AN INVESTIGATION INTO Tl-IE IMPACT OF SEX ROLE IDENTITY ON THE EFFECTIVE PERFORMANCE OF SALESWOMEN by Lucette Bowers Comer Dissertation submitted to the faculty of the Graduate School of the University of Maryland in partial fulfillment of the requirements for the degree of Doctor of Philosophy 1988 Advisory Conmrittee: Professor Marvin A. Jolson Associate Professor William G. Nickels Associate Professor Ruth B. Smith Associate Professor M. Susan Taylor Associate Professor Mary M. Leonard '\{'\ \ I \, \ I L ... I . \'{ \ ) ;' ( r I \ \ ' ? A.BSTRACT Title of Dissertation: An Investigation into the Impact of Sex Role Identity on the Performanc e Effectiveness of Saleswomen Lucette Bowers Comer, Doc tor of Philosophy, 1988 Disser t a tion directed by: Marvin A. Jolson Professor Department of Marketing Because of a n increasing s hortage of qualified salespersonnel, recruiter s for sales position s are very receptive to female applicants. Despite this, sex-discrimi nation is sti.11 detect :ible in the ma rket place. Some sales ma nage r s resist bringing women into male-oriented areas of selling, criticizing thern for ;-il leged weaknesses. Their c riticisms stem from the belief that saleswomen will behave acco rd i ng to ge nd e r stereotypes on the job and th:1t this behavior will impact negatively on sell ing per(ormance. Sales managers need assurance that the saleswomen they hire will perform well on the job. This resea r c h investigated the usefulness of the concept o( II sex role identity" ns [\ basis for segrnenting the pool of Eem;JlP. applicants by their potential for effec tive performance. ,\ s ur vey was conducted of sales managers in three traditionally male areas of selling, who described saleswomen under their supervision. The relationships between sales managers' perceptions of gender stereotypic behavior, selling effectiveness, and sex role identity were examined. Saleswomen' s gender stereotypic behavior was defined as perceived weaknesses in three areas: "selling ability," "human relations," and "motivation ." Sellin g effectiveness was measured as perceived proficiency in performance of six functions of selling and non-selling activities. Saleswomen were classified into sex role types on the basis of their sales managers' perceptions of their masculine ("instrumental") and feminine ("expressive") traits in their sex role identities on the Bern Sex Role Inventory. Seven re sea re h hypotheses were tested usin g univariate and multivariate a nalysis of variance and co rrelationnl analyses. The result s showed that sales mann gers perceived some gender stereotypic behavior in the mark e tplace and that some o[ this behavior was associ? ated w;th reduce '! se 11 i? L L n g e ff ec t 1? _veness. Sex role types of saleswomen related to both pe r ceived gender stereotypic behavior and selli ng effectiveness. Androgynous n nd masculine saleswomen were perceived '?'S both th e 1. u east stereotype ct and the most effective performers. The findings g ive partial support for a two-dimensional mod 1 of selling effectiveness -~ -- -------- defined by masculin e "instrumentality" and f eminine "expressiveness." The result s have implications for the selection, training , and supervision of saleswome n. This dissertation is dedicated to my mother: Ella Smith Bowers, to the memory of my father: Ray Landis Bowers, and to the memory of my husband: John Preston Comer, Jr. l l ~-- - -- ------ ------ ACKNOWLEDGEMENTS I should like to extend my appreciation to the mem bers of my committee: to Marvin A. Jolson, my advisor, who first introduced me to the "wonderful world of selling" and whose rigorous standards of excellence kept me striving to attain my highest level of potential throughout; to William G. Nichols, whose commitment to a balance of "academic" and ."social" considerations helped me transve rse the "cultural shock" associated with being an older women in a "traditionally male" business college and whose unfa iling enthusiasm, clear thinking and support were always available; to Ruth B. Smith, who contributed her special technical expe rtise a nd who was always ready with advice and support; to M. Susan Tayl or, who contributed her special expertise in the area of organizational behavior and whose clear thinking was particularly valuable; and to Mary M. Leonard, who introduced me to the "woman's" point of view and helped me to achieve the critical, hut delicate, balance between "sexist" and "feminist" perspectives. I would like to extend my appreciation to my doc t ora l student colleagues. In