mothers and children while
using plans. International Journal of Behavioral Development, 25, 176-183.
Dean, R., & Woodcock, R.W. (1999). The WJ-R and Bateria-R in
146
Neuropsychological Assessment . Itasca, IL: Riverside.
Dearing, E., McCartney, K., & Taylor, B. A. (2001). Change in family income-to-
needs matters more for children with less. Child Development, 72, 1779-1793.
Deater-Deckard, K., Pylas, M. V., & Petrill, S. A. (1997). The Parent-child
interaction system (PARCHISY). London, UK: Institute of Psychiatry.
Devine, S., Welsh, M.C., Retzlaff, P., Yoh, M., & Adams, C. (2001). Explicit and
implicit cognitive processes underlying Tower of Hanoi performance. Journal
of the International Neuropsychological Society, 2, 250.
Diamond, A. (1990). The development and neural bases of memory functions as
indexed by the AB and delayed response tasks in human infants and infant
monkeys. Annals of the New York Academy of Sciences, 608, 267-317.
Dodici, B.J., Draper, D.C., & Peterson, C.A. (2003). Early parent-child interactions
and early literacy development. Topics in Early Childhood Special Education,
23, 124-136.
Duffy, J.D., & Campbell, J.J. (2001). Regional prefrontal syndromes: A theoretical
and clinical overview. In S.P. Salloway, P.F. Malloy, & J.D. Duffy (Eds.), The
frontal lobes and neuropsychiatric illness (pp.113-124). Washington, DC:
American. Psychiatric Publishing Inc.
Duncan, G.J., Yeung, W.J., Brooks-Gunn, J., & Smith, J.R. (1998). How much does
childhood poverty affect the life chances of children? American Sociological
Review, 63, 406-423.
Dunn, L. M. & Dunn, L. M. (1981). Peabody Picture Test-Revised Manual for
Forms L and M. Circle Pines, MN: American Guidance Service.
147
Eddy, J. M., Dishion, T. J. & Stoolmiller, M. (1998). The analysis of intervention
change in children and families: Methodological and conceptual issues
embedded in intervention studies. Journal of Abnormal Psychology, 26(1),
53-69.
Egeland, B., & Hiester, M. (1993). Teaching task rating scales. Institute of
Child Development, University of Minnesota.
Ellis, S. A., & Siegler, R. S. (1997). Planning as a strategy choice. Why don't
children plan when they should? In S. Friedman & E. Scholnick (Eds.), Why,
how, and when do we plan? The developmental psychology of planning (pp.
183-208). Hillsdale, NJ: Erlbaum.
Emick, J.E., & Welsh, M. (2005). Association between formal operational thought
and executive function as measured by the Tower of Hanoi-Revised
Learning and Individual Differences, 15, 177-188.
Epstein, J.N., Erkanli, A., Conners, C.K., Klaric, J., Costello, J.E., & Angold, A.
(2003) Relations Between Continuous Performance Test Performance
Measures and ADHD Behaviors. Journal of Abnormal Child Psychology, 31,
543-554.
Espy, K. A., Kaufmann, P. M., Glisky, M. L., & McDiarmid, M. D. (2001). New
procedures to assess executive functions in preschool children. Clinical
Neuropsychologist, 15, 46-58.
Espy, K.A., McDiarmid, M.D., Cwik, M.F., Senn, T.E., Hamby, A., & Stalets, M.M.
(2004). The contributions of executive functions to emergent mathematic
skills in preschool children. Developmental Neuropsychology, 26, 465-486.
148
Fan, J., Wu, Y., Fossella, J.A., & Posner, M.I. (2001). Assessing the heritability of
attentional networks BMC Neuroscience, 14.
Farver, J. M., & Howes, C. (1993). Cultural differences in American and Mexican
mother-child pretend play. Merrill-Palmer Quarterly, 39, 344 -358.
Fein, G., & Fryer, M. (1995). Maternal contributions to early symbolic play
competence. Developmental Review, 15, 367-381.
Flanagan, D. P., McGrew, K. S. & Ortiz, S. O. (2000). The Wechsler Intelligence
Scales and Gf-Gc theory: A contemporary interpretive approach. Boston,
MA: Allyn & Bacon.
Flick, L.H., & McSweeney, M. (1987). Measures of mother-child interaction: A
comparison of three methods. Research in Nursing and Health, 10, 129-137.
Frazier, P.A., Tix, A.P., & Barron, K.E. (2004). Testing moderator and mediator
effects in counseling psychology research. Journal of Counseling Psychology,
51(1), 115-134.
Freund, L. S. (1990). Maternal regulation of children?s problem-solving behavior and
its impact on children?s performance. Child Development, 61, 113?126.
Friedman, N. P., Miyake, A., Corley, R. P., Young, S. E., DeFries, J. C., & Hewitt, J.
K. (2006). Not all executive functions are related to intelligence.
Psychological Science, 17, 172-179.
Friedman, S.L., & Scholnick, E.K. (1997). An evolving ?blueprint? for planning:
Psychological requirements, task characteristics, and social-cultural
influences. In S.L. Friedman & E.K. Scholnick (Eds.), The developmental
psychology of planning: Why, how, and when do we plan? (pp. 3?22).
149
Mahwah, NJ: Erlbaum.
Friedman, S.L., & Sherman, T.L. (1985). Mothers as mediators of cognitive
development: A coding manual (ERIC Clearinghouse on Elementary and
Early Childhood Education, ED 272 318).
Frye, D., Zelazo, P.D., & Palfai, T. (1995). Theory of mind and rule based
reasoning. Cognitive Development, 10, 483-527.
Fuster, J. M. (1997). The prefrontal cortex: Anatomy, physiology, and
neuropsychology of the frontal lobe (3rd ed.). New York: Raven.
Gardner, F. (2000). Methodological issues in the use of observational methods for
measuring parent child interaction. Clinical Child and Family Psychology
Review, 3, 185-198.
Gardner, F., Ward, S., Burton, J. & Wilson, C. (2003). Joint play and the early
development of conduct problems in children: A longitudinal observational
study of pre-schoolers. Social Development, 12, 361-379.
