1 
 

 Connecting Spaces: Gender, Video Games and Computing in the Early Teens 
 

Jennifer Ashlock1, Miodrag Stojnic, and Zeynep Tufekci 
  
 
Abstract 
Informed by evidence that computing attitudes may be uniquely constructed in informal 
contexts and that the early teens are a key period for academic decision-making, we 
investigate lines of practice that connect computing skills, attitudes, and videogames. 
We compare the relationship between computer skill, computer efficacy, and activities 
associated with gaming using a data set of 3,868 children in middle school. The time 
that children spend gaming has very modest association with skill and efficacy. 
Accounting for the frequency with which children modify games, engage in social 
gaming activities, and the salience of gamer identity explains the gender gap in 
computer skill and significantly narrows the gender gap in computer efficacy. We find 
support for the argument that computer skill and efficacy are dependent on children 
connecting often isolated social contexts, a socially embedded characteristic of the 
digital divide. 
 

Keywords 

gender, video games, STEM education, computer science, digital technology 

 

 

 

 

 

 
1 Corresponding Author: 

Jennifer Ashlock, Department of Sociology, University of Maryland, College Park, MD  

ashlock [at] umd [dot] edu 

 

 

 



2 
 

 Computing has significant consequences for communities, industries and the 

state, but it has been challenging to bring different voices to the field. Often discussed in 

relationship to computer literacy (Gee 2003), video games have grown in popularity 

amongst girls and the general public (Lenhart et al. 2008, 2015; Rideout, Peebles, and 

Robb 2022; Rideout and Robb 2019) surpassing film box office returns in 2018 (Shieber 

2019), but the association with early computing paths is unclear. A recent survey of 

children in grades 7 through 12 shows a rather large gender difference: 25% of girls are 

interested in learning computer science (CS) compared to 50% of boys (Google and 

Gallup 2020). This is of particular concern since middle school is when children begin 

decision-making about college preparation (DeWitt, Archer, and Osborne 2013; Legewie 

and DiPrete 2014) and computing is a subject largely decoupled from U.S. K-12 

education (Puckett and Gravel 2020).  

Videogames could be associated with children’s skills and attitudes towards 

learning computing, but its role in early CS paths is rarely unpacked. In middle school, 

performances of gender are especially prized, perhaps more so than at any other time in 

the life course (Crouter et al. 2007; Williams 2006) and gaming may advantage boys (Fox 

and Tang 2017). Boys spend more time playing videogames than girls (Lenhart et al. 

2015; Rideout et al. 2022), but usage patterns may be less influential than the values 

and practices associated with more socially embedded aspects of gaming such as 

programming, information seeking and playing with friends. For children, these activities 

are often not connected into “lines of practice” that sustain interest or coalesce around 

tech-savvy identities (Azevedo 2011; Carter, Gibbs, and Harrop 2012; Consalvo 2017; 

Garcia 2017; Ito et al. 2013). Accounting for engagement in advanced gaming activities 

and identity in this key period may elucidate gender differentiated efficacy in 

computing, especially in the current educational environment where children usually 

learn digital technology outside of school. 

In the current paper we examine gaming activities, identity and their association 

with computer skill and efficacy. Using a data set of 3868 children in middle school we 



3 
 

assess gender differences in frequency of play, time playing on various devices, 

knowledge-based and social gaming activities, as well as gaming identity. We consider 

the different gender gaps across the activities, examine which forms of gaming are 

significantly associated with skill and confidence, and then evaluate gender separate 

models. While our study is limited in that it is cross-sectional, our results demonstrate 

how children learn outside of school and amongst peers in ways that sustain interest in 

a subject, thus shedding light on the socially embedded aspects of the digital divide 

(DiMaggio et al. 2004).  

LITERATURE REVIEW 

Participation in early computing paths may be influenced by experience with 

electronic devices and perceived ability to learn computing. Academic efficacy, belief in 

the capacity to learn a specific curriculum or field (Bandura et al. 2001), is associated 

with effort regulation and persistence towards learning activities (Richardson, Abraham, 

and Bond 2012). Computer efficacy in particular is correlated with elementary and 

middle school children’s experience with digital devices at home (Blanchard, Gardner-

McCune, and Anthony 2018) and programming class learning outcomes such as 

computational thinking (Ketenci et al. 2019). Skills and attitudes towards learning 

computing may be mutually reinforcing because efficacy is associated with risk-taking, 

experimentation and practices that are generative in computing paths (Gee 2003; 

Klimmt and Hartmann 2006). Middle school may be a pivotal time for computing 

interest because educational trajectories become more salient (Legewie and DiPrete 

2014). But, like science efficacy which appears to decline before the transition to high 

school (Lofgran, Smith, and Whiting 2015), computer efficacy also declines across the 

middle school grades, even as computer skill appears to increase (Ashlock, Stojnic, and 

Tufekci 2022).  

The earlier in life that children have experiences and develop practices with 

electronic devices, the more familiarity and confidence they may associate with them, 



4 
 

but there is evidence of gender differences. As early as elementary school, girls may 

spend less time playing with computer games and technological toys (Cherney and 

London 2006), but more recent evidence from the UK suggests that the gender gap in 

digital play is closing (Livingstone and Pothong 2021). Alternatively, culturally shaped 

beliefs about innate ability, the history of the field, and gender schemas may dampen 

girls’ interest in computer science more so than other STEM fields (Cheryan et al. 2017; 

Correll 2004; Huffman, Whetten, and Huffman 2013; Leslie et al. 2015; Master et al. 

2017; Meyer, Cimpian, and Leslie 2015; Moote et al. 2020). Taking account of culturally 

shaped beliefs and social structural processes, a body of scholarship shows that like-

minded peer groups, adult support and belongingness facilitate child development of 

science efficacy and “science-possible selves” (Gauthier et al. 2017; Hill et al. 2017; 

Robnett 2016; Robnett and Leaper 2013; Rogers et al. 2021). Much less is known about 

how children develop computer efficacy which is often navigated in the private sphere. 

Support in Early Computing Paths  

Gender differentiated computer interest and efficacy may derive in part from the 

fragmented institutional and cultural support for computer science education in the U.S. 

K-12 education system. Technology education has been outside the purview of the U.S. 

Department of Education with separate state boards of education and “vocation-

technical” schools. While parents, school administrators and teachers tend to agree that 

CS education is important (Google and Gallup 2020), the integration of technology 

throughout the curriculum is highly varied in part due to the emphasis that No Child Left 

Behind places on math and reading achievement over other subject matter (Gray et al. 

2010; Puckett and Gravel 2020; Rothstein, Jacobsen, and Wilder 2008). There is 

evidence of change, but about 40% of middle schools do not offer any CS courses 

(Google and Gallup 2016, 2020) and when available, they are often not required which 

may disadvantage girls. Many children do not have a good understanding of what 



5 
 

computing careers actually entail (Grover, Pea, and Cooper 2014), believing that 

computer scientists, for example, usually work in isolation (Hansen et al. 2017).  

In social interaction, adult support for children’s computing interest is also 

varied. Positive feedback from teachers and parent-led activities related to science and 

math are reliant on adult perception and discernment of child interest, which often 

advantages boys (Alexander, Johnson, and Kelley 2012; Archer et al. 2012; Lareau 2003; 

McKellar et al. 2019). Weak social support for teaching computing (Mazur and 

Woodland 2018) and cultural norms shape the framing of student digital skills in schools 

such that teachers have difficulty recognizing skills learned at home or do not see them 

as relevant (Paino and Renzulli 2013; Rafalow 2018). 

Especially in the early teens, peers may play an important role in the 

development of computer skill and efficacy. As suggested by Gee (2003) and others, 

some forms of routinized play can become “affinity spaces” where similarly interested 

participants set goals and learn together by sharing passions and practices. Children may 

develop interest and confidence when equal access is provided in these low-stakes 

experiences, thus countering stereotypes (Gee and Hayes 2010). Shared hobbies are 

beneficial in science learning if they provide a forum for discussing interests and 

intersections of identity across multiple social contexts (Azevedo 2011), but since digital 

technology learning often takes place outside of school both in physical space and 

online, computing skills and efficacy may be dependent on peers and connected affinity 

spaces (Livingstone and Sefton-Green 2016). 

 

Connecting Computer Skill, Efficacy and Gaming Activities into Lines of Practice 

While much focus has been placed on gaming as a potential solution to the 

varied support for computer learning in the U.S., simply playing is likely insufficient 

because such contexts may be isolated in social networks and unassociated with affinity 

spaces. Even though physical and online spaces are often conceptualized by parents and 



6 
 

others as mutually exclusive domains, this does not reflect reality. Rather, computer 

learning is an embedded social practice, a dimension of the digital divide which children 

may experience acutely (Livingstone and Sefton-Green 2016). The computer literacy and 

learning scholarship shows that children sustain their interest by actively connecting 

affinity spaces and advanced practices across multiple contexts such as school and at 

home, on gaming platforms and in the same physical space (Esteban-Guitart, Coll, and 

Penuel 2018; Ito et al. 2013, 2020). Analogous to Bourdieu’s emphasis on the interplay 

of agency and structure (1977), connected learning experiences may depend on children 

knowing and interacting with peers who share their interests and practices such as 

modifying games with code, seeking information about the latest gaming technologies 

and developing the skills necessary to set up a gaming session which allows for multiple 

players. Called “paratexts” by some, children engage in activities beyond the game itself 

such that conversations, friends and practices overlap and emerge as “lines of practice” 

(Azevedo 2011; Carter et al. 2012; Consalvo 2017). Absent multiple connections to 

affinity spaces and social networks, gaming may function more as entertainment than as 

learning in early academic paths (Ferguson et al. 2014; Hartanto, Toh, and Yang 2018; 

Vedechkina and Borgonovi 2021). Lines of practice may contain several components 

which we describe below. 

Knowledge-based gaming activities. 

 “Constructivist” practices like programming may build skill and confidence if 

connected to social contexts. Children enhance games through “mods”, a form of 

programming that allows for various customizations, different progressions through 

games, and other options that are associated with computer efficacy (Bush, Gilmore, 

and Miller 2020; Consalvo 2009; Gee and Hayes 2010). Use of mods, cheats and other 

forms of programming rely on research into new games and strategies (Kahila et al. 

2021; Lenhart et al. 2008; Williamson and Facer 2004). A recent study from Finland 

shows that 12 to 15 year old children engage in information seeking activities more 



7 
 

often than discussing games, and both are more common than modifying games (Kahila 

et al. 2021). 

Social gaming activities. 