particular, Darlene Smith, made some valuable contributions. I would also like to thank my fami.ly and fri e nd s : my friend \vendy fleit, who through discussions at innumerable ijj breakfasts contributed a good "ear" nnd provided some excellent feedback a nd some useful suggestions. Others that I should mention include my brother Ray Bowers and his wife Helen, my friends Martha Beard, Toni Cooper, Dan Glick, and my cousin Mary B. Whitney (who did not live t o see the project come to fruition, but who was gr ea tly interested in i ts outcome). I would like to extend my appreciation to Barbara Pletcher and the menber s of the National Association for Professional Saleswomen (NAPS), who were of gr eat help in identifying respondents for the study and who contributed insights into the lives of professional sale s women. Special thanks go to the many sales managers, too numerous to name, who ga ve their time to complete the lengthy questionnaire a nd without whose cooperation the researc h would not have bee n possi ble . I would also like t o exp r ess my ap preciation to the !Jnive r sity 0f Maryland. To the College of Business and Ma nagement that prov iJed a dmini s trative s upport, a nd to the Compu ter Science Cent e r that pro vided the compute r facilities. Special thanks go to Margarite IIall and Leslie Co lema n who s uggested potenti.al respond e nts , to Jeanne Fine ran and Linda Pete r s a nd all of the staff of the FaculLy Services offi ce , who performed the important f unct i.on of safely receiving the compl eted question naires , and to Mary Cin dri c of the Doctoral S tud ies Office who provi. i. n ') I positions of authority o n th e j o b ( Ho ffma n L1)?S) ? ._ -1 Thi s s t e r eotypr 55 stems from the notion tha t wome n would rathe r do r o ut i ne wo r k (e.g., clerical) in return for a good de al o f pra i se f rom a ma l e authority figure, than do more demanding wo rk for bette r mone t a r y rewards. This manner of thinking has be en used as a jus tif ica tion for assigning women to low-paying, routine work (e.g. , Bergma nn 1986). The existence of stereotypic thinking of thi s typ e on salesforces is attested to by the state ment in the tra ining materials of the 0rtho Pharmaceutical Company that "women like praise, they respond to it" (Kurtz 1986, p . A.3). Perceived Dependency: Avoiding Unplens an t Ac ti v l ties . On e aspect of this stereotype that women over- value affiliative reward s from their work is that they may tend to place dl s proportionc1t e importance on their relationships with customers. Young and Mond y 0982) suggest that salespersons, who work in the field, with o u t frequent contacts with supervisors or peers may be g in to place disproportionate importance on their relation s hips with c us t omer s , may have approval needs that cire not satisfied by supe rvi s ors o r peers, and may turn to customers to s a tisfy those ne ed s . As a result, they become vulnerable to the withdrawal o f appr oval by customers. Blake and Mout o n (1980) s uggest that s uc h s ales pe r so ns avoid situations which involve pers ona l unpleasa ntness , s uch as calling on dissatisfied cus tomers . They mi gltt r e a rt t o interruptions durin g s ales interv i e ws a nd t o o bjec ti o ns r aised by 56 customers as personal a ttacks. Such salespersons might experience excessive stress whe n confronted with conflicting demands from customers that they are unable to satisfy . The perception of relatively high role conflict a mon g sa l eswome n reported by Futrell (1980) may, in part, reflect such problems. Summary This chapter has discussed criticisms of saleswomen by sales managers in the context of gender stereotypes. It h:::is suggested the manner in which conformity to the s t e r eoty pe has the potential to impact nega tively upon the per forma nce of saleswomen. The proposed researc h will investiga tc whether behavior that is perceived by sales managers t o be in accorda nce wi.th gender stereotypes actually impacts negatively upo n the sel Ling effectiveness of saleswomen. The next cha pter extends the literature review into the areas of selling effec tiveness anrl sex role identity. The model that serves as the b sis for the res,~arc h is developed. 57 Endnotes I. The sex distribution of the samples in these studies reflects the sex composi tion of the jobs and illustrates the potential for bias. The sample obtained by Swan, Futrell and Todd ( 1978), from a male-oriented sellino area consisted of 396 males and 35 females. 