Gauvain, M., & Rogoff, B. (1989). Collaborative problem solving and
children's planning skills. Developmental Psychology, 25, 139-151.
Gioia, G.A., Isquith, S.C., Guy, L., & Kenworthy, L. (2000). Behavior Rating I
Inventory of Executive Function. Odessa, FL: Psychological Assessment
Goel, V., Pullara, D., & Grafman, J. (2001). A computational model of frontal lobe
dysfunction: Working memory and the Tower of Hanoi. Cognitive Science, 25
(2), 287-313.
Goldberg, T., Saint-Cyr, J., & Weinberger, D. (1990). Assessment of procedural
learning and problem solving in schizophrenic patients by Tower of Hanoi
150
type tasks.Journal of Neuropsychology, 2, 165-173.
Golden, C. J. (1981). Luria Nebraska Children's Battery: Theory and formulation. In
G.W. Hynd & J.E. Obzut (Eds.), Neuropsychological assessment of the
school-aged child (pp. 277-302). NT: Grune and Stratton.
Greenberg, M. T., Lengua, L. J., Coie, J. D., Pinderhughes, E. E., & Conduct
Problems Prevention Research Group (1999). Predicting developmental
outcomes at school entry using a multiple-risk model: Four American
communities. Developmental Psychology, 35, 403-417.
Hancock, G. R., & Mueller, R. O. (Eds.). (2006). Structural Equation Modeling: A
Second Course. Greenwich, CT: Information Age Publishing, Inc.
Halperin, J. M., Sharma, V., Greenblatt, E., & Schwartz, S. T. (1991). Assessment of
the continuous performance test: Reliability and validity in a nonreferred
sample. Journal of Consulting and Clinical Psychology, 3, 603-608.
Helms, J.E. (1997). The triple quandary of race, culture, and social class in
standardized cognitive ability testing. In D.P. Flanagan, J.L. Genshaft, & P.L.
Harrison (Eds.), Contemporary intellectual assessment: Theories, tests, and
issues (pp. 517?532). New York: Guilford Press.
Henderson, B., & Moore, S. G. (1980). Children?s responses to object differing in
novelty in relation to level of curiosity and adult behavior. Child
Development, 51, 457-465.
Hess, R.D., & Shipman, V.C. (1965). Early experience and the socialization of
cognitive modes in children. Child Development, 36 (4), 869-886.
Hoyle, R. H. (ed.). Structural Equation Modeling, Concepts, Issues, and Applications,
151
Sage Publications, Thousand Oaks, CA, 1995.
Hoyle, R.H., & Panter, A.T. (1995). Writing about structural equation modeling. In
R.H. Hoyle (Ed.), Structural Equation Modeling: Concepts, Issues, and
Applications (pp. 158?176).Thousand Oaks, CA: Sage.
Hubbs-Tait, L., Culp, A. M., Culp, R. E., & Miller, C. E. (2002). Relation of maternal
cognitive stimulation, emotional support, and intrusive behavior during Head
Start to children?s kindergarten cognitive abilities. Child Development, 73,
110 ?131.
Hutton, U. M. Z., & Towse, J. N. (2001). Short-term memory and working memory
as indices of children's cognitive skills. Memory, 9, 383-394.
Isquith, P.K., Gioia, G.A., & Espsy, K.A. (2004). Executive Function in preschool
children: Examination through everyday behavior. Developmental
Neuropsychology, 26(1), 403-422.
Johnson, J., Christie, J., & Yawkey, T. (1999). Play and early childhood
development. (2nd ed.). New York: Longman.
Joseph, R. M., McGrath, L., & Tager-Flusberg, H. (2005). Executive dysfunction and
its relationship to language impairment in school-age children with autism.
Developmental Neuropsychology. 27(3), 361?378.
Kalback, S.H. (2004). The assessment of developmental language differences,
executive functioning, and social skills in deaf children. Unpublished
dissertation. Gallaudet University.
Kaufman, A.S., & Kaufman, N.L. (2004). Kaufman Assessment Battery for Children
(2nd ed., K-ABC-II). Circle Pines, MN: American Guidance Service.
152
Keller, H., Lohaus, A., Volker, S., Elben, C., Ball, J. (2003). Warmth and
contingency and their relationships to maternal attitudes toward parenting. The
Journal of Genetic Psychology, 164, 275-292.
Kline, R. B. (2006). Reverse arrow dynamics: Formative measurement and feedback
loops. In G. R. Hancock & R. O. Mueller (Eds.), Structural Equation
Modeling: A Second Course (pp. 43-68). Greenwich, CT: Information Age
Publishing.
Kopecky, H., Chang, H. T., Klorman , R., Thatcher, J. E., & Borgstedt, A. D. (2005).
Performance and private speech of children with Attention-
Deficit/Hyperactivity Disorder while taking the Tower of Hanoi test: Effects
of depth of search, diagnostic subtype, and methylphenidate. Journal of
Abnormal Child Psychology 33, 625-638.
Landry, S.H., Miller-Loncar, C.L., Smith, K.E., & Swank, P.R., (2002). The role of
early parenting in children?s development of executive processes.
Developmental Neuropsychology, 21(1), 15?41.
Landry, S. H., Smith, K. E., Miller-Loncar, C. L., & Swank, P. R. (1997). Predicting
cognitive-linguistic and social growth curves from early maternal behaviors in
children at varying degrees of biologic risk. Developmental Psychology, 33,
1?14.
Landry, S. H, Smith, K. E., Swank, P. R., & Miller-Loncar, C. L. (2000). Early
maternal and child influences on children?s later independent cognitive and
social functioning. Child Development, 71, 358?375.
Lewis, C., Kier, C., Hyder, C., Prenderville, N, Pullen, J., & Stephens, A. (1996).
153
Observer influences on fathers and mothers: and experimental manipulation of
the structure and function of parent infant conversation. Early Development &
Parenting, 5, 57-68.
Little, L.D., Lindenberger, U., & Nesselroade, J.R. (1999). On selecting indicators for
multivariate measurement and modeling with latent variables: When ?good?
indicators are bad and ?bad? indicators are good. Psychological Methods, 4,
192?211.