Children who are particularly engaged with knowledge-based activities may 

select friends with whom they can discuss their enjoyment of gaming (De Grove 2014; 

Eklund and Roman 2017, 2019). Playing with friends in large online role-playing games 

may build skills and efficacy because they organize thousands of players for discussions 

of play, strategy, and programming (Deci and Ryan 2014; Steinkuehler and Williams 

2006) and the information available in such spaces can facilitate dialog about the 

gaming industry (Paaßen, Morgenroth, and Stratemeyer 2017; Williamson and Facer 

2004) which may be crucial for learning the different paths into computing (Grover et al. 

2014). Though children who play videogames together in the same physical space may 

find a more supportive environment less vulnerable to hostile gaming cultures, online 

play usually requires more complex technologies and thus the characteristics of in-

person play may constrain rather than expand experimentation and skill development 

(Klimmt and Hartmann 2006). 

Gaming identity. 

Now frequently referenced in popular culture, a salient gamer identity may stitch 

together goals, motivations and actions that are otherwise fragmented or isolated. 

Although definitions vary (Paaßen et al. 2017), gaming identity is associated with social 

activities as well as practices (Eklund and Roman 2017) thus connecting well-being, 

commitment and behavior in a feedback loop (Tajfel 2010). Identity is often embedded 

in one’s social structure (Stets and Burke 2000; Stryker and Burke 2000) and thus 

depends on multiple social supports.  

 

Gender Differences in Gaming Activities 

A popular assumption about videogames in American culture is that boys will be 

advantaged in what is perceived to be an undifferentiated masculine domain. If the 



8 
 

cultures around gaming closely parallel those of computing such that the traditional 

“geek” stereotype is associated with both, girls may be less interested (Cheryan et al. 

2009; Cheryan, Drury, and Vichayapai 2013) and this tends to be reflected in gender 

differences in gaming frequency (Lenhart et al. 2015; Rideout et al. 2022). As shown by 

the delineation of the associated advanced activities above, however, gaming is 

multifaceted and some activities and identities may be less risky than others. Boys and 

girls may consider the social costs to various gaming activities and invest accordingly 

(Grooten and Kowert 2015). 

For example, linguistic profiling of female players and hostile treatment appear 

common in popular first-person shooter games, particularly when cues of feminine 

identity are apparent (Gray 2014; Ivory et al. 2014; Kuznekoff and Rose 2013). Perhaps 

reflecting girls’ preference for in-person experiences that are less hostile, gender 

differences for teens’ play in person appear smaller than play with friends online 

(Lenhart et al. 2015). Stigma can also limit the acknowledgement of interest to others 

who may perceive the activity as incongruent to identity (Shaw 2012) and this appears 

to be supported in patterns of gamer identity. In a national sample of teens, Burch and 

Wiseman found that approximately 35% of girls answered “yes” when asked if they 

identify as gamers, compared to 69% of boys (2015). By comparison, gaming activities 

which do not require social interaction may have narrow gender differences. A survey of 

middle schoolers completed between 2009 and 2012 (N=5720) shows that 28% of boys 

and 18% of girls reported that they “create games on their computer” (Webb and Miller 

2015). According to a more recent national study (N=754), modifying games is a 

relatively rare activity, with only 3% of 8 to 12 year-olds reporting that they do this often 

(Rideout and Robb 2019).  

 

Exploring Gender Differences in the Benefits of Advanced Gaming and Identity 



9 
 

While bridging efficacy, practices and social contexts is challenging for many 

children, video games are popularly perceived as a masculine space where boys and 

men who play may have higher status (Klimmt and Hartmann 2006). Therefore, in 

comparison to boys, girls may get less of a return on gaming in terms of pay-off for 

efficacy or skills. Though “trash talk” during play is often normalized (Ortiz 2019), the 

harassment which occurs in these spaces may disadvantage girls (Fox and Tang 2017; 

Kuznekoff and Rose 2013; Richard and Gray 2018). Some girls continue to play despite 

these challenges, but assume a secondary status in gaming communities (Fox and Tang 

2014) thus gaining less skill and confidence from these experiences (Kaye and 

Pennington 2016; Richard 2017; Shaw 2012; Vermeulen et al. 2016; Vermeulen, Van 

Bauwel, and Van Looy 2017). Given the likely presence of stereotypes, as well as actual 

hostile behaviors or microaggressions in these spaces, girls may not experience the 

same boost to efficacy or skills from gaming activities as boys. 

While traditional computer science “geek” identity may be unattractive to girls 

especially looking to avoid its anti-social connotations (Cheryan et al. 2013), “gamers” 

are well known in popular culture. Experimental research with adult subjects in the UK 

shows that gamer identity may offer protection from stereotypes and other cultural 

beliefs that otherwise weaken performance and enjoyment (Kaye 2019; Kaye and 

Pennington 2016). In other words, evidence suggests that gaming identity might be even 

more important for women’s outcomes than men’s as it serves a buffering effect from 

stereotyping. 

 

The Current Project 

Despite interest in gaming as a generative path to CS, its role is largely unknown 

in the crucial middle school period, a time in which young people are likely to avoid 

activities that do not easily mesh with their evolving gender identity. Informed by the 

research above, this project evaluates how children develop lines of practice by bridging 



10 
 

contexts for learning computing as they pertain to gaming. Specifically, we examine a 

sample of middle school students and evaluate the relationship between skill, computer 

efficacy and the frequency with which children engage in the aforementioned gaming 

activities. In multivariate regressions, we control for household SES as previous 

scholarship has shown a relationship with social class and interest in computing (Charles 

and Bradley 2006; Yardi and Bruckman 2012), and for grade level as efficacy may decline 

in the early teens (Lofgran et al. 2015) 

First, in middle school there are likely to be gender differences in skill and 

computer efficacy, as found in other research. Bivariate analyses will assess this 

hypothesis. 

 
Hypothesis 1: On average, boys will report more skill and computer efficacy than 
girls. 
 

For children in their early teens, gaming may be a rare opportunity to develop 

computer skills, confidence and tech-savvy identities, but as described in the research 

above, boys and girls may differentially engage in gaming activities due their varied 

social risks.  

 
Hypothesis 2: Boys will report more hours of play on all devices with higher 
frequency than girls. Boys will also report higher frequency of knowledge-based 
activities, social activities and more salient gaming identity than girls. 

 
In light of the research outlined above, the multivariate regressions may show 

that accounting for lines of practice may reduce the gender difference in skill and 

computer efficacy, net of controls.  

 
Hypothesis 3(a): Net of controls, accounting for modes of play, frequency of play, 
knowledge-based activities, social activities and gaming identity will reduce the 
gender differences in skill. 
 



11 
 

Hypothesis 3(b): Net of controls, accounting for modes of play, frequency of 
play, knowledge-based activities, social activities and gaming identity will reduce 
the gender differences in computer efficacy. 
 

 
Some aspects of videogames have a masculine culture and this may lead to girls and 

boys having different experiences in affinity spaces. Evidence suggests that gaming 

identity, when salient, protects from stereotyping and hostile environments. 

 
Hypothesis 4(a): Net of controls, gaming activity and identity will predict skill 
differently for girls than boys.  
 
Hypothesis 4(b): Net of controls, gaming activity and identity will predict efficacy 
differently for girls than boys. 
 
 

METHODS 

Sample and Data Collection 

This study reports analyses of a sample of middle school students in the 

Southeastern United States. Our methodology has been published previously (Ashlock 

et al. 2022) so we provide an abridged summary here. After focus groups and survey 

piloting in early 2015, the survey was administered in classrooms in Fall 2015 through 

Spring of 2016, in a paper-and-pencil format, resulting in an initial sample of 5235 

students.  

Respondents came from a stratified sample of high school students in three 

school districts in the Southeastern United States. We used a non-random sampling 

technique described in Cohen (1992), statistical power analysis, which places the focus 

on obtaining a sample that is diverse and large enough to allow for sufficient statistical 

power in the analyses. The choice of schools was based on the proportion of children 

eligible for reduced lunch, as well as the proportion of black, Latin American and Asian 

students according to publicly available information. In total, permission was obtained 



12 
 

from fifteen schools, including 13 public schools (magnet and nonmagnet). The 

proportion of students surveyed in each school depended on the degree of access and 

resources of surveyors. The student response rate was high, in part due to IRB approval 

for an “opt out” form. We aimed for a near census of the schools, and according to 

publicly available information, we took an average sample of 83% of the school 

population. When we were unable to survey all students due to staffing constraints and 

other issues, a convenience sample was achieved, although a representative sample was 

made whenever possible. Thus, the proportion of white, black, Hispanic, and Asian 

students surveyed in each school is very similar to the public information of each 

school.1 

As mentioned above, the three counties surveyed are located in the 

Southeastern United States. The U.S. South has higher poverty rates than the national 

average, and there is some evidence that gender differences in math and English 

language arts scores in the Southern region reflect more traditional stereotypes about 

gender-appropriate careers (Pope and Sydnor 2010). Reardon et al. (2019) find that the 

gender difference in math and English language arts state accountability tests tend to 

reflect the Pope and Sydnor findings but there is considerable variation by school 

district. For the three counties surveyed the gender difference in state accountability 

tests for math is -.05 to .05 standard deviation units (with upper bounds for boys) and -

.35 to -.25 standard deviation units for English language arts (with upper bounds for 

girls) which indicates that they are representative of most counties in the United States 

(Pope and Sydnor 2010). The proportion of respondents with parents working in a STEM 

field is higher than the national average and thus may be more knowledgeable of STEM 

fields than other children in the United States.  

A consistent analytic sample was specified with information for all models. 

Listwise deletion reduced our sample by 26% (from 5235 to 3868) but did not 

 
1 Additional information is available upon request. 



13 
 

significantly change the mean SES, the gender differences across the models, nor the 

substantive results. Demographic characteristics of the sample are described in Table 1. 

Overall 36% of the sample is estimated to receive reduced cost lunch. Typically 

underrepresented in surveys of this kind, Black students comprise 16.4% of the data set 

(634 respondents), Hispanics 11.7% (454 respondents), Asian Americans 10.8% (419 

respondents) and multi-racial children 16% (621).  

(Table 1 about here) 

Measures 

Dependent variables.   

Our task-based skill measure (or “task-skill” for brevity) is a scale generated from 

a list of twelve tasks that students reported to have ever done on a computer (0 to 12). 

Respondents netted an additional point on the scale for each task completed. Self-

reported task familiarity constructs have good predictive power of actual skills 

completed in a lab environment (Hargittai 2005, 2008; Hargittai and Hsieh 2011; 

Hargittai and Kim 2010). Our task-skill index consists of a battery of self-reported 

literacy items in using software, hardware, and the internet and is similar to a more 

recent survey, the National Assessment of Educational Progress Computer Access and 

Familiarity Study (National Assessment of Educational Progress (NAEP) 2019:23). The 

Cronbach’s alpha for the tasks is .76. Since self-reported tasks do not necessarily 

indicate actual skill, readers should take caution when interpreting the results.  