0 The sample obtained by Skinner, Dubinsky and Cronin (1984) from n female-oriented area consisted of 34 males and 123 females. 2. A two-dimensional model of selling effectiveness co nsisting of instrumental and expressive dimensions is introduced (Blake and Mouton 1980; Buzzotta, Lefton and Sherberg 1982 ; Jolson 1984). According to this model, salespersons who beha ve in acco rdan ce with masculine stereotypes would be more likely t o excel at tnsks associated with the instrumental dimension of sellin g , whil e those \vho behave in accordance with feminine stereotypes would he more likely to excel at tasks associated with the expressive dimension. Salespersons 1vho excel at tasks associ :1ted wi th both i.nstrumentnl and expressive dimension s are co nsidered the r:iost effocti.vc . Salespersons ivho excel at tasks associnted wi.th o ne d i.lllen.s i.o n at the expense of the other, are less effective . Thus, the most effe c t?iv c salespe rsons are defined as androgynous i.n their apprnn c li 58 to selling. 3. Busch and Bush (1978) speculated that women, as a group, might have a greater need for clarity than do men. Alternatively, saleswomen may not be receiving sufficient feedback about their performance to become secure on their jobs, since sales managers seemed to be reluctant to discuss the negative aspects of their performance with them (Futrell 1980) . It is a lso possible that the female respond en ts were simply more honest about expressing their insecurities about the job than were the men (e.g., Spiegler and Liebert 1970; Wilson 1967), and that the responses reflected the general tendency of females to und e r-value their own competence (e . g ., Deaux 1984). 4. The "kernel of truth": behind women's perceived inferiority in the ability to learn technical mater i al received some support i.n the analysis of Maccoby and Jacklin (1974) who found women to be weaker than men in mathematical and spatial abilities , although the differences they isolated were very small. Subsequent meta-analysis has revealed that differences in mathematical ability and in spatial ability accou nt ed for o nl y 1% and 4.5% respectively of the variance betwee n the sexes (llyde 1981). While researcher s have continued to claim superior mathematical ability for males (Benhoe and Stanley 1980; 1983), c ritics have disco unted the result s (see Fox, Tobin and Brody 1979, for a review of these 59 ---=-::=- -=-::.--- ---- arguments). The important contributions that have been made to the field of mathematics by women historically (Osen 1984), weakens the argument for the innate superiority of males, a lthough recent research into the sex typing of another technica l field, that of computer use, argues for a male dominant stereotype in the field (Lockheed 1985 ; Fetler 1985) . 5. Females might appear t o have difficulty responding quickly to situations because their traditional socialization does not provide sufficient opportunities for them to develop the necessary Skills. Block (1984) sugges ts that the traditional socialization of females causes them to live in a more s tructured world in which exploratory behavior is generally discouraged. Boys live in a "l ess predictable world" where "ad hoc strategies are frequently demanded." Girls do not have as many opportunities to encounter situations in which they can develop and prac tice independ ent and improvisational behaviors and are not rewarded f or s uch behaviors. Differences have been identified in the types of ga mes played by boys and girls (e.g., Lever 1976 ; 1978 ) and the degree to which the sexes differentially engage in rough and tum ble plny (Di Pietro 1981). 6 . It should be noted that a t least one sales exec utive has gone on record as preferring sa l eswomen over salesmen beca use he 60 feels they think faster on their feet ( c ited in Ca rte r and Br yan 1978). 7. There is little evidence to support the truth o f thi s stereotype. In reviews of existing research, neithe r Maccoby a nd Jacklin (1974) nor Eagly (1978) found convincing evidence o f s e x differences in influencability. Nonetheless , s ubseriuent meta-analyses revealed a tendency for males to conform le ss t ha n females in group pressure situations (Cooper 1979) and in non-gr oup situations when a communicator presented arg uments in s upport o f a an issue (Eagly and Carli 1981). Researc h h;1s s hown th:t t the ability of women to be persuasive in their interactions with me n relates to the sex-typing of the communication. In s itu,:it io ns where the intent of the communication is to give informa tion, ;:i nrl in which the subject is one in which female s a r e e xpe c ted t o h vc expertise, or about which men are relatively disinte r es t e d, men a r e more likely to be persuaded by women (Goldberg 1974; 1975; Mo r e loc k 1980). 