Lock, C., Welsh, M.C., Adams, C., & Kurtz, A. (2002). Tower of Hanoi: Influence of
strategy instruction and extended practice on performance. Journal of the
International Neuropsychological Society, 8, 229.
Luria, A.R. (1961) The role of speech in the regulation of normal and abnormal
behavior. Liveright Publishing Corporation, New York.
MacCallum, R. C., & Austin, J. T. (2000). Applications of structural equation
modeling in psychological research. Annual Review of Psychology, 51, 201-
226.
Malloy, P.F., & Richardson, E.D. (2001). Assessment of frontal lobe functions. In
S.P. Salloway, P.F. Malloy, & J.D. Duffy (Eds.), The frontal lobes and
neuropsychiatric illness (pp.125-138). Washington, DC: American Psychiatric
Publishing, Inc.
Marsh, H.W., Hau, K.T., Balla, J. R., & Grayson, D. (1998). Is more ever too much?
The number of indicators per factor in confirmatory factor analysis.
Multivariate Behavioral Research, 33, 181?220.
McGrew, K. (1986). A review of the differential predictive validity of the
154
Woodcock- Johnson Scholastic Aptitude clusters. Journal of
Psychoeducational Assessment, 4, 307-3l7.
Mezzacappa, E., (2004). Alerting, orienting, and executive attention:
developmental and socio-demographic properties in an epidemiological
sample of young, urban children. Child Development 75, 1373? 1386.
Milner, B. (1963) Effects of different brain lesions on card sorting. Archives of
Neurology, 9, 90-100.
Mistry, R. S., Biesanz, J. C., Taylor, L. C., Burchinal, M., & Cox, M. J. (2004).
Family income and its relation to preschool children's adjustment for families
in the NICHD study of early child care. Developmental Psychology, 40, 727-
745.
Miyake, A., Friedman, N.P., Emerson, M.J., Witzki, A. H., Howerter, A., & Wager,
T.D. (2000). The unity and diversity of executive functions and their
contributions to complex ?frontal lobe? tasks: A latent variable analysis.
Cognitive Psychology, 41, 49-100.
Morrison, E. F., Rimm-Kauffman, S., & Pianta, R. C. (2003). A longitudinal study of
mother-child interactions at school entry and social and academic outcomes in
middle school. Journal of School Psychology, 41(3), 185?200.
Mundy, P., Fox, N., & Card (2003). Joint attention, EEG coherence and early
vocabulary. development. Developmental Science, 6, 48 - 54.
Nabors, N. A., Evans, J. D., & Strickland, T. L. (2000). Neuropsychological
assessment and intervention with African Americans. In E. Fletcher-Janzen,
T.L. Strickland,and C.R. Reynolds (eds.), Handbook of cross-cultural
155
neuropsychology. (pp. 31-42) New York: Kluwer Academic/Plenum
Publishers.
NICHD Early Child Care Research Network. (1993). The NICHD Study of Early
Child Care: A comprehensive longitudinal study of young children's lives.
(ERIC Document Reproduction Service No. ED3530870)
NICHD Early Child Care Research Network. (1999). Child care and mother-child
interaction in the first three years of life. Developmental Psychology, 35,
1399-1413.
NICHD Early Child Care Research Network. (2000). The relation of child care to
cognitive and language development. Child Development, 71 (4), 960-980.
NICHD Early Child Care Research Network (2001). Nonmaternal Care and Family
Factors in Early Development: An Overview of the NICHD Study of Early
Child Care. Applied Developmental Psychology, 22, 457-4.
NICHD Early Child Care Research Network. (2003). Do children?s attention
processes mediate the link between family predictors and school readiness?
Developmental Psychology, 39, 581-593.
NICHD Early Child Care Research Network. (2005a). Predicting individual
differences in attention, memory, and planning in first graders from
experiences at home, child care, and school. Developmental Psychology, 41,
99-114.
NICHD Early Child Care Research Network. (2005b). Duration and developmental
timing of poverty and children?s cognitive and social development from birth
through third grade. Child Development, 76 (4), 795-810.
156
NICHD Study of Early Child Care (1995) Phase I Manuals. Retrieved June 13, 2005
from http://secc.rti.org/
NICHD Study of Early Child Care (1999) Phase I Instrument Document.
Retrieved June 13, 2005 from http://secc.rti.org/
NICHD Study of Early Child Care (2001) Phase II Manuals. Retrieved June 13, 2005
from http://secc.rti.org/
NICHD Study of Early Child Care and Youth Development (n.d.) Phase II Instrument
Document. Retrieved June 13, 2005 from http://secc.rit.org/
Nichols, K., Fox, N., & Mundy, P. (2005). Joint attention, self-recognition and
neurocognitive functioning. Infancy, 7, 35-52.
Noble, K. G., Norman, M. F., & Farah, M. J. (2005). Neurocognitive correlates of
socioeconomic status in kindergarten children. Developmental Science, 8, 74-
87.
Osmon, D. C., Smerz, J. M., Braun, M. M., & Plambeck, E. (2006). Processing
abilities associated with math skills in adult learning disability. Journal of
Clinical and Experimental Neuropsychology, 28, 1-12.
Ornstein, P. A., Haden, C. A., & Hedrick, A. M. (2004). Learning to remember:
Social-communicative exchanges and the development of children?s memory
skills. Developmental Review, 24, 374?395.
Ozonoff,S., Pennington, B.F., & Rogers,S. J. (1991).Executive function deficits in
high-functioning autistic individuals: Relationship to the theory of mind.
Journal of Child Psychology and Psychiatry. 32, 1081?1105.
Passler, M.A., Isaac, W., & Hynd, G.W. (1985). Neuropsychological development of
157
behavior attributed to frontal lobe functioning in children. Developmental
Neuropsychology, 1, 349-370.
Perner, J., & Lang, B. (2000). Theory of mind and executive function: Is there a
Developmental relationship? In S. Baron-Cohen, H. Tager-Flusberg, & D.