Computer efficacy is a measure which evaluates perceived competence in a 

specific field of academic study and was developed by Bandura et al. (2001). Students in 

the survey were asked: “Please rate how well you feel you are able to do each of the 

things described below” with the numbers 1 through 10 listed for them to circle. “Learn 

computers” was included in the list of five items (see Figure 1). Our measure is future-

oriented and specific to academic learning of computers.  

(Figure 1 about here) 



14 
 

Variables of interest.   

There are two groups of gaming variables used in this paper: general and 

advanced. The general gaming measures, mode and frequency, are broad indicators of 

gaming activity. The mode variables are hours per week that the respondent plays on 

various devices. They derive from a battery of questions which ask the respondent to 

report the number of hours (on schooldays and weekend days) for different uses of 

technology (chatting, doing homework, surfing the web, watching videos, watching TV), 

including separate items for playing games on specific devices (cell phone/tablet (a 

combined question), console, computer, or something else). We use the cell 

phone/tablet, console and computer gaming measures only as these uses were 

overwhelmingly the most common in our sample and chatting, homework and surfing 

were not associated with computer skill or efficacy. (Additional information is available 

upon request.) 

The second general gaming measure is gaming frequency which is reported on a 

scale of 1 to 5 (“almost every day” to “I don’t play video games at all”). This measure 

was then reverse coded and converted to 0 to 4 to facilitate more intuitive 

interpretations of the regression coefficients. 

The advanced gaming measures include two knowledge-based activities, three 

social activities, and gamer identity. The two knowledge-based activities are (1) the 

frequency with which the respondent looks up game reviews, gaming strategies or 

“cheats” (a form of coding that gains access to additional content or customization), and 

(2) uses mods or scripts when playing video games. The social activities include talking 

to friends about video games; playing video games with friends in person, in the same 

room; or playing video games online with friends who are not in the same room. The 

scale for all of the above items ranges from “often”, “sometimes”, “rarely”, and “never” 

and are reverse coded such that higher values indicate greater frequency (1 to 4). 

Gaming identity is a single-item measure derived from a section of the survey in which 



15 
 

respondents were presented with a battery of six different statements, “I consider 

myself a _______” (gamer, fashionista, jock, geek, music buff, nerd). Respondents were 

asked to report how much they felt each statement was true on a scale of “Not at all 

true of me” and “Very true to of me” (1 to 4). Only the gamer measure is included in our 

analyses. 

Control variables.  

Gender is a dichotomous variable, in which self-identified girls were coded as 1 

and self-identified boys were coded as 0. 

Socio-economic status (SES) is a composite adapted from the measure of SES in 

PISA (Programme for International Student Assessment 2014:16), a well-known cross-

national comparison of 15-year-old student achievement. PISA combines parents’ 

highest education, highest occupation socio-economic index, and an index of 23 

household items intended to measure wealth, cultural possessions, and educational 

resources. The weights assigned to each component are empirical; they come from the 

principal component analysis. We emulate the PISA measure of SES, the only exception 

being that we use a different index in the place of home possessions. More detailed 

information about the SES measure can be found in Ashlock et al. (2022). 

Variables describing race and ethnic origins derive from an item on the survey, 

“What is your race or origin? Check all that apply” with the following categories: 

“White”, “Black or African American”, “Hispanic/Latino/Spanish”, “Asian (including 

India/Pakistan)”, “Native American” and “Other. Specify:”. Our measure is similar to the 

combined race and ethnicity question described in Krogstad and Cohn (2014). Only 

those respondents who placed themselves in the single race category were coded as 

such. The “Other” category was checked for answers consistent with multi-racial 

identity. Children who indicated “White” and also reported European American lineage 

in the “Other” section were coded as White. The “other” category consists of Native 

American and multiracial children. Readers thus need to be careful when interpreting 



16 
 

this category or making comparisons to it. The “race” categories in all regression 

analyses are dummy variables. The omitted category is “White”. 

The computer ownership measure is constructed from a question on the survey 

which asked respondents if they own a computer and if they have their own device, if 

they shared it with people in their family, or if they didn’t have the device. Answers 

were coded as “owning”, “sharing”, or “not having” and ambiguous answers coded as 

missing.  

Dummy variables for cell phone, tablet and game console ownership are also 

included in the models to distinguish students who “never” play games with students 

who own the devices but do not use them. The coefficients are not reported in the 

models as they are not statistically significant and a note is included in the table for 

readers. 

Math grades and science grades are self-reported. Higher values indicate As, 

lower values indicate Ds, etc. Self-reported grades are adequate measures of academic 

performance as students tend accurately respond to such questions in surveys. It should 

be noted that respondents may inflate their performance when such an item is 

sensitive, as in the case that they are lower performing (Rosen, Porter, and Rogers 

2017). Readers are advised to take these issues into account.  

Since other research finds declines in science efficacy across the middle school 

grades (Lofgran et al. 2015), dummy variables for grade level are also included, with 6th 

grade as the omitted category. 

Missing Values 

Stata applies listwise deletion to observations with at least one missing value in 

any variable. We identified a consistent analytic sample with information for all models. 

The following descriptive statistics, correlations and regression models include only 

those respondents who have non-missing data across all models. While this reduces our 



17 
 

sample size overall by about 26% (from 5235 to 3868), the substantive conclusions 

reached did not change. 

Analytical Strategy 

Our analysis takes place in two main parts. First bivariate analyses evaluate the 

extent of the gender differences across the dependent variables (hypothesis 1) and 

variables of interest (hypothesis 2) and then a correlation matrix is presented. Second, 

multivariate analyses using ordinary least square regressions evaluate gender 

differences in computer task-skill as well as computer efficacy, net of the general 

gaming measures (mode and frequency) and advanced gaming measures. Skill is 

evaluated first to test support for hypotheses 3(a) and 4(a) and then efficacy is tested to 

evaluate support for hypotheses 3(b) and 4(b). Gaming identity is included as a separate 

step in the models to evaluate if it is associated with the other measures or whether it 

has an independent effect. The task-skill measure is also added to the efficacy models to 

evaluate its effect independently from the other gaming activities, thus connecting 

“lines of practice”. In order to make comparisons across the two dependent variables in 

the multivariate analyses, we standardized them. The coefficients in tables 4 and 5 

should be interpreted as changes in standard deviation units as they are associated with 

the dependent variable, abbreviated as SD units. To account for possible correlated 

residuals amongst students in each of the 15 schools, the cluster() option was used in 

Stata. Post-hoc tests for evaluating the significance of the coefficients in the gender-

separate models include the ANOVA with Barlett test for equal variances and Lavene’s 

test. 

RESULTS 

Bivariate Analyses 

The bivariate results provide support for some of the predicted gender 

differences (table 2). While there are gender differences in both dependent variables 



18 
 

and this supports hypothesis 1, the gender difference in computer skill is smaller (0.21 in 

comparison to 0.55 SD units). 

Overall, children engage in a wide variety of video game play. As predicted in 

hypothesis 2, boys play more frequently and for more hours on all devices. The gender 

difference for time spent per week playing on computers is 0.34 SD units, about half the 

size of the difference for time playing gaming consoles (0.64 SD units).  

Turning to the gender differences in the advanced gaming activities, frequency 

of engagement is higher for boys overall, further supporting hypothesis 2. Gender 

differences in the knowledge-based activities are smaller than for the social activities, 

similar to previous research (Lenhart et al. 2015; Webb and Miller 2015). When asked if 

they consider themselves a gamer, the average for girls is closer to “2”, “mostly not true 

of me”, and for boys it is closer to “3”, “a little true of me”. This gender difference in 

identity amounts to one SD unit. The gender difference grows as commitment to the 

identity rises. About 40% percent of boys in the sample indicated that this was “very 

true of me” in comparison to 10% of girls, a significant difference which is larger than 

previously found (Burch and Wiseman 2015).  

 (Table 2 about here) 

Correlation Matrix 

The bivariate correlations offer an overview of the interconnected gaming 

activities (table 3). Task-based skill is correlated with gaming frequency at .1 and is 

significantly correlated with hours gaming on a computer, but not gaming on cell 

phones/tablets and consoles. While the knowledge-based activities of looking up 

reviews and modding games are similarly associated with skill and efficacy, most of the 

social activities and identity are more strongly correlated with computer efficacy. 

Gender is most strongly correlated with the social activities and gamer identity.  



19 
 

Gaming identity salience is strongly associated with all of the advanced gaming 

activities. The strongest correlates with gaming identity are talking to friends about 

games (.71) and playing online with friends (.61). The other advanced gaming activities 

are also strongly associated: looking up game reviews (.58), modding games (.53), and 

playing in person with friends (.49) and frequency of play (.47). 

Grade in school (6, 7 and 8) is associated with more computer skills, suggesting 

accumulating skill, but grade level is associated with significantly less computer efficacy. 

Self-reported science grades are negatively correlated with all of the gaming measures 

though they are not significant for frequency of play, talking about games, and playing in 

person. 

Multivariate Analyses 

Computer task-skill. Overall our measures explain 22 percent of the variance in 

computer task-skill, as shown in table 4. The gender difference in computer skill is -0.21 

SD units in the constrained model (see model 1). The addition of the general gaming 

measures (model 2) reduces the gender difference in computer skill by .09 SD units or 

43% such that the gender coefficient is no longer statistically significant. The hours per 

week played using a computer has a large role in reducing the gender gap. The addition 

of the knowledge-based activities and social activities in model 3 reduces the gender 

coefficient another .12 SD or 98%. This step in the analysis reveals that playing on a 

computer is associated with skill-enhancement through the knowledge-based activities. 

Playing games online with friends appears to be associated with computer gaming in 

particular. Gaming identity has a modest association with skill and does not explain 

additional variance in the dependent variable. When identity is added to the model, the 

coefficient for talking about games becomes significant and negative. Overall these 

results suggest that the gender gap in task-based skill is explained by the knowledge 

component of gaming. These findings generally support hypothesis 3(a). 



20 
 

Hypothesis 4(a) predicted that gaming activities would be experienced 

differently by boys and girls. Time played on a computer and modding are associated 

with more skill enhancement for boys than girls and supports hypothesis 4(a), congruent 

with stereotyping. Curiously, playing online has similar association with skill for boys and 

girls, however (comparing models 5 and 6). As the descriptive results show, boys more 

frequently engage in modding (table 2), and they may get more out of this activity as 

well, perhaps by engaging in more intensive programming. Gamer identity has a modest 

association with skill and similarly experienced by boys and girls. Hypothesis 4(a) is 

partially supported. 