8. This is attested to by the consist e nt repor ts o f satisfaction with their relationships with c us t ome r s e xpr essed by saleswomen (Swan, Futrell and Todd l 978 ; Futrell 1980) . Swan et al. (1984) concluded that s aleswome n n r e pe r ceive r! fa vo r .::ibly by indus trial purchasing a gents. Pe r ceptio ns by c us t ome r s 0f automobile saleswomen con f irm thi s ge nerall y fa vor a ble 8 ttitHd e 61 toward saleswomen (Lundstrom and Ashworth 1983). 9. It should be stressed that concern for the sale, unba lanced by concern for the customer, can cause the opposite effec t. Salespersons can upset the balance by neglecting the servicing of established accounts in favor of ferreting out new ones. 10. It is possible that affiliative rewards are less important to women in a sex-neutral situation than they are in situations where women are numerically rare. 11. Even women with strongly dominant personalities have been found to decline assuming l eadership positions in relationships with men (Carbonell 1984; Megeree 1969; Nyquest a nd Spence 1986). The "kernel of truth" behind this stereotype may lie in the nature of male-female social interaction patterns. The behavior of male s may make it difficult for females to a s sume leadership positions . Studies of male-female interactions illustrate some of the ways n1en control the conversations: males interrupt females, and change subjects more frequently than the other way ;:irou nd (e . g ., Markel, Lon g and Saine 1976; Zimmerma n and West 1975; \vest and l'. i.rnmerm.:in l978; Fishman 1978). If customers were to exhibit s uch behavi o r s Ln their interactions with saleswomen, sales1vo111e 11 could not c ont ro 1 the interview. 12. Henni g :rnd Jardim (1977) argue that men learn 62 - - --=:.=- organizational behavioral skills when participating in team sports, an activity that is not part of the traditional female socialization experience. These are the same skills that are needed to function effectively in traditionally male-oriented organizations. Lacking these early experiences, females who have been socialized traditionally, may not have the sense of participation as a member of a team when on their jobs. 13. Busch and Bush (1978) interpreted a significant correlation between job clarity and satisfaction with supervision as evidence of communication difficulties between saleswomen and their sales managers. Pharmaceutical saleswomen have reported low satisfaction with their relationships with bo th supervisors (Busch and Bush 1978; Swan, Futrell and Todd 1978; Futrell 1980) and others salespersonnel (Swan, Futrell and Todd 1978; Futrell 1980). 14. It should be noted that the actions of male salespersonncl contribute to the perceived difficulties of saleswomen in functioning as team members . For exampl e, reports have indicated that women are not generally admitted to informal bullsessions of male salespersonne l (e . g ., Women i n Selling, 1974; Busch and Bush , 1978), suggesting that salesmen ar e not willing to accept the 111 ns equals on the snlesforce . 15. The "kernel of truth" und erlying this stereotype is the 63 -.=:- - -=:-==- fact that our culture encourages males to suppress their emotions, Permitting females to express theirs (e . g., Pleck 1976). Women are encouraged to use crying as an emotional catharsis but men are not. Men are permitted to expr ess their emotions through anger, an emotion that is is deemed unattractive in a woman (Braiker 1986; Brownmiller 1984). Resea rch has somewhat confirmed this stereotype. Lombardo et al. (1983) i ndicated a greater Self-reported likelihood of crying among female than male subjects, whil e Kraemer and Hastrop (1986) have reported more crying among females . in natural settings. 16. Greenberg and Greenberg (1976, 1985) argue that salespersons need great "ego strength" t o give them the emotional fortitude to bounce back from such continual rejections, but found no d l? ? ference between salesmen and saleswomen in the degree to \vhich they possessed the quality of "ego-strength ." In contrast, ffennig and Jardim (1977) argue that the vulnera bility to personr1l hurt l? s greater in women because they have not d eveloped the sense of Participating in a game in their approach to their work. 17. In the words of one saleswomen, "as a woman comins in ?;1ith every disadvantage, with no history in business, I'm wi.lli.ng t o use every edge I've got" (quoted in Hyatt 1979, p. 145). 