Cohen (Eds.), Understanding other minds: Perspectives from developmental
cognitive neuroscience (2nd ed., pp. 150?181).Oxford: Oxford University
Press.
Piaget, J. (1954). The construction of reality in the child. New York, NY: Basic
Books.
Piaget, J. (1971). The theory of stages in cognitive development. In D. R. Green
(Ed.), Measurement and Piaget. (pp. 1-11). New York: McGraw.
Pianta, R. C. (1994). Patterns of relationships between children and kindergarten
teachers. Journal of School Psychology, 32, 15-31.
Pianta, R. C., & Harbers, K. L. (1996). Observing mother and child behavior in a
problem-solving situation at school entry: Relations with academic
achievement. Journal of School Psychology, 34(3), 307-322.
Portes, P. R. (1991). Assessing Children's Cognitive Environments through Parent-
Child Interaction: Estimation of a general zone of proximal development in
relation to Scholastic Achievement. Journal of Research in Education 24 (3),
Spring, 30-38.
Qi, C.H., Kaiser, A. P., Milan, A.P., Milan, S.E., Yzquierdo, Z., & Hancock, T.B.
(2003). The performance of low-income African American Children on the
Preschool Language Scale?3. Journal of Speech, Language, and Hearing
158
Research, (46), 576?590.
Rescorla, L., & Alley, A. (2001). Validation of the Language Development Survey
(LDS): A parent report tool for identifying language delay in toddlers. Journal
of Speech, Language, and Hearing Research, 44, 434?445.
Reynell, J. (1990). Reynell Developmental Language Scales (U.S. Edition). Los
Angeles, CA: Western Psychological Services.
Riccio, C. A., Reynolds, C. R., & Lowe, P. (2001). Clinical applications of
continuous performance tests: Measuring attention and impulsive responding
in children and adults. New York: John Wiley & Sons, Inc.
Rogoff, B. (1990). Apprenticeship in thinking: Cognitive development in social
context. New York: Oxford University Press.
Rogoff, B. (2003). The cultural nature of human development. New York:
Oxford University Press.
Roopnarine, J. L., Johnson, J. E., & Hooper, F. H. (Eds.). (1994). Children?s play in
diverse cultures. Albany, NY: SUNY Press.
Rosvold, H. E., Mirsky, A. F., Sarason, I., Bransome Jr., E. D., & Beck, L. H. (1956).
A continuous performance test of brain damage. Journal of Consulting
Psychology, 20, 343-350.
Ruff, H.A., & Rothbart, M.K. (1996). Attention in early development. New York:
Oxford University Press.
Russell, J. (ed.) (1997) Autism as an executive disorder. Oxford: OUP.
Sabbagh, M. A., Xu, F., Carlson, S. M., Moses, L. J., Lee, K., (2006). Executive
functioning and of mind in preschool children from Beijing, China:
159
Comparisons with U.S. preschoolers, Psychological Science, 17, 74-81.
Satorra, A. & Bentler, P.M. (1988). Scaling Corrections for Chi-square Statistics in
Covariance Structure Analysis. Proceedings of the American Statistical
Association, 308-313.
Saxton, M. (1997). The contrast theory of negative input. Journal of Child Language,
24,139-161.
Schachar, R., & Logan, G. D. (1990). Impulsivity and inhibitory control in normal
development and childhood psychopathology. Developmental Psychology, 26,
710-720.
Schachar, R., & Logan, G. D. (1990). Impulsivity and inhibitory control in normal
development and childhood psychopathology. Developmental Psychology, 26,
710-720.
Schmand, B., Brand, N., & Kuipers, T. (1992). Procedural learning of cognitive and
motor skills in psychotic patients. Schizophrenia Research, 8, 157-170.
Scholnick, E.K., & Friedman, S.L. (1987). The planning construct in the
psychological literature. In S.L. Friedman, E.K. Scholnick, & R. Cocking
(Eds.), Blueprints for thinking: The role of planning in cognitive development
(pp. 1?38). New York and London: Cambridge University Press.
Scholnick, E. K., & Friedman, S. L. (1993). Planning in context: Developmental
and situational considerations. In M. E. Lachman (Ed.), Planning
and control processes across the life span (pp. 145-167). East
Sussex, England: Erlbaum.
Scholnick, E.K., Friedman, S.L., & Wallner-Allen, K.E., (1997). What do they really
160
measure? A comparative analysis of planning tasks. In S. L. Friedman & E.K.
Scholnick (Eds.), The developmental psychology of planning: Why, how and
when do we plan? (pp. 127-156). Hillsdale, NJ: Lawrence Erlbaum
Associates.
Senn, T.E., Espy, K.A., & Kaufmann, P.M. (2004). Using path analysis to understand
Executive Function organization in preschool children. Developmental
Neuropsychology, 26(1), 445-464.
Sergeant, J.A., Geurts, H., Oosterlaan, J. (2002) How specific is a deficit of executive
functioning for attention-deficit/hyperactivity disorder? Behavior Brain
Research, (130) 3?28.
Sethi, A., Mischel, W., Aber, J., Shoda, Y., & Rodriquez, M. (2000). The role of
strategic attention in the development of self-regulation: Predicting
preschoolers?delay of gratification from mother?toddler. Developmental
Psychology, 36, 1488?1495.
Shoda, Y., Mischel, W., & Peake, P. K. (1990). Predicting adolescent cognitive and
self-regulatory competencies from preschool delay of gratification:
Identifying diagnostic conditions. Developmental Psychology, 26, 978-986.
Sigel, I.E., Stinson, E.T. & Flaugher, J. (1991). Socialization of representational
competence in the family: The distancing paradigm. In L. Okagaki & R.J.
Sternberg (Eds.), Directors of development: Influences on the development of
children's thinking, 121-44. Hillsdale, NJ: Lawrence Erlbaum.
Simon, H.A. (1975). The functional equivalence of problem solving skills. Cognitive
Psychology, 18, 643-661.
161
Snitz, B.E., Bieliauskas, L.A., Crossland, A., Basso, M.R., & Roper, B. (2000). The
PPVT-R as an estimate of premorbid intelligence in older adults. Clinical
Neuropsychology, 14, 181-186.