 (Table 4 about here) 

Computer efficacy.  Overall the models explain 29 percent of the variance in 

computer efficacy. The gender difference in efficacy, -0.55 SD units, is over twice as 

large as the difference in the constrained task-based skill model (comparing model 1 in 

table 4 to model 1 in table 5). While including the general gaming measures reduces the 

gender difference in efficacy by 0.08 SD units or 15%, the addition of the knowledge-

based and social activities reduces the difference by another 32% to -0.31 SD units 

(comparing models 2 and 3 in table 5). Knowledge-based activities and talking to friends 

about games have significant associations.  

When gaming identity is added to model 4 it does not substantially change the 

gender co-efficient, but the coefficients for time played on a computer and talking to 

friends about games are no longer statistically significant. Gaming console time also 

shifts to a significant negative association. Overall, this suggests that children who 

identify with gaming play more hours on a computer, modify games and discuss games 

in ways that enhance their efficacy more than children who do not identify. Hours 

played on a console, when not associated with gaming identity, tend to be negatively 

associated with computer efficacy. The coefficient for identity (.16 in model 4) explains 

more variance than the knowledge-based activities, an effect which appears 



21 
 

independently associated with computer efficacy. Gaming identity remains a robust and 

independent component of efficacy when skill is added to the model. Overall these 

findings provide support for hypothesis 3(b). 

Comparing models 6 and 7 to evaluate hypothesis 4(b), there are no significant 

gender differences for the general gaming measures nor the gaming activities in the 

computer efficacy models. This suggests that gamer identity has a similar effect for girls 

and boys and thus does not support hypothesis 4(b).  

A separate analysis (not shown) does not find any significant gender differences 

prior to the addition of the skill measure in model 5, including for computer efficacy by 

grade level. The significant gender difference by grade level found in models 6 and 7 

suggests that while boys appear to experience a similar downward trend in efficacy, the 

gap across the grades appears more severe for girls, net of skill. 

A few overarching themes emerge in the results. Overall, the gender gap in 

computer efficacy is larger than the gap for skill. Skill appears largely associated with the 

knowledge-based activities as well as playing with friends online. Efficacy, on the other 

hand, has a direct socio-psychological component (gamer identity) which is independent 

of task-based skill. Efficacy also has an indirect knowledge-based component through an 

association with task-based skill. While popular, in-person play does not have a 

significant relationship with computer skill or efficacy in the models, though it is 

correlated with gaming frequency and talking about games with friends.  

Playing games on a computer appears to be associated with more frequent 

participation in gaming activities overall, perhaps because children who use computers 

to play games find that they facilitate more forms of play, experimentation and social 

interaction than other devices, thereby enhancing lines of practice. Even though many 

of the gaming activities were negatively associated with math and science grades in the 

correlation matrix, math and science grades are positively associated with computer skill 



22 
 

and efficacy net of the controls. This suggests that strong academic performance is 

positively associated with early computing paths. 

 (Table 5 about here) 

DISCUSSION  

Informed by scholarship on the fragmented support for computing in the U.S. 

and the importance of children developing lines of practice that connect efficacy, 

practices and social activities, we evaluated the relationship between different forms of 

gaming on skill and computer efficacy. As described above, we find support for 

hypotheses 1, 2, 3. Hypothesis 4(a) is partially supported and 4(b) is not supported. The 

gender difference in computer efficacy is much larger than the gender difference in 

computer skill which supports other work which points to more socio-psychological 

explanations of the gender gap in early computing paths, as opposed to skill or intrinsic 

qualities such as ability. We find gendered patterns in most of the gaming variables, and 

the contrasts are especially evident in social activities and gamer identity which may 

have greater social risks. We also find that the advanced gaming activities such as 

programming games, social activities and gamer identity explain the gender difference 

in computer skill and reduce the gender differences in computer efficacy by half. The 

gender stratified models on the whole show advanced gaming has similar association 

with skill and computer efficacy for boys and girls, including gaming identity. The results 

in total do not support the common claim that girls are initially less interested in gaming 

or that it has different meanings for them. Rather, the differences at this stage appear 

to be about engaging in practices and cultivating identities because generally when girls 

do this at rates similar to boys, they will develop the same skills and efficacy.  

An ongoing concern about the digital divide is that literacy includes knowledge-

based and social practices that take place in the private and public sphere. As 

Livingstone and others have noted, children may experience these disconnects in 

especially acute ways when computing is not integrated in early education. Thus, the 



23 
 

generative elements of digital gaming are not necessarily encapsulated by the net hours 

in which children play or frequency of play, but developing practices and identities that 

consistently build upon one another. For example, we find that knowledge-based 

activities such as modifying games are associated with skill development as found in 

other research (DiSalvo, Crowley, and Norwood 2008; Gee 2007), but that boys engage 

in this activity more frequently and in more skill-enhancing ways than girls, perhaps 

because they play on a computer more frequently. Boys also play online with friends 

more frequently where together they may find information about modifying games, but 

this activity may be socially risky for girls.  

The most potentially stigmatized dimension of gaming is perhaps the salience of 

gamer identity. Gaming identity appears to bundle frequent play, computer play, 

programming, skill, and online play together in such a way that they are efficacious and 

associated with skill. When combined, attitudes towards computing, identity and 

knowledge-based activities may build tacit knowledge, a kind of understanding that 

becomes “second nature” and requires less cognitive effort (Azevedo 2011; Consalvo 

2017; Eklund and Roman 2017; Garcia 2017; Lizardo 2017). But since girls may not have 

as much access to friends who validate this identity, they may find more difficulty 

engaging in social gaming and knowledge-based practices that sustain learning. Though 

playing video games is likely not the only generative activity in computing paths (Master 

et al. 2017; Nugent et al. 2019), the gender differences in advanced gaming practices 

and identity reflect the difficulty - especially for girls – to game in ways that sustain 

interest in computer science (Wong 2016) and may lead some to depart computing 

paths.  

Though limited, our results support the idea that interests are associated with 

both self-conceptualizations and practices. Our results suggest that gamer identity 

depends on connecting and engaging in frequently isolated contexts – modifying games, 

playing on a computer, online gaming, and talking with friends. Girls and boys receive 



24 
 

the same boost in computer efficacy when they have the same level of identity. A body 

of research shows that STEM identities in particular may have long term benefits for the 

pipeline, over and above that of field-specific confidence (Cech et al. 2011; Ceci and 

Williams 2010; Simpkins, Davis-Kean, and Eccles 2006; Stets et al. 2017). Prior research 

has shown that STEM self-expressions act as a kind of emotional “buy-in” in early paths 

(Charles 2017:11) and in the case of gaming this may often require negotiating gender 

expressions in the private sphere. Characterizations of gamers in popular culture 

personify gamers as masculine which may contribute to perceived entitlement in these 

spaces for children who identify. Regardless of gender, however, our results suggest 

that a critical aspect is engaging in the specific activities associated with this identity, as 

well as not avoiding this identity when one participates in such activities. 

There are alternative explanations for our findings. For example, since we do not 

account for participation in computer classes, after school activities, or out-of-school 

learning activities that incorporate game programming or social activities, we are not 

able to account for the extent to which social contexts are isolated for our respondents 

(Alexander et al. 2012). Girls may do more work to bridge contexts than boys due to 

emphasis on same-sex friendships at this age, for example. We do not account for play 

with parents and siblings and the bonding which takes place in families may be a 

generative factor (Eklund 2015; Harvey 2015). It is also difficult to determine how online 

and in-person gaming differs in early computing paths since children often play both 

online and in person. All else held equal, in-person gaming does not appear have a 

significant association with skill or efficacy in our results.  

When it comes to digital divides, advantages come in the form of access, various 

forms of device usage, as well as the information available in social networks (DiMaggio 

et al. 2004). The present research highlights the ways that identity may enhance or 

hinder lines of practice. Beyond the contribution of material resources and cognitive 



25 
 

skill, a child’s confidence in their ability to learn digital devices is a form of cultural 

capital that may have long term consequences for career paths.  

Limitations 

Our results should be evaluated with the knowledge that our methodology 

allows for a comparison of gender differences in gaming, but it does not offer the ability 

to evaluate the processes by which these relationships come to be. Our study is cross-

sectional and thus our results do not evaluate causality, only correlation. We also 

excluded survey items such as intended college major as they may have primed 

respondents in the survey and biased answers. Thus, our findings may only be relevant 

to this age group. Finally, due to our research site location, parent labor force 

participation in STEM fields is very likely above the U.S. average. While controls for 

parent occupation did not change the regression results, our respondents’ knowledge of 

computing may be higher than other children in the U.S. 

Future Work 

Our results inform future research on children’s computer learning. Research 

into additional gaming activities, social contexts, gaming identity and connections with 

learning would be well-served. For example, literacy practices in libraries and 

afterschool are increasingly popular as are gaming walkthroughs (such as those on 

Twitch.tv). These may add additional dimensions to connected learning practices (Ito et 

al. 2020). Stereotyping and boundary work in gaming spaces may discourage 

underrepresented groups from participation and additional aspects of identity could be 

incorporated.  

As others have noted, informal learning spaces such as advanced gaming fill a 

gap in the current K-12 curriculum but our results further support the movement to 

bring more computer learning into early education. Many educators, researchers and 

gamers address this issue, calling for interventions much earlier in the pipeline, the 

funding and development of educational games that facilitate community learning, and 



26 
 

computing to be further integrated into K-12 curriculum (Ito et al. 2020; Scholes, Mills, 

and Wallace 2021; Tekinbaş 2020).  

 

CONCLUSION  

Informed by evidence which emphasizes the importance of skills and computer 

efficacy in early computing paths, as well as the unique circumstances in which children 

learn digital technology, our analyses investigated the role of advanced gaming 

activities. Accounting for knowledge-based activities, social activities and gamer 

identity, we fully explain the gender difference in computer skill and half the variance in 

computer efficacy. We find that knowledge-based activities and gaming identity are 

likely a feature of early computing paths because they connect often socially distant 

activities into lines of practice, but that girls may be disadvantaged in the current social 

context of gaming. 

 

ACKNOWLEDGEMENTS 

We wish to thank the editor and the three anonymous reviewers whose comments and 

suggestions helped to improve and clarify the manuscript. 

 

REFERENCES 

 

Alexander, Joyce M., Kathy E. Johnson, and Ken Kelley. 2012. “Longitudinal Analysis of 
the Relations between Opportunities to Learn about Science and the Development of 
Interests Related to Science.” Science Education 96(5):763–86. 