18. The male occupational envi.ronment i_s a more sexuali z ed 64 environment than is the female occ upa tio nal e nvi r onme nt. Males enjoy "talking about sex, a pproa ching women a bout sex , a n d displaying a readiness for sexual inte raction" (Gutek 1985, p. 166). 19. Many women feel that fulfillment o f the i r gend e r r ol e i s a prerequisite for seeking a career. A woman who f a ils in he r obligations in her role as a woman experie nc e s g r eat g uilt (3erna rd 1981; Caplan 1985); a woman who succe eds at both rol e s doe s s o only at the price of great personal stress (Braiker 1986). 20. Research has generally disconfirmed the truth of these allegations. For example, a study by Conc ora n an d Dunca n (19 79) demonstrated by regression analysis that only 3% of th e wage ga p between men and could be explaine d by the 11 lness o f work e r s , t he illness of family members, or of any other e videnc e of l ow l abo r force participation. 21. These mixed results may be explaine d by th e s ituat i o nnJ nature of women's self-confide nce . Wome n ha ve lowr~r self-confidence than do men in s ituations where s t e r eo t y pes attribute a relatively hig h l e vel o f compe tence to me n (Le nny 1981). Results of a t tr ibution reseucc h s upport c1 t e nd e nc y ;1mo ng women to undervalue their own compe t e nce s ince they attri. b11 te thei r successes to factor s o ther than their a bllity s uc h as luc k (Bar-Tal 65 - - - --=...-:::-- -=--= and Frieze 1977; Simon and Feather 1973 ) and effort (Wiegers and Frieze 1977; Erkut 1983 ; Chrombie 1983), but attribute their f.ailures to their lack of ability (Nicholls 1975), a l though some studies have not reported this phenomenon (e.g., Lugenbuhl, Crowe, and Kahan 1975; McMahan 1973). This situation may be changing. Recent reseach has uncovered no differences between men and women in such attributions (Kaufman and Shikiar 1985). 22. Indeed, Brunig and Snyder (1985) presented evidence that low or negative expectancies on self-evaluations may lead to dysfunctional behavior on the job, such as absenteeism, 9oor performance, and turnover tendencies . 23. Research has established a "kernel of truth" underlying these perceived differences between the sexes in the manner in which they are motivated by affiliation and achievement needs. The need for achievement and the need for affil iation seem to be linked in women, and sex roles appear t o influence these relationships (Sutherland and Veroff , 1985) . For males , the need for achievement and the need for affiliation are nega tively correlated (Sutherland, 1980; Veroff, 1983 ). For females, they seem to be posjtively correla ted (Sutherland, 1980; Ve r off, 1983), implying that women With high achievement needs also have hi gh affilintivc needs. 24. Hoffman (1975) theorized that females have 3renter needs 66 for love and approval than do males. She argued that girls experience too much maternal support and protection in their early Years and carry the resulting dependency into their adulthood, retaining the child's eagerness to please adults" in their achievement behaviors. Thus, according to her theory, women look to those i?n authority positions on their jobs to satisfy their approval needs. Alagna (1982) has confirmed that women are sensitive to evaluation by others and respond favorably to approval by others. She argues that women who perform traditionally male roles benefit from peer approval of their activities. 67 I 1 J CHAPTER THREE TIIE RELEV Alff LITERATURE: II The preceeding chapter reviewed the literature concerning gender stereotypes in selling. The discussion now turns to the concepts of selling effectiveness and sex role identity. This chapter begins with a review of the literature about selling effectiveness and its measurement and establishes the rationale for the use of the functions of selling as the basis of a measure of the quality of selling performance. The next c luster of literature conce rn s sex role identity and its measurement. The Bern Sex Role Inventory (BSRI) is introduced along with the theory underlying it s construc tion, the relationship of sex role identity to behavi 1Jr, and its applicability to vocational behavior. The three cl usters of literature are then brought together to establish the conceptual framewor!< for the s tud y . A model is introduced that relates th e sex ro1 e t yp es of Su'' lc s wome n t o perceived gender stereotypic behavior ln the areas of selling ability, human relations, and motivatio n, as well as to e[fectiv(~ performance of the functions of selling . 