Stevens, E., Blake, J., Vitale, G., & MacDonald, S. (1998). Mother-infant object
involvement at 9 and 15 months: Relation to infant cognition and early
vocabulary. First Language, 18, 203?222.
Stuss D.T., & D.F. Benson (1986). The frontal lobes. New York: Raven Press.
Swanson, H.L., & Cooney, J.B. (1991). Learning disabilities and memory. In B.Y.L.
Wong (Ed.), Learning about learning disabilities (pp. 103-127). San Diego,
CA: Academic Press, Inc.
Tamis-LeMonda, C. S., Chen, L. A., & Bornstein, M. H. (1998). Mothers? knowledge
about children?s play and language development: Short-term stability and
interrelations. Developmental Psychology, 34, 115?124.
Tamis-LeMonda, C.S., Shannon, J.D., Cabrera, N.J., Lamb, M.E. (2004). Fathers and
mothers at play with their 2 and 3 year-olds: Contributions to language and
cognitive development. Child Development, 75, 1806-1820.
Valenzuela, M. (1997). Maternal sensitivity in a developing society: The
context of urban poverty and infant chronic undernutrition. Developmental
Psychology, 33, 845?855.
Vandell, D. L., & Pierce, K. M. (1998). Measures used in the Study of After-School
Care: Psychometric properties and validity information. Unpublished manual,
University of Wisconsin-Madison.
Vandermaas-Peeler M., Way E., & Umpleby, J. (2003). Parental guidance in a
162
cooking activity with preschoolers. Journal of Applied Developmental
Psychology, 24, 75-89.
Vaughan, A., Mundy, P., Crowson, M., Burnette, C., Delgado, C., Gomez, Y., Meyer,
J.,Neal, R., & Pomares, Y. (2003). Child, caregiver and temperament
contributions to infant joint attention. Infancy, 4, 603?616.
Vygotsky, L.S. (1962). Thought and Language. Cambridge, MA: MIT Press.
Vygotsky, L. S. (1978). Mind in society. Cambridge, MA: MIT Press.
Wallner, K. E. (1996). Executive functioning and its relation to planning skill in
seven-year-old children. Unpublished dissertation. University of Maryland.
Welsh, M.C. (1991). Rule-guided behavior and self-monitoring on the Tower of
Hanoi disk-transfer task. Cognitive Development, 6, 59-76.
Welsh, M.C. (2002). Developmental and clinical variations in executive functions.
In U. Kirk & D. Molfese (Eds.), Developmental variations in language and
learning. (pp. 139-186). New Jersey: LEA.
Welsh, M.C. (2005). An examination of Tower of Hanoi performance in children
participating in the NICHD Study of Early Child Care. Unpublished report.
University of Northern Colorado.
Welsh, M.C., Friedman, S.L., Spieker, S.J. (2006). Executive functions in developing
children: Current conceptualizations and questions for the future. In K.
McCartney & D. Phillips, Blackwell Handbook of Early Childhood
Development. Oxford: Blackwell Publishing.
Welsh, M.C., & Huizinga, M. (2001). The development and preliminary validation of
the Tower of Hanoi-Revised. Assessment, 8 (2), 167-176.
163
Welsh, M.C. & Huizinga, M. (2005). Tower of Hanoi disk-transfer task: Influences of
strategy knowledge and learning on performance. Learning and Individual
Differences, 4, 283-298.
Welsh, M. C., Huizinga, M., Granrad, M., Cooney, J., Adams, C., & van der Molen,
M. (2002). A structural equation model of executive function in young adults.
A Paper Presented Annual meeting of the International Neuropsychological
Society, Toronto, Canada.
Welsh, M. C., & Pennington, B. F. (1988). Assessing frontal lobe functioning in
children: Views from developmental psychology. Developmental
Neuropsychology, 4, 199?230.
Welsh, M.C., Pennington, B.F., & Groisser, D.B. (1991). A normative-developmental
study of executive function: A window of prefrontal function in children.
Developmental Neuropsychology, 7(2), 131-149.
Welsh, M.C., Satterlee-Cartmell, T., & Stine, S. (1999). Towers of Hanoi and
London: Contribution to working memory and inhibition to performance.
Brain and Cognition, 41, 231-242.
Wertsch, J. V. (1979). From social interaction to higher psychological processes.
Human Development, 22, 1?22.
Wood, D. J., & Middleton, D. (1975). A study of assisted problem- solving. British
Journal of Psychology, 66, 181-191.
Woodcock, R. W., & Mather, N. (1989). Woodcock-Johnson Psychoeducational
Battery--Revised. Allen, TX: DLM Teaching Resources.
Yu, C. Y., & Muthen, B. O. (2002). Evaluation of model fit indices for latent variable
164
models with categorical and continuous outcomes. (Technical Report). Los
Angeles, CA: University of California, Los Angeles, Graduate School of
Education and Information Studies.
Zimmerman, I. L., Steiner, V. G., & Pond, R. E. (1992). The Preschool Language
Scale-3: Examiner?s Manual. The Psychological Corporation: Harcourt Brace,
Inc.

Jessica E. Emick
Home Address: Contact Information:
1381 E. University Ave. #302 Email: jessica.emick@gmail.com
Las Vegas, NV 89119
Education
University of Maryland, College Park, MD
Doctor of Philosophy, Expected May, 2007
Major: School Psychology
Dissertation: ?The development of executive functioning and maternal behaviors:
Secondary analysis of longitudinal data?
Dissertation Chair: Dr. William Strein, PhD
G.P.A. 3.8/4.0
University of Northern Colorado, Greeley, CO
Master of Arts Degree, May 2003
Major: Human Neuropsychology
Thesis: ?Formal operational reasoning as a predictor of performance on the Tower of
Hanoi-Revised?
Thesis Advisor: Dr. Marilyn Welsh, PhD
G.P.A. 3.8/4.0
University of Toledo, Toledo, OH
Bachelor of Arts Degree, May 2001
Major: Psychology with a developmental concentration
Minor: Biological Chemistry
Senior Project: ?The effect of contour deletion on object recognition in preschool
children?