Archer, Louise, Jennifer DeWitt, Jonathan Osborne, Justin Dillon, Beatrice Willis, and 
Billy Wong. 2012. “Science Aspirations, Capital, and Family Habitus: How Families Shape 
Children’s Engagement and Identification with Science.” American Educational Research 
Journal 49(5):881–908. 



27 
 

Ashlock, Jennifer, Miodrag Stojnic, and Zeynep Tufekci. 2022. “Gender Differences in 
Academic Efficacy across STEM Fields.” Sociological Perspectives 65(3):555–79. 

Azevedo, Flávio S. 2011. “Lines of Practice: A Practice-Centered Theory of Interest 
Relationships.” Cognition and Instruction 29(2):147–84. 

Bandura, Albert, Claudio Barbaranelli, Gian Vittorio Caprara, and Concetta Pastorelli. 
2001. “Self-Efficacy Beliefs as Shapers of Children’s Aspirations and Career Trajectories.” 
Child Development 72(1):187–206. 

Blanchard, Jeremiah J., Christina Gardner-McCune, and Lisa Anthony. 2018. “How 
Perceptions of Programming Differ in Children with and without Prior Experience.” P. 
1099 in Proceedings of the 49th ACM Technical Symposium on Computer Science 
Education, SIGCSE ’18. New York, NY, USA: Association for Computing Machinery. 

Bourdieu, Pierre. 1977. Outline of a Theory of Practice Cambridge University Press. New 
York, NY: Cambridge University Press. 

Burch, Ashly, and Rosalind Wiseman. 2015. “Curiosity, Courage and Camouflage: 
Revealing the Gaming Habits of Teen Girls.” in Vol. 2. 

Bush, Jeffrey B., Monica R. Gilmore, and Susan B. Miller. 2020. “Drag and Drop 
Programming Experiences and Equity: Analysis of a Large Scale Middle School Student 
Motivation Survey.” Pp. 664–70 in Proceedings of the 51st ACM Technical Symposium on 
Computer Science Education. New York, NY, USA: Association for Computing Machinery. 

Carter, Marcus, Martin Gibbs, and Mitchell Harrop. 2012. “Metagames, Paragames and 
Orthogames: A New Vocabulary.” Pp. 11–17 in Proceedings of the International 
Conference on the Foundations of Digital Games, FDG ’12. New York, NY, USA: 
Association for Computing Machinery. 

Cech, Erin, Brian Rubineau, Susan Silbey, and Caroll Seron. 2011. “Professional Role 
Confidence and Gendered Persistence in Engineering.” American Sociological Review 
76(5):641–66. 

Ceci, Stephen J., and Wendy M. Williams. 2010. “Sex Differences in Math-Intensive 
Fields.” Current Directions in Psychological Science 19(5):275–79. 

Charles, Maria. 2017. “Venus, Mars, and Math: Gender, Societal Affluence, and Eighth 
Graders’ Aspirations for STEM.” Socius 3:1–16. 

Charles, Maria, and Karen Bradley. 2006. “A Matter of Degrees: Female 
Underrepresentation in Computer Science Programs Cross-Nationally.” in Women and 
Information Technology. Cambridge, MA: MIT Press. 



28 
 

Cherney, Isabelle D., and Kamala London. 2006. “Gender-Linked Differences in the Toys, 
Television Shows, Computer Games, and Outdoor Activities of 5- to 13-Year-Old 
Children.” Sex Roles 54(9–10):717–26. 

Cheryan, Sapna, Benjamin J. Drury, and Marissa Vichayapai. 2013. “Enduring Influence 
of Stereotypical Computer Science Role Models on Women’s Academic Aspirations.” 
Psychology of Women Quarterly 37(1):72–79. 

Cheryan, Sapna, Victoria C. Plaut, Paul G. Davies, and Claude M. Steele. 2009. “Ambient 
Belonging: How Stereotypical Cues Impact Gender Participation in Computer Science.” 
Journal of Personality and Social Psychology 97(6):1045–60. 

Cheryan, Sapna, Sianna A. Ziegler, Amanda K. Montoya, and Lily Jiang. 2017. “Why Are 
Some STEM Fields More Gender Balanced than Others?” Psychological Bulletin 
143(1):1–35. 

Cohen, Jacob. 1992. “A Power Primer.” Psychological Bulletin 112(1):155–59. 

Consalvo, Mia. 2009. Cheating: Gaining Advantage in Videogames. Cambridge, MA: MIT 
Press. 

Consalvo, Mia. 2017. “When Paratexts Become Texts: De-Centering the Game-as-Text.” 
Critical Studies in Media Communication 34(2):177–83. 

Correll, Shelley J. 2004. “Constraints into Preferences: Gender, Status, and Emerging 
Career Aspirations.” American Sociological Review 69(1):93–113. 

Crouter, Ann C., Shawn D. Whiteman, Susan M. McHale, and D. Wayne Osgood. 2007. 
“Development of Gender Attitude Traditionality Across Middle Childhood and 
Adolescence.” Child Development 78(3):911–26. 

De Grove, Frederik. 2014. “Youth, Friendship, and Gaming: A Network Perspective.” 
CyberPsychology, Behavior & Social Networking 17(9):603–8. 

Deci, Edward L., and Richard M. Ryan. 2014. “Autonomy and Need Satisfaction in Close 
Relationships: Relationships Motivation Theory.” Pp. 53–73 in Human Motivation and 
Interpersonal Relationships: Theory, Research, and Applications, edited by N. Weinstein. 
Dordrecht: Springer Netherlands. 

DeWitt, Jennifer, Louise Archer, and Jonathan Osborne. 2013. “Nerdy, Brainy and 
Normal: Children’s and Parents’ Constructions of Those Who Are Highly Engaged with 
Science.” Research in Science Education 43(4):1455–76. 

DiMaggio, Paul, Eszter Hargittai, Coral Celeste, and Steven Shafer. 2004. “Digital 
Inequality: From Unequal Access to Differentiated Use.” Pp. 355–400 in Social 
Inequality. New York: Russell Sage Foundation. 



29 
 

DiSalvo, Betsy James, Kevin Crowley, and Roy Norwood. 2008. “Learning in Context: 
Digital Games and Young Black Men.” Games and Culture 3(2):131–41. 

Eklund, Lina. 2015. “Bridging the Online/Offline Divide: The Example of Digital Gaming.” 
Computers in Human Behavior 53:527–35. 

Eklund, Lina, and Sara Roman. 2017. “Do Adolescent Gamers Make Friends Offline? 
Identity and Friendship Formation in School.” Computers in Human Behavior 73:284–89. 

Eklund, Lina, and Sara Roman. 2019. “Digital Gaming and Young People’s Friendships: A 
Mixed Methods Study of Time Use and Gaming in School.” YOUNG 27(1):32–47. 

Esteban-Guitart, Moisès, César Coll, and William R. Penuel. 2018. “Learning across 
Settings and Time in the Digital Age.” Digital Education Review. 

Ferguson, Christopher J., Cheryl K. Olson, Lawrence A. Kutner, and Dorothy E. Warner. 
2014. “Violent Video Games, Catharsis Seeking, Bullying, and Delinquency: A 
Multivariate Analysis of Effects.” Crime and Delinquency 60(5):764–84. 

Fox, Jesse, and Wai Yen Tang. 2017. “Women’s Experiences with General and Sexual 
Harassment in Online Video Games: Rumination, Organizational Responsiveness, 
Withdrawal, and Coping Strategies.” New Media and Society 19(8):1290–1307. 

Garcia, Antero. 2017. “Space, Time, and Production: Games and the New Frontier of 
Digital Literacies.” Pp. 198–210 in Handbook of Writing, Literacies, and Education in 
Digital Cultures, edited by K. A. Mills, A. Stornaiuolo, A. Smith, and P. J. Zacher. New 
York: Routledge. 

Gauthier, G. Robin, Patricia Wonch Hill, Julia McQuillan, Amy N. Spiegel, and Judy 
Diamond. 2017. “The Potential Scientist’s Dilemma: How the Masculine Framing of 
Science Shapes Friendships and Science Job Aspirations.” Social Sciences 6(1):1–21. 

Gee, James Paul. 2003. What Video Games Have to Teach Us about Learning and 
Literacy. Palgrave Macmillan. 

Gee, James Paul. 2007. Good Video Games + Good Learning: Collected Essays on Video 
Games, Learning, and Literacy. New York, NY: Peter Lang. 

Gee, James Paul, and Elisabeth R. Hayes. 2010. Women and Gaming: The Sims and 21st 
Century Learning. New York, NY: Springer. 

Google and Gallup. 2016. Trends in the State of Computer Science in U.S. K-12 Schools. 
Washington, DC: Gallup. Retrieved September 1, 2022 
(https://services.google.com/fh/files/misc/trends-in-the-state-of-computer-science-
report.pdf). 



30 
 

Google and Gallup. 2020. Current Perspectives and Continuing Challenges in Computer 
Science Education in U.S. K-12 Schools. Washington, DC: Gallup. Retrieved September 1, 
2022 (https://services.google.com/fh/files/misc/computer-science-education-in-us-
k12schools-2020-report.pdf) 

Gray, Kishonna. 2014. Race, Gender, and Deviance in Xbox Live: Theoretical Perspectives 
from the Virtual Margins. New York: Routledge. 

Gray, Lucinda, Nina Thomas, Laurie Lewis, and Peter Tice. 2010. Teachers’ Use of 
Educational Technology in U.S. Public Schools: 2009. Washington, D.C.: National Center 
for Education Statistics, Institute for Educational Sciences, U.S. Department of 
Education. 

Grooten, Jan, and Rachel Kowert. 2015. “Going Beyond the Game:  Development of 
Gamer Identities Within Societal Discourse and Virtual Spaces.” Loading... The Journal of 
the Canadian Games Studies Association 9(14):70–87. 

Grover, Shuchi, Roy Pea, and Stephen Cooper. 2014. “Remedying Misperceptions of 
Computer Science Among Middle School Students.” Pp. 343–48 in Proceedings of the 
45th ACM Technical Symposium on Computer Science Education, SIGCSE ’14. New York, 
NY: ACM. 

Hansen, Alexandria K., Hilary A. Dwyer, Ashley Iveland, Mia Talesfore, Lacy Wright, 
Danielle B. Harlow, and Diana Franklin. 2017. “Assessing Children’s Understanding of the 
Work of Computer Scientists: The Draw-a-Computer-Scientist Test.” Pp. 279–84 in 
Proceedings of the 2017 ACM SIGCSE Technical Symposium on Computer Science 
Education, SIGCSE ’17. New York, NY, USA: Association for Computing Machinery. 