68 Measuring Selling Performance Measurement issues cloud r esearch findings about the cornparati.ve performa nce O f male s a nd females in traditionally male areas of selling. In parallel surveys of the same industry, Swan and Futrell (1978) reported saleswomen's performance to be significantly lower than that of comparable salesmen, while I3u s ch and Bush (1978) reported no differences between the sexes . The potential for bias in these studies has been discussed. The Functions of Selling Performance of six functions of selling served as the basis for the measure of selling effectiveness in the present study. The functions of selling constitute a model of the personal s elli.ng process involving s ix seq uentia l promot i.onal tcisks that cir e performed by salespeople (Jolson 1974a; 1977b ; 1984; 198Gb). The quality of their performance by saleswomen is, thus, an i.ndi c ,ti.o n of selling effec tiveness . Effective Performance of the Functions of Se lling The secti?o,1 th a t f o 11 ows revi?e ws t h e 1i? terature relcvJnt to 69 7 effective performance of the functions of selling. 1t specifies the tasks that must be accomplished in pe r forming each function and describes effective performance of each. The primary i..ssues involvecl in performance of each function a re also addressed. The tasks to be accomplished by salespersons when performing the functions are enumerated in Figure 3-l. Prospecting. As defined here, the functi..on of prospecti..ng refers to the process of identifying qua lified pros pects for the company's products, culminating in the fi..r s t o ne-on-one relationship between the salesperson and the prospec t. The tasks performed by salespersons when prospecting involve sea r ching for prospects a ncl eva luating their suitability as potential c ustomers . Effective prospectors must be fami..liar with, and use , the best available sources of lead s (Anderson c1nd Hair 1983) . They must caref ully "qualify" prospects l to ensure that compan y re sources are apportioned effectively among the bes t potential c ustomers . Prospects may be qualified either in advance o f the contact, during the contact, or at both times. The key issue i..n prospect i.n g is the> freq ue n t requi remcnt tor sa lespersons to per form effort prospecting (contnctin,.,, persons without company-supplied lead s) . '1any snlespersons ct i.sl ikc? effo r t pros pecting since they must work with indivi.du:cils 1vlio h:..ivc n()t 70 Figure 3-1 Sequence of Tasks to be Accomplished when Performing The Six Functions of Selling Prospectin_g_ Identify Marketing Opportunities Use best available sources for leads Identify many leads Keep meticulous records Maintain search for leads on continuous basi.s Pre-Screen Leads Determine prospect has a need for product Determine ability to pay for product Select proper buying personnel Determine willingness to mee t with salesperson Prepare for Interview Obtain background information about compa ny Obtain background information about buyer Contacting_ Sc hedul e Interview Gai n entry Schedule interviews efficiently Follow up, as appropriate Establish Initial Contact with pr ospect Make good first impressions Establish rapport with prospect Establish prestige of company and self Probing for Needs Acquire Information from Environment Tour faci.lities Study data 71 I 'J Figure 3-1 (cont'd-I) Sequence of Tasks to be Accomplished when Performing the Six Functions of Selling Probing for Needs (cont'd) Elicit Information from Prospect Encourage prospect to talk Appear interested in prospect's concerns Listen attentively to prospect Maintain eye contact with prospec t Di r ect Probing Process Ask clarifying questions (as appropriate) Ask elaborative questions (as appropriate) Change topics easily (when appropriate) Analyze Information Identify main points of prospect's conversa tion Establish why prospect buys Be sensitive to non-verbal cues Retain Information Accurately restate prospect's points Take notes Stimulating Desire Establi s h and maintain interest of customer Address prospect ' s concerns Use demonstrations, dramatizations, o r visual aids Check on attentiveness pe riodi c ly Transmit information accurately Explain technical features Explain product's us~ge Match prospect's needs to product benefit s Use language familiar to prospect Anticipate a nd deal with objections Answer questions Maintain credibility Establish own experti s e Appear trustworthy