Graduated with departmental and university honors (G.P.A. 3.5/4.0)
Honors and Awards
CAPS Dissertation Award, University of Maryland Department of Counseling and
Personnel Services, 2006
Goldhauber Travel Grant, University of Maryland Graduate School, 2006
Colorado Graduate Fellowship, University of Northern Colorado Department of
Psychology, 2002-2003
Excellence in Psychology Award, University of Toledo, 2001 (Cash Award)
Dean?s and President?s List, University of Toledo, 1997-2001
Publications
Emick, J., Mann, H., & Kopriva, R. (2006). New ELLs and large-scale test validity.
Manuscript under preparation.
Kopriva, R., Emick, J., & Hipolito-Delgado, C. (2006). Do proper accommodation
assignments make a difference? Examining the impact of improved decision-
making on scores for ELLs. Manuscript accepted for publication.
Kopriva, R., Winter, P., Emick, J., & Chen, S. (2006). Achieving accurate results for
diverse learners- Access-based item development. Manuscript accepted for
publication.
Strein, W., Rahill, S.A., Emick, J.E., & Benn, A.E. (2005). Social competency
interventions for students with learning disabilities: A meta-analytic review.
Manuscript submitted for publication.
Winter, P., Kopriva, R., Chen, S., Emick, J. (2006). Exploring individual and item
factors that affect assessment validity for diverse learners: Results from a
large-scale cognitive lab. Learning and Individual Differences, 16, 267-276.
Emick, J.E. & Welsh, M. (2005). Association between formal operational thought and
executive function as measured by the Tower of Hanoi-Revised
Learning and Individual Differences, 15, 177-188.
Karlin, N.J., Emick, J., Mehls, E. & Murry, F.R. (2005). Comparison of Efficacy and
Ageism Levels Between Psychology and Nursing Students. Gerontology and
Geriatrics Education, 26(2), 81-96.
Conference Presentations
Johnstone, C., Allman, C., Altman, Emick, J., Hall, T., Ketterlin-Gellar, L., Thurlow,
M., & Tindal, G. (April, 2007). Universal Design of Assessment: Report from
a Researchers? Summit. Presented at the American Educational Research
Association national meeting, Chicago, IL.
Emick, J., & Kopriva, R. (April, 2007) The Validity of Large-Scale Assessment
Scores for ELLs Under Best Practices Testing Conditions: Does Validity Vary
by Language Proficiency? Presented at the American Educational Research
Association national meeting, Chicago, IL.
Emick, J. (April, 2006). Joint play and executive function: Secondary analysis of
longitudinal data. A poster presented at the 2006 American Educational
Research Association national meeting, San Francisco, CA.
Emick, J., Mann, H., Kopriva, R., & Ming-Yi, C. (April, 2006). Language Liaisons:
A novel accommodation for new ELLs. A paper discussion presented at the
2006 American Educational Research Association national meeting, San
Francisco, CA.
Emick, J., Kopriva, R., Chen Su, C., Mislevy, R., & Carr, T. (April, 2006). Achieving
accurate results for diverse learners: Access-enhanced item development. A
paper discussion presented at the 2006 American Educational Research
Association national meeting, San Francisco, CA.
Koran, J., Emick, J., & Kopriva, R. (April, 2006). Teacher evaluation of a
computerized accommodations recommendation system for English
Language Learners. A paper discussion presented at the 2006 American
Educational Research Association national meeting, San Francisco, CA.
Koran, J., Kopriva, R., Emick, J., Monroe, R., Walker Webb, P., & Garavaglia, D.
(April, 2006). A multiple measures approach for gathering information to
make individualized test accommodation decisions for English Language
Learners. A paper presentation presented at the 2006 National Council on
Measurement Education meeting, San Francisco, CA.
Welsh, M., Tlustos, S., Howell, Z., & Emick, J. (February, 2005). Formal operational
thought and executive functions in college freshmen and seniors. A poster
presented at the 2005 International Neuropsychological Society meeting, St.
Louis, MI.
Emick, J.E., Welsh, M., Howe, Z, & Chandran, S. (Feburary, 2004). Formal
operational reasoning as a predictor of performance on the Tower of Hanoi-
Revised. A poster presented at the 2004 International Neuropsychological
Society meeting, Baltimore, MD.
Karlin, N.J., Emick, J.E., & Mehls, E. E. (April, 2003). Ageism and nursing students.
A presentation at the Annual Conference of the Rocky Mountain
Psychological Association, Denver, CO.
O?Reilly-Wolf C., Welsh M., McNeil H., Emick J.E., & Gonzalez S. (February,
2003).Individual differences in self-efficacy and tower task performance. A
poster presented at the 2003 International Neuropsychological Society
meeting, Honolulu, HI.
O?Reilly-Wolf C., Welsh M., McNeil H., Emick J.E., & Gonzalez S. (May, 2002).
Individual differences in self-efficacy and tower task performance. A poster
presented at the 2002 Annual University of Northern Colorado Research
Symposium, Greeley, CO.
Emick, J.E. & Haaf, R.A. (May, 2001). The role of contour deletion in object
recognition by preschool children. A poster presented at the Annual
Conference of the Midwestern Psychological Association, Chicago, IL.
Related Professional Activities
Project Manger, Center for the Study of Assessment Validity and Evaluation,
2005-2006
Duties: Served as project manger for multi-year grant, Achieving Accurate Results
for Diverse Learners. Over-saw the creation of access-based large-scale test items for
ELLs and students with high frequency disabilities in conjunction with the South
Carolina Department of Education. Assisted in the development of the research
design for a comparability field test. Maintained responsibility for daily management
and allocation of resources.
Graduate Assistant, Center for the Study of Assessment Validity and
Evaluation, 2003-2005
Duties: Provided research, statistical, writing, and administrative assistance to multi-
year federal grants (Valid Assessment of English Language Learners, The Taxonomy
for English Language Learners, and Achieving Accurate Results for Diverse
Learners) geared to assessing the effectiveness of accommodation and
accommodation use for English Language Learners. Served as science content editor
in the development of access-enhanced large-scale achievement test items.