Hargittai, Eszter. 2005. “Survey Measures of Web-Oriented Digital Literacy.” Social 
Science Computer Review 23(3):371–79. 

Hargittai, Eszter. 2008. “An Update on Survey Measures of Web-Oriented Digital 
Literacy.” Social Science Computer Review 27(1):130–37. 

Hargittai, Eszter, and Yuli Patrick Hsieh. 2011. “Succinct Survey Measures of Web-Use 
Skills.” Social Science Computer Review 30(1):95–107. 

Hargittai, Eszter, and Su Jung Kim. 2010. The Prevalence of Smartphone Use among a 
Wired Group of Young Adults. Evanston, IL: Institute for Policy Research Northwestern 
University. 

Hartanto, Andree, Wei Xing Toh, and Hwajin Yang. 2018. “Context Counts: The Different 
Implications of Weekday and Weekend Video Gaming for Academic Performance in 
Mathematics, Reading, and Science.” Computers and Education 120:51–63. 



31 
 

Harvey, Alison. 2015. Gender, Age, and Digital Games in the Domestic Context. New 
York: Routledge. 

Hill, Patricia Wonch, Julia McQuillan, Eli Talbert, Amy Spiegel, G. Robin Gauthier, and 
Judy Diamond. 2017. “Science Possible Selves and the Desire to Be a Scientist: Mindsets, 
Gender Bias, and Confidence during Early Adolescence.” Social Sciences 6(2):36–53. 

Huffman, Ann Hergatt, Jason Whetten, and William H. Huffman. 2013. “Using 
Technology in Higher Education: The Influence of Gender Roles on Technology Self-
Efficacy.” Computers in Human Behavior 29(4):1779–86. 

Ito, Mizuko, Richard Arum, Dalton Conley, Kris Gutierrez, Ben Kirshner, Sonia 
Livingstone, Vera Michalchik, William Penuel, Kylie Peppler, Nichole Pinkard, Jean 
Rhodes, Tekinbas Katie Salen, Juliet Schor, Julian Sefton-Green, and S. Craig Watkins. 
2020. The Connected Learning Research Network: Reflections on a Decade of Engaged 
Scholarship. Irvine, CA: Connected Learning Alliance. 

Ito, Mizuko, Kris Gutiérrez, Sonia Livingstone, Bill Penuel, Jean Rhodes, Katie Salen, Juliet 
Schor, Julian Sefton-Green, and S. Craig Watkins. 2013. Connected Learning: An Agenda 
for Research and Design. Irvine, CA, USA: Digital Media and Learning Research Hub. 

Ivory, Adrienne Holz, Jesse Fox, T. Franklin Waddell, and James D. Ivory. 2014. “Sex Role 
Stereotyping Is Hard to Kill: A Field Experiment Measuring Social Responses to User 
Characteristics and Behavior in an Online Multiplayer First-Person Shooter Game.” 
Computers in Human Behavior 35:148–56. 

Kahila, Juho, Matti Tedre, Sanni Kahila, Henriikka Vartiainen, Teemu Valtonen, and Kati 
Mäkitalo. 2021. “Children’s Gaming Involves Much More than the Gaming Itself: A Study 
of the Metagame among 12- to 15-Year-Old Children.” Convergence 27(3):768–86. 

Kaye, Linda K. 2019. “The Role of Gamer Identity on Digital Gaming Outcomes.” Pp. 460–
65 in Design, User Experience, and Usability. Design Philosophy and Theory, edited by A. 
Marcus and W. Wang. Cham: Springer International Publishing. 

Kaye, Linda K., and Charlotte R. Pennington. 2016. “‘Girls Can’t Play’: The Effects of 
Stereotype Threat on Females’ Gaming Performance.” Computers in Human Behavior 
59:202–9. 

Ketenci, Tuba, Brendan Calandra, Lauren Margulieux, and Jonathan Cohen. 2019. “The 
Relationship Between Learner Characteristics and Student Outcomes in a Middle School 
Computing Course: An Exploratory Analysis Using Structural Equation Modeling.” 
Journal of Research on Technology in Education 51(1):63–76. 

Klimmt, Christoph, and Tilo Hartmann. 2006. “Effectance, Self-Efficacy, and the 
Motivation to Play Video Games.” Pp. 133–45 in Playing Video Games: Motives, 



32 
 

Responses, and Consequences, edited by P. Vorderer and J. Bryant. New York, NY: 
Routledge Taylor & Francis Group. 

Krogstad, Jens Manuel, and D’Vera Cohn. 2014. U.S. Census Looking at Big Changes in 
How It Asks about Race and Ethnicity. Washington, D.C.: Pew Research Center. 

Kuznekoff, Jeffrey H., and Lindsey M. Rose. 2013. “Communication in Multiplayer 
Gaming: Examining Player Responses to Gender Cues.” New Media and Society 
15(4):541–56. 

Lareau, Annette. 2003. Unequal Childhoods: Class, Race, and Family Life. Berkeley, CA: 
University of California Press. 

Legewie, Joscha, and Thomas A. DiPrete. 2014. “The High School Environment and the 
Gender Gap in Science and Engineering.” Sociology of Education 87(4):259–80. 

Lenhart, Amanda, Joseph Kahne, Ellen Middaugh, Alexandra R. MacGill, Chris Evans, and 
Jessica Vitak. 2008. Teens, Video Games, and Civics. Washington, D.C.: Pew Internet & 
American Life Project. 

Lenhart, Amanda, Aaron Smith, Monica Anderson, Maeve Duggan, and Andrew Perrin. 
2015. Teens, Technology and Friendships. Washington, DC: Pew Research Center. 

Leslie, Sarah-Jane, Andrei Cimpian, Meredith Meyer, and Edward Freeland. 2015. 
“Expectations of Brilliance Underlie Gender Distributions across Academic Disciplines.” 
Science 347(6219):262–65. 

Livingstone, Sonia, and Kruakae Pothong. 2021. Playful by Design: A Vision of Free Play 
in a Digital World. London: 5Rights Foundation. 

Livingstone, Sonia, and Julian Sefton-Green. 2016. The Class: Living and Learning in the 
Digital Age. New York: NYU Press. 

Lizardo, Omar. 2017. “Improving Cultural Analysis: Considering Personal Culture in Its 
Declarative and Nondeclarative Modes.” American Sociological Review 82(1):88–115. 

Lofgran, Brandi B., Leigh K. Smith, and Erin F. Whiting. 2015. “Science Self-Efficacy and 
School Transitions: Elementary School to Middle School, Middle School to High School.” 
School Science and Mathematics 115(7):366–76. 

Master, Allison, Sapna Cheryan, Adriana Moscatelli, and Andrew N. Meltzoff. 2017. 
“Programming Experience Promotes Higher STEM Motivation among First-Grade Girls.” 
Journal of Experimental Child Psychology 160:92–106. 

Mazur, Rebecca, and Rebecca H. Woodland. 2018. “A Fringe Topic in a Fragile Network: 
How Digital Literacy and Computer Science Instruction Is Supported (or Not) by Teacher 
Ties.” ACM Transactions on Computing Education 18(4):1–20. 



33 
 

McKellar, Sarah E., Aixa D. Marchand, Matthew A. Diemer, Oksana Malanchuk, and 
Jacquelynne S. Eccles. 2019. “Threats and Supports to Female Students’ Math Beliefs 
and Achievement.” Journal of Research on Adolescence 29(2):449–65. 

Meyer, Meredith, Andrei Cimpian, and Sarah-Jane Leslie. 2015. “Women Are 
Underrepresented in Fields Where Success Is Believed to Require Brilliance.” Frontiers in 
Psychology 6:235–235. 

Moote, Julie, Louise Archer, Jennifer DeWitt, and Emily MacLeod. 2020. “Science Capital 
or STEM Capital? Exploring Relationships between Science Capital and Technology, 
Engineering, and Maths Aspirations and Attitudes among Young People Aged 17/18.” 
Journal of Research in Science Teaching 57(8):1228–49. 

National Assessment of Educational Progress (NAEP). 2019. Computer Access and 
Familiarity Study, General Information, Grades 4, 8, & 12. National Center for 
Educational Statistics (NCES), Institute of Education Sciences, U.S. Department of 
Education. 

Nugent, Gwen, Bradley Barker, Houston Lester, Neal Grandgenett, and Dagen Valentine. 
2019. “Wearable Textiles to Support Student STEM Learning and Attitudes.” Journal of 
Science Education and Technology 28(5):470–79. 

Ortiz, Stephanie M. 2019. “‘You Can Say I Got Desensitized to It’: How Men of Color 
Cope with Everyday Racism in Online Gaming.” Sociological Perspectives 62(4):572–88. 

Paaßen, Benjamin, Thekla Morgenroth, and Michelle Stratemeyer. 2017. “What Is a True 
Gamer? The Male Gamer Stereotype and the Marginalization of Women in Video Game 
Culture.” Sex Roles 76(7):421–35. 

Paino, Maria, and Linda A. Renzulli. 2013. “Digital Dimension of Cultural Capital: The 
(In)Visible Advantages for Students Who Exhibit Computer Skills.” Sociology of Education 
86(2):124–38. 

Pope, Devin G., and Justin R. Sydnor. 2010. “Geographic Variation in the Gender 
Differences in Test Scores.” Journal of Economic Perspectives 24(2):95–108. 

Programme for International Student Assessment. 2014. PISA 2012 Technical Report. 
Organisation for Economic Co-operation and Development (OECD). 

Puckett, Cassidy, and Brian E. Gravel. 2020. “Institutional Ambiguity and De Facto 
Tracking in STEM.” Teachers College Record 122(8):1–38. 

Rafalow, Matthew H. 2018. “Disciplining Play: Digital Youth Culture as Capital at School.” 
American Journal of Sociology 123(5):1416–52. 



34 
 

Richard, Gabriela T. 2017. “Video Games, Gender, Diversity, and Learning as Cultural 
Practice: Implications for Equitable Learning and Computing Participation Through 
Games.” Educational Technology 57(2):36–43. 

Richard, Gabriela T., and Kishonna L. Gray. 2018. “Gendered Play, Racialized Reality: 
Black Cyberfeminism, Inclusive Communities of Practice, and the Intersections of 
Learning, Socialization, and Resilience in Online Gaming.” Frontiers: A Journal of Women 
Studies 39(1):112–48. 

Richardson, Michelle, Charles Abraham, and Rod Bond. 2012. “Psychological Correlates 
of University Students’ Academic Performance: A Systematic Review and Meta-
Analysis.” Psychological Bulletin 138(2):353–87. 