Graduate Research Assistant, Center for the Study of Assessment Validity and
Evaluation, 2004
Duties: Assisted Rebecca Kopriva, PhD in the writing and editing of Building
Validity for English Language Learners into Large Scale Achievement Tests
(Erlbaum Publishing).
Graduate Research Assistant, Philip Merril College of Journalism, 2003
Duties: Assisted the university dean with gathering, configuring, and reporting
enrollment data and programmed career events for journalism students.
Graduate Assistant, Assault Survivors Advocacy Program, 2001-2003
Duties: Provided victim advocacy/crisis services for students and significant others,
developed and evaluated prevention education programs, recruited, trained, and
supervised student advocates/interns, and assisted with grant writing and grant
administration.
Counselor, Residential Youth Treatment Center, 2002-2003
Duties: Maintained a safe and therapeutic environment through the principles of
behaviorism for kids aged 4-12 and provided therapeutic crisis intervention.
Counselor, National Depression Screening Day, 2002
Duties: Assessed scores, provided individual education, referrals, and support to
clients.
Counselor, National Alcohol and Drug Screening Day, 2002-2003
Duties: Assessed scores, provided individual education, referrals, and support to
clients.
Program Assistant, Women's Resource Center, 2000-2001
Duties: Facilitated domestic violence/self sufficiency support groups, provided
comprehensive multi-program evaluation, supervised student interns, and assisted in
planning national and campus events.
Resident Advisor, Resident Life Staff, 1998-2000
Duties: Provided programmatic, mediation, and resident support to 50+ freshman
women while maintaining a safe and positive living environment.
Counselor, Upward Bound, 2000
Duties: Coordinated summer learning initiative, served as a resource and conflict
intermediary, and specialized in educational diversity.
Professional Membership
American Psychological Association (2003-Current)
Maryland School Psychologists? Association (2004-2006)
American Educational Research Association (2005- Current)
Skills
Experience handling large sets of data.
Ability to effectively train/present to individuals and groups on a variety of topics.
Experience aligning access-based test items to educational agencies standards.
Excellent knowledge of Front page, Publisher, Excel, and Internet Research.
Training Experience/Invited Presentations
Invited Participant, National Center on Educational Outcome: Universal
Design Summit, 2006
Duties: Invited ELL expert for Universal Design Summit. The summit brought
together experts to share research related to Universal Design and develop common
definitions and approaches for dissemination.
Presenter, University of Maryland, 2005-2006
Duties: Co-presented Enhancing School Success to the University of Maryland
community with Dr. Beth Warner. The talk provided practical ways to improve a
child?s educational success and presented information on the special education and
assessment process in the state of Maryland.
Trainer, Riverside Publishing Company, 2004-2007
Duties: Consultant for Riverside Publishing training customers (ranging from 10 to
50 participants per 7 hour training session) regionally and nationally on the technical
adequacy and administration of the Battelle Developmental Inventory 2
nd
Edition.
Clients have included: Prince George County Public School District, Delaware
Department of Education, North Cambria School District, and Capital Area
Independent Unit.
Trainer, Valid Assessment of English Language Learners, 2004
Duties: Provided training for 300 undergraduate student employees on how to
administrator accommodations for large-scale standardized assessments in elementary
schools.
Guest lecturer, Women and Violence, 2002
Duties: Invited to lecture to 150 women?s health students on the topic of the cultural
context of women and violence.
Guest lecturer, Sexual Assault Prevention and Gender Education, 2001-2003
Duties: Developed educational presentations for campus wide use, lead classroom
discussion, and trained graduate and undergraduate volunteers to be educational
lecturers.
Teaching Assistant, Leadership U.T., 2000-2001
Duties: Co-taught third year leadership class, developed leadership theory-based
curriculum, evaluated assignments, and initiated highly successful service learning project.
Practica and Internship Experience
Internship, Clark County School District, Las Vegas, NV, 2006-2007 (Supervisor:
Dr. Joan Silverstein, Ph.D)
Duties: Acted as a school psychologist under the supervision of a licensed
psychologist in two elementary schools with high populations of English Language
Learners and one high school for the performing arts with exceptional mental health
needs. Served as a member of the Student Intervention Team and multidisciplinary
team and provided consultation for staff and families. Received exceptional training
in response to intervention, comprehensive mental health services including group
and individual therapy, and autism assessment. Training included individual and
group supervision, as well as didactic and professional development training.
Externship, University Parent Consultation & Child Evaluation Service/University
Counseling Center, 2005-2006 (Supervisor: Dr. Beth Warner, Ph.D)
Duties: Engaged in individual and group therapy as well as parent and teacher
consultation through observation and direct services, conducted psychological
assessments, supervision of graduate students on cognitive assessments, and co-
presented parent workshops (Enhancing School Success & Managing Externalizing
Behaviors).
Group Counseling, Prince George County Schools, 2005 (Supervisor: Dr. Lee
Rothman, Ph.D)
Duties: Co-facilitated counseling groups for elementary aged students using the
Structure/Themes/Open Communication/Reflection/Individuality/Experiential
Learning/Social Problem-Solving (STORIES) program.
Practica, Prince George County Schools, 2004-2005 (Supervisor: Dr. Lee
Rothman, Ph.D)
Duties: Completed comprehensive assessments including review of records,
classroom observations, structured and unstructured assessments for students with a
range of disabilities/diagnoses (including cognitive, emotional, and developmental),
assisted in the planning and coordination of interventions, and made
recommendations for IDEA classification and DSM diagnosis.
Practica, Psychological and Educational Evaluation and Research
Psychological Services, 2004-2005 (Supervisor: Dr. Hedwig Teglasi, Ph.D)
Duties: Completed comprehensive assessments for students referred to University School
Psychology Clinic, assisted in the planning and coordination of interventions, and
provided individual feedback for parents and if appropriate the client.