Rideout, Victoria J., A. Peebles, and Michael Robb. 2022. Common Sense Census: Media 
Use by Tweens and Teens, 2021. San Francisco: Common Sense Media. 

Rideout, Victoria J., and Michael Robb. 2019. Common Sense Census: Media Use by 
Tweens and Teens, 2019. San Francisco: Common Sense Media. 

Robnett, Rachael D. 2016. “Gender Bias in STEM Fields: Variation in Prevalence and 
Links to STEM Self-Concept.” Psychology of Women Quarterly 40(1):65–79. 

Robnett, Rachael D., and Campbell Leaper. 2013. “Friendship Groups, Personal 
Motivation, and Gender in Relation to High School Students’ STEM Career Interest.” 
Journal of Research on Adolescence 23(4):652–64. 

Rogers, Adam A., McKay Boyack, Rachel E. Cook, and Emma Allen. 2021. “School 
Connectedness and STEM Orientation in Adolescent Girls: The Role of Perceived Gender 
Discrimination and Implicit Gender-Science Stereotypes.” Sex Roles 1–17. 

Rosen, Jeffrey A., Stephen R. Porter, and Jim Rogers. 2017. “Understanding Student Self-
Reports of Academic Performance and Course-Taking Behavior.” AERA Open 3(2):1–14. 

Rothstein, Richard, Rebecca Jacobsen, and Tamara Wilder. 2008. Grading Education: 
Getting Accountability Right. New York, NY: Teachers College Press. 

Scholes, Laura, Kathy A. Mills, and Elizabeth Wallace. 2021. “Boys’ Gaming Identities and 
Opportunities for Learning.” Learning, Media and Technology 0(0):1–16. 

Shaw, Adrienne. 2012. “Do You Identify as a Gamer? Gender, Race, Sexuality, and 
Gamer Identity.” New Media & Society 14(1):28–44. 

Shieber, Jonathan. 2019. “Video Game Revenue Tops $43 Billion in 2018, an 18% Jump 
from 2017.” TechCrunch. Retrieved August 6, 2021 
(https://social.techcrunch.com/2019/01/22/video-game-revenue-tops-43-billion-in-
2018-an-18-jump-from-2017/). 



35 
 

Simpkins, Sandra D., Pamela E. Davis-Kean, and Jacquelynne S. Eccles. 2006. “Math and 
Science Motivation: A Longitudinal Examination of the Links between Choices and 
Beliefs.” Developmental Psychology 42(1):70–83. 

Steinkuehler, Constance A., and Dmitri Williams. 2006. “Where Everybody Knows Your 
(Screen) Name: Online Games as ‘Third Places.’” Journal of Computer-Mediated 
Communication 11(4):885–909. 

Stets, Jan E., Philip S. Brenner, Peter J. Burke, and Richard T. Serpe. 2017. “The Science 
Identity and Entering a Science Occupation.” Social Science Research 64:1–14. 

Stets, Jan E., and Peter J. Burke. 2000. “Identity Theory and Social Identity Theory.” 
Social Psychology Quarterly 63(3):224–37. 

Stryker, Sheldon, and Peter J. Burke. 2000. “The Past, Present, and Future of an Identity 
Theory.” Social Psychology Quarterly 63(4):284–97. 

Tajfel, Henri. 2010. Social Identity and Intergroup Relations. New York, NY: Cambridge 
University Press. 

Tekinbaş, Katie Salen. 2020. Raising Good Gamers: Envisioning an Agenda for Diversity, 
Inclusion, and Fair Play. Irvine, CA: Connected Learning Alliance. 

Vedechkina, Maria, and Francesca Borgonovi. 2021. “A Review of Evidence on the Role 
of Digital Technology in Shaping Attention and Cognitive Control in Children.” Frontiers 
in Psychology 12:487. 

Vermeulen, Lotte, Elena Núñez Castellar, Dirk Janssen, Licia Calvi, and Jan Van Looy. 
2016. “Playing under Threat. Examining Stereotype Threat in Female Game Players.” 
Computers in Human Behavior 57:377–87. 

Vermeulen, Lotte, Sofie Van Bauwel, and Jan Van Looy. 2017. “Tracing Female Gamer 
Identity. An Empirical Study into Gender and Stereotype Threat Perceptions.” 
Computers in Human Behavior 71:90–98. 

Webb, David C., and Susan B. Miller. 2015. “Gender Analysis of a Large Scale Survey of 
Middle Grades Students’ Conceptions of Computer Science Education.” Pp. 1–8 in 
Proceedings of the Third Conference on GenderIT. Philadelphia, PA: ACM. 

Williams, Bronwyn T. 2006. “Girl Power in a Digital World: Considering the Complexity of 
Gender, Literacy, and Technology.” Journal of Adolescent and Adult Literacy 50(4):300–
307. 

Williamson, Ben, and Keri Facer. 2004. “More Than ‘Just a Game’: The Implications for 
Schools of Children’s Computer Games Communities.” Education, Communication & 
Information 4(2/3):255–70. 



36 
 

Wong, Billy. 2016. “‘I’m Good, but Not That Good’: Digitally-Skilled Young People’s 
Identity in Computing.” Computer Science Education 26(4):299–317. 

Yardi, Sarita, and Amy Bruckman. 2012. “Income, Race, and Class: Exploring 
Socioeconomic Differences in Family Technology Use.” Pp. 3041–50 in Proceedings of 
the SIGCHI Conference on Human Factors in Computing Systems, CHI ’12. New York, NY, 
USA: Association for Computing Machinery. 

  



37 
 

 

TABLES (1-5) 

Table 1. Demographics of combined sample (N = 3868) 

    N Percent 
Gender   

 

female  1883 48.7 
male  1985 51.3 

   
 

Race/ethnicity   
 

White  1746 45.0 
Black/African American  634 16.4 

Hispanic/Latino  454 11.7 
Asian American  419 10.8 

Other  621 16.1 
   

 

Receive reduced cost lunch (est)   36.0 
   

 

Grade   
 

6th grade  1113 28.8 
7th grade  1401 36.2 
8th grade  1354 35.0 

    
Computer    

own  1912 49.4 
share with others in home  1748 45.2 

no home computer  208 5.4 
    

Cell phone (shared or own)  3305 85.4 
    

Tablet (shared or own)  3078 80.5 
    

Game console (shared or own)  3411 88.2 
 

 



38 
 

 

Table 2: Bivariate Analyses for Dependent Variables and Variables of Interest   
    All  Girls  Boys 

    N = 3868  N = 1985  N = 1883 
Variable   Range M (SD)   M (SD)   M (SD) 
Dependent variables           
 Task-based Computer skill 0-12 7.16 (2.81)  6.84 (2.61)  7.48 (2.97) * 
 Computer efficacy 1-10 6.75 (2.60)  6.04 (2.62)  7.50 (2.35) * 
             
Controls (abridged)           
 Math grades 0-4 3.40 (0.83)  3.40 (0.82)  3.40 (0.85)  
 Science grades 0-4 3.46 (0.81)  3.51 (0.76)  3.42 (0.86) * 
             
eneral gaming measures           
 Modes of play           
  Cell & tablet game hrs/wk (continuous) 6.29 (7.52)  5.75 (7.47)  6.86 (7.54) * 
  Console gaming hrs/wk (continuous) 6.06 (9.48)  3.12 (6.98)  9.16 (10.70) * 
  Computer gaming hrs/wk (continuous) 4.60 (6.70)  3.47 (5.83)  5.79 (7.31) * 
             
 Gaming frequency           
  Gaming frequency (never - every day) 0-4 3.00 (1.25)  2.63 (1.40)  3.40 (0.91) * 
             
Advanced gaming measures           
 Knowledge-based activities           
  Looks up game strategies, reviews & "cheats" 1-4 2.37 (1.08)  2.03 (1.03)  2.72 (1.01) * 
  Uses game mods 1-4 1.89 (1.07)  1.62 (0.94)  2.18 (1.11) * 
             
 Social activities           
  Talks about games w/friends 1-4 2.49 (1.17)  1.91 (1.04)  3.10 (0.98) * 
  Plays in person w/friends 1-4 2.43 (1.11)  2.04 (1.08)  2.84 (0.98) * 
  Plays online w/friends 1-4 2.35 (1.26)  1.79 (1.09)  2.93 (1.16) * 
             
 Gaming identity 1-4 2.36 (1.19)  1.81 (1.04)  2.94 (1.06) * 
                          
Note: SD = standard deviation           
*Statistically significant gender difference (p<.001)           

 

 

 

 

 



39 
 

 

 

Table 3. Correlation matrix - variables of interest (N = 3868) `

Variable 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
1. Female indicator 1
2. SES 0.00 1
3. Math grade (1-4) 0.00 0.37 * 1
4. Science grade (1-4) 0.06 * 0.39 * 0.49 * 1
5. Own computer (0-1) - 0.01 0.21 * 0.08 * 0.08 * 1
6. Grade in school 0.01 0.01 0.01 - 0.02 0.09 * 1
7. Cellgame hrs/wk - 0.07 * - 0.20 * - 0.20 * - 0.16 * - 0.04 - 0.10 * 1
8. Console hrs/wk - 0.32 * - 0.25 * - 0.24 * - 0.23 * - 0.01 0.02 0.38 * 1
9. Compgame hrs/wk - 0.17 * - 0.10 * - 0.15 * - 0.12 * 0.15 * - 0.07 * 0.36 * 0.28 * 1
10. Gamefreq (0-4) - 0.31 * - 0.05 - 0.08 * - 0.04 - 0.01 - 0.11 * 0.31 * 0.27 * 0.30 * 1
11. Lookup reviews (1-4) - 0.32 * - 0.08 * - 0.11 * - 0.09 * 0.06 * - 0.04 0.20 * 0.31 * 0.30 * 0.39 * 1
12. Mod games (1-4) - 0.26 * - 0.06 * - 0.10 * - 0.09 * 0.11 * - 0.06 * 0.17 * 0.26 * 0.37 * 0.31 * 0.56 * 1
13. Talk games (1-4) - 0.50 * - 0.04 - 0.05 * - 0.05 0.05 - 0.07 * 0.18 * 0.35 * 0.34 * 0.44 * 0.58 * 0.51 * 1
14. Play in person (1-4) - 0.36 * - 0.00 - 0.02 - 0.05 0.03 - 0.06 * 0.16 * 0.30 * 0.20 * 0.37 * 0.41 * 0.39 * 0.59 * 1
15. Play online (1-4) - 0.45 * - 0.07 * - 0.07 * - 0.08 * 0.10 * - 0.01 0.18 * 0.39 * 0.35 * 0.40 * 0.50 * 0.51 * 0.66 * 0.50 * 1
16. Gamer identity (1-4) - 0.47 * - 0.07 * - 0.09 * - 0.06 * 0.05 - 0.08 * 0.19 * 0.40 * 0.39 * 0.47 * 0.58 * 0.53 * 0.71 * 0.49 * 0.61 * 1
17. Task-based skill (0-12) - 0.11 * 0.18 * 0.12 * 0.13 * 0.23 * 0.21 * - 0.02 0.03 0.13 * 0.10 * 0.19 * 0.27 * 0.18 * 0.13 * 0.24 * 0.21 * 1
18. Comp efficacy (1-10) - 0.28 * 0.14 * 0.15 * 0.14 * 0.12 * - 0.04 0.01 0.06 * 0.15 * 0.17 * 0.26 * 0.27 * 0.31 * 0.20 * 0.27 * 0.35 * 0.41 *



40 
 

Table 4. Linear regression of task-based skill (std) by gaming measures
m1 m2 m3 m4 m5 m6

Girls Boys
b/se b/se b/se b/se b/se b/se

Controls
Girls -0.209 ** -0.119 -0.002 0.016

(0.06) (0.06) (0.06) (0.06)
SES 0.073 ** 0.087 ** 0.088 ** 0.089 ** 0.093 ** 0.099 *

(0.02) (0.02) (0.02) (0.02) (0.03) (0.04)
Black 0.021 0.011 0.002 0.006 0.069 -0.063

(0.07) (0.06) (0.05) (0.05) (0.06) (0.10)
Hispanic -0.057 -0.035 -0.063 -0.065 0.036 -0.144

(0.08) (0.07) (0.07) (0.07) (0.09) (0.09)
Asian 0.308 ** 0.333 *** 0.276 ** 0.269 ** 0.286 ** 0.240 *

(0.08) (0.08) (0.08) (0.08) (0.08) (0.10)
Other race 0.158 *** 0.158 *** 0.124 ** 0.123 ** 0.124 * 0.104

(0.04) (0.04) (0.04) (0.04) (0.05) (0.06)
Math grades (0-4) 0.031 0.05 * 0.058 ** 0.059 ** 0.072 ** 0.048

(0.02) (0.02) (0.02) (0.02) (0.02) (0.03)
Science grades (0-4) 0.084 * 0.088 * 0.092 ** 0.088 ** 0.056 0.111 **

(0.03) (0.03) (0.03) (0.03) (0.05) (0.03)
No home comp (omitted: own) -0.430 *** -0.346 *** -0.307 *** -0.315 *** -0.279 *** -0.340 ***

(0.06) (0.05) (0.04) (0.04) (0.06) (0.07)
Shared comp (omitted: own) -0.322 *** -0.282 *** -0.240 *** -0.242 *** -0.204 *** -0.269 ***

(0.04) (0.04) (0.03) (0.03) (0.03) (0.04)
Grade 7 (omitted: 6th grade) 0.229 *** 0.245 *** 0.246 *** 0.251 *** 0.252 *** 0.250 ***

(0.05) (0.05) (0.04) (0.04) (0.05) (0.03)
Grade 8 (omitted: 6th grade) 0.442 *** 0.484 *** 0.495 *** 0.501 *** 0.509 *** 0.486 ***

(0.03) (0.03) (0.03) (0.03) (0.03) (0.03)

General gaming measures
  Modes of play

Cell & tablet game hrs/wk (std) -0.035 -0.022 -0.017 0.005 -0.032
(0.02) (0.02) (0.02) (0.03) (0.02)

Console game hrs/wk (std) 0.015 -0.031 -0.040 0.025 -0.044
(0.02) (0.02) (0.02) (0.04) (0.03)

Comp game hrs/wk (std) 0.122 *** 0.046 * 0.037 * -0.045 0.068 * ꬸ
(0.02) (0.02) (0.02) (0.03) (0.02)

  Gaming frequency  (0-4) 0.065 ** 0.019 0.012 0.013 -0.001
(0.02) (0.01) (0.01) (0.01) (0.02)

Advanced gaming measures
  Knowledge-based activities

Looks up reviews (1-4) 0.035 * 0.024 0.058 ** 0.000
(0.02) (0.02) (0.02) (0.02)

Uses game mods (1-4) 0.172 *** 0.163 *** 0.077 ** 0.203 *** §

(0.02) (0.02) (0.02) (0.02)
  Social activities

Talks to friends re:gaming (1-4) -0.018 -0.042 * -0.002 -0.075 ***
(0.02) (0.02) (0.04) (0.03)

Plays in person w/friends (1-4) -0.009 -0.010 -0.004 -0.023
(0.02) (0.02) (0.02) (0.03)

Plays online w/friends (1-4) 0.101 *** 0.093 *** 0.104 *** 0.103 ***
(0.02) (0.02) (0.02) (0.02)

  Gamer identity  (0-4) 0.075 ** 0.061 * 0.095 *
(0.02) (0.03) (0.03)

Constant -0.591 ** -0.901 *** -1.388 *** -1.410 *** -1.481 *** -1.205 ***
(0.17) (0.16) (0.15) (0.15) (0.14) (0.20)

R-sqr 0.145 0.167 0.222 0.225 0.194 0.254
* p<0.05, ** p<0.01, *** p<0.001;  ꬸ Stat. sign. gender difference at p<0.01;  § Stat. sign. gender difference at p<0.0001
All models include controls for cell phone, tablet and game console ownership  



41 
 

Table 5. Linear regression of Computer efficacy (std) by gaming measures
m1 m2 m3 m4 m5 m6 	 m7

Girls Boys
b/se b/se b/se b/se b/se b/se b/se

Controls
Girls -0.546 *** -0.462 *** -0.314 *** -0.276 *** -0.281 ***

(0.04) (0.05) (0.04) (0.04) (0.04)
SES 0.049 * 0.062 * 0.064 ** 0.066 ** 0.036 0.057 ** 0.011

(0.02) (0.02) (0.02) (0.02) (0.02) (0.02) (0.03)
Black 0.004 -0.003 -0.011 -0.002 -0.004 0.022 -0.048

(0.05) (0.05) (0.05) (0.04) (0.03) (0.06) (0.07)
Hispanic 0.035 0.056 0.045 0.041 0.063 0.097 0.009

(0.05) (0.05) (0.04) (0.04) (0.03) (0.07) (0.09)
Asian 0.132 0.155 0.091 0.076 -0.014 0.045 -0.086

(0.08) (0.08) (0.08) (0.08) (0.05) (0.07) (0.05)
Other race 0.027 0.031 0.000 -0.003 -0.043 -0.04 -0.051

(0.06) (0.06) (0.06) (0.05) (0.05) (0.06) (0.05)
Math grades (0-4) 0.086 * 0.103 ** 0.111 ** 0.113 ** 0.094 ** 0.135 *** 0.055

(0.03) (0.03) (0.03) (0.03) (0.03) (0.03) (0.04)
Science grades (0-4) 0.108 ** 0.111 ** 0.112 ** 0.104 ** 0.075 * 0.044 0.097 *

(0.03) (0.03) (0.03) (0.03) (0.03) (0.04) (0.04)
No home comp (omitted: own) -0.306 ** -0.227 * -0.195 -0.212 -0.107 0.040 -0.247 **

(0.10) (0.10) (0.11) (0.11) (0.11) (0.16) (0.15)
Shared computer (omitted: own) -0.177 *** -0.140 *** -0.109 ** -0.115 ** -0.034 -0.018 -0.061

(0.03) (0.03) (0.03) (0.03) (0.03) (0.04) (0.04)
Grade 7 (omitted: 6th grade) -0.124 * -0.110 * -0.106 * -0.094 * -0.178 *** -0.234 *** -0.117

(0.05) (0.05) (0.04) (0.04) (0.03) (0.05) (0.06)
Grade 8 (omitted: 6th grade) -0.122 ** -0.083 * -0.064 -0.051 -0.219 *** -0.275 *** -0.154 ** ª

(0.04) (0.03) (0.04) (0.04) (0.03) (0.05) (0.04)

General gaming measures
  Modes of play

Cell & tablet game hrs/wk (std) -0.036 -0.025 -0.015 -0.009 -0.044 0.025
(0.02) (0.02) (0.02) (0.02) (0.03) (0.02)

Console game hrs/wk (std) 0.012 -0.033 -0.054 ** -0.040 * 0.001 -0.067 *
(0.02) (0.02) (0.02) (0.01) (0.04) (0.03)

Comp game hrs/wk (std) 0.115 *** 0.045 * 0.024 0.012 -0.012 0.023
(0.02) (0.02) (0.02) (0.02) (0.04) (0.03)

  Gaming frequency  (0-4) 0.063 ** 0.008 -0.008 -0.012 0.008 -0.047 *
(0.02) (0.02) (0.01) (0.01) (0.02) (0.02)

Advanced gaming measures
  Knowledge-based activities

Looks up reviews (1-4) 0.078 *** 0.053 *** 0.045 ** 0.062 * 0.035
(0.01) (0.01) (0.01) (0.02) (0.02)

Uses game mods (1-4) 0.108 *** 0.088 *** 0.033 0.015 0.044 *
(0.01) (0.01) (0.02) (0.03) (0.02)

  Social activities
Talks to friends re:gaming (1-4) 0.082 ** 0.028 0.042 0.021 0.065 *

(0.03) (0.03) (0.02) (0.03) (0.02)
Plays in person w/friends (1-4) -0.013 -0.016 -0.013 -0.033 0.001

(0.01) (0.01) (0.01) (0.02) (0.01)
Plays online w/friends (1-4) 0.038 0.020 -0.011 -0.007 -0.01

(0.02) (0.02) (0.02) (0.04) (0.02)
  Gamer identity  (0-4) 0.165 *** 0.140 *** 0.161 *** 0.128 ***

(0.02) (0.01) (0.02) (0.02)
Computer skill

Task-based skill (std) 0.334 *** 0.367 *** 0.302 ***
(0.02) (0.03) (0.02)

Constant -0.323 -0.619 *** -1.157 *** -1.206 *** -0.735 *** -1.020 *** -0.660 *
(0.16) (0.15) (0.14) (0.15) (0.13) (0.14) (0.26)

R-sqr 0.130 0.149 0.194 0.209 0.292 0.212 0.273
* p<0.05, ** p<0.01, *** p<0.001 ª Statistically significant gender difference at p<.05   
All models include controls for cell phone, tablet and game console ownership



42 
 

FIGURE 
 
 

Figure 1. Measures of Academic Efficacy in Survey