Practica, Howard County Schools, 2004-2005 (Supervisor: Mary Nelpa, Eds)
Duties: Provided instructional consultation for teachers around the instructional needs
of students and provided consultation for the administrative team around systematic
concerns about the current character education program. Completed curriculum-based
assessments and assisted in the development and measurement behavioral
assessments and interventions.
Provision of Supervision
Supervisor, University Parent Consultation & Child Evaluation
Service/University Counseling Center, 2005-2006
Duties: Provided individual supervision for 2 University of Maryland school
psychology practicum students. Oversaw the administration, scoring, interpretation,
writing, and feedback of cognitive and achievement assessments.
Reviewer
Reviewer, Current Issues in Education, 2005-2006
Research Experiences
2004-Current The development of executive functioning and joint play: Secondary
analysis of longitudinal data. William Strein, Ph.D (dissertation chair) University of
Maryland.
2004-2006 Achieving Accurate Results for Diverse Learner. Rebecca Kopriva. Ph.D.
(center director) University of Maryland.
2003-2006 Valid Assessment of English Language Learners. Rebecca Kopriva Ph.D
(center director) University of Maryland.
2003-2005 Meta Analysis. Social Skills Training. William Strein, Ph.D (advisor)
University of Maryland.
2003-2005 The Taxonomy of English Language Learners. Rebecca Kopriva Ph.D
(center director) University of Maryland.
2002-2003 Investigating formal operational reasoning as a predictor of performance
on the Tower of Hanoi-Revised. Marilyn Welsh Ph.D (advisor) University of
Northern Colorado.
2002-2003 Investigating nursing student?s attitudes towards ageism. Nancy Karlin
Ph.D and Elizabeth Mehls, FNP University of Northern Colorado.
2002 Individual differences in self-efficacy and tower task performance. Marilyn
Welsh Ph.D (advisor) University of Northern Colorado.
2001-2002 Validation of sexual attitudes and behaviors survey. Jennie Bruner Psy.D
(supervisor) University of Northern Colorado.
2001-2003 Development of sexual assault prevention program based on social norms
theory and previous research. Jennie Bruner Psy.D (supervisor) University of
Northern Colorado.
2002 Investigating male attitudes toward social norm messages. Teresa Wroe
(supervisor) Colorado Coalition Against Sexual Assault.
2000-2001 Investigating the effect of contour deletion on Biederman?s theory of
object perception by preschool children. Robert Haaf Ph.D (advisor) University of
Toledo.
2001 Investigating the effect of various therapies on children of substance abusers:
National Longitudinal Study. Christie Berger Ph.D (supervisor) University of Toledo.
2000-2001 Investigating influence of naming objects on infant?s ability to form
concepts of those objects. Anne Fulkerson Ph.D. (supervisor) University of Toledo.
2000-2001 Investigating the effect of parsing on object recognition of preschoolers.
Robert Haaf Ph.D (supervisor) University of Toledo.
Grants and Contracts
2006 Contract ($500) with Psychological Corporation for the national standardization
of the PALS-II.
2005-2006 Contract ($500) with the Riverside Publishing Company for the try-out of
the Cognitive Assessment System-Revised (CAS-R).
2004-2006 Contract ($700) with Riverside Publishing Company, per training, for
providing private trainings regionally and nationally on the Battelle Developmental
Inventory 2
nd
Edition (BDI-2).
2004-2006 Contract ($3000) with grant program Assisting Young Mothers for
assessments using the Battelle Developmental Inventory 2
nd
Edition to measure
participant?s growth and assess programmatic performance.
2005 Contract ($200) with The Psychological Corporation for the national
standardization of the Differential Abilities Scale 2
nd
Edition (DAS-II).
2004 Contract ($200) with The Psychological Corporation for the national
standardization of the Bayley Scales of Infant Development 3
rd
Edition (BSID-III)
2002-2003 Contract ($1500) with Riverside Publishing Company for national
standardization (validity and stability) and field test (palm software) of the Battelle
Developmental Inventory 2
nd
Edition (BDI-2).
2002 Contract ($2000) with Riverside Publishing Company for national
standardization (convergent validity and stability) of the Stanford-Binet Intelligence
Scale Fifth Edition.
Trainings Attended
Cognitive Assessment System- Revised, December, 2005, Itasca, IL
Training: 12 hour training on the administration, scoring and interpretation of the
Cognitive Assessment System- Revised.
American Psychological Association, June 2005, Chapel Hill, NC
Training: 40 hour Advanced Training Institute in Using Large-Scale Databases: The
NICHD's Study of Early Child Care sponsored by American Psychological
Association and Research Triangle Institute.
Battelle Developmental Inventory Train-the-Trainer Event, August 2004, Itasca,
IL
Training: 16 hour training on the administration of the Battelle Developmental
Inventory 2
nd
Edition and training on the presenting the assessment sponsored by
Riverside Publishing Company.
Ending Violence against Women Training, January 2003, Greeley, CO
Training: 8 hour training on campus dating violence sponsored by the Department of
Justice.
Battelle Developmental Inventory Training, December 2002, Little Rock, AK
Training: 12 hour training in administration of the Battelle Developmental Inventory
2
nd
Edition sponsored by Riverside Publishing Company.
Therapeutic Crisis Intervention Training, July 2002, Loveland, CO
Training: 30 hour training in therapeutic crisis intervention, behavioral techniques,
and childhood restraint sponsored by Namaqua Residential Treatment Center.
PASS Theory and the Cognitive Assessment System, May 2002, Greeley, CO
Training: 8 hour training in theory and administration of the CAS system sponsored
by Riverside Publishing Company.
Sexual Assault Prevention Training, May 2002, Montrose, CO
Training: 30 hour training in sexual assault prevention education and program
evaluation sponsored by the Department of Justice.
Stanford-Binet Intelligence Scale Fifth Edition, February 2002, Denver, CO
Training: 8 hour training in administration of the Stanford-Binet Intelligence Scale
Fifth Edition sponsored by Riverside Publishing Company.
Sexual Assault Advocate Training/Trainer, January 2002, Greeley, CO
Training: 30 hour training in sexual assault advocacy sponsored by University of
Northern Colorado. Certified as a confidential sexual assault advocate and trained
other advocate.