Persuade prospect to buy 72 Figure 3-1 (cont'd-2) Sequence of Tasks to be Accomplished when Performing The Six Functions of Selling Closing Anticipate and ove rcome objections Invites prospect ' s summary comments Ask for the order Close sale Finalize details Arrange future visits Conclude interview Retaining Retain sale Insure prospect receives product On time ln good condition With access to servic e a s needed Maintain availability Eor servi ce ~ake periodic call s Ha ndle dissat i_ s Eaction and cr)mpl3i~nt s Retain c ustomer Transform prospect into loy, l cu s t ome r Establish relationship with c us tome r Maintain availability Develop account by obtaining: Repeat orders Enlarged orders New ord ers Begin Prospecting Agnin 73 expresse d prior inte r est i n their products ?nd they frequently ? e ncounter c onsiderab l e sa l es r esistance (Jolson l986b).- Some salespe rsons a void the pr os pectin g f unction a l togeth er when it involves effort pros pecting ( e . g ., De uscher , Marshal l and Burgoyne 198 2). Contacting. It is diffi c ul t t o sepa r a t e t he f unctions of prospecting and contac tin g , s ince t hey invo lve ove rl a p ping . . . 3 a c t ivities. The f unctio n of contac tin g i n vo lves : (1) gaining access to cus tomers, (2) making good fir st impressio ns on t he m, (3) redirecting their thought s from pre viou s concern s , (4) es t a b lishing rapport with them, (S) e sta blishing c r e dibil i t y , a nd (6) maki ng n tran s ition from initial c onve r s atio n t o the s t a r t of t he act uc1 l sales pr e senta t ion. Effe ctive c ontac t o r s es t? b li sh the degree of their prospect ' s receptivene ss to the sa l es inte r vi_ew , maki ng ce r tai n t hat sufficient time is a vail a ble, tha t t he physica l co nd itio ns ? re favorable , and that the re will be a mi nimum of i nt erru ptio ns d uring the call ( Young a nd Mo nd y 1982) . They s pe nd no mo r e that n minute o r two in redir e ct i ng t he tho ugh ts of prospects (1?om previ.011 s 4 c oncern s by us e o f " s mal l t a l k " (Jolson 1977b) . To estr1blish good first impressi o ns , the y dr ess ;J t t r ac ti. ve l y nnd :1p pr o pr i;1tely , h ? ve good pos ture a nd a r eady s mile (Yo u ng ,me\ Mo ndy 1()82) . They 74 e stablish rapport quickly. As they guide the conversation to the st art of the actual presentation, they both arouse the interest of Prospects in the product and re-qualify the prospects. 5 The key issues in contacting are: (1) the importance of establishing rapport with customers, and (2) the importance of e stablishing credibility. Establishing rapport with customers is an ? lmportan t aspect of selling effectiveness (Wilson and Ghingold 1980). One way for salespersons to build rapport is by establishing Similarities between their customers and themselves. Success in selling has been associated with similarities in demographic and l?f l estyle characteristics (Evans 1963). O'Shaughnessy (1971-1972) stressed th e importance of ideologica l or attitudinal similarity between salespersons and customers as well as of likabi.1.ity. ile conceived of likability as salespersons' appearance of friendship, approval, and support of the self images for their customers. Ile Suggested that in the absence of similarity, salespersons c3n establish rapport by being empathetic to the needs of c ustomers, determi?ni?no 0 the best way to approac h them, a nd adjustin g their behaviors accordingly. This notion of adapti ng behavi.or t o characteristics of indivi.dual customers has been s s territories, firms, and industries . Multi_-dimensionalltv. Many c ommonl y used ,neasuces of se ll. i_ ng effectiveness ref_ le c t onl y ,3 s i? n g l e 3-s' p ect o f n s c.1l esperso 11' s performance, distorting the r esul ting r ating of ove r Jl l effectiveness. (e . g ., Ru s h 1953 ; Lev in 1963) . \ good nc ,1su i?e -- should reflect all relevant dimensions, and should not reflect irrelevant dimensions (Cocanougher and 1vancevich 1978). The dimensions chosen to underlie performance should be theoretically meaningful, tapping aspects of the job that are common to all sales and should possess face validity, jobs in varying degrees, measuring characteristics that are easily recognizable by the sales managers who use the instrument (Heneman 1974). Six dimensions underlie the functions; these dimensions are complete. Since they are performed sequentially, they are exclusive; since they e ncompass all the selling activi.ties . ll They are performed ex aust1.ve. by the salesperson, they are h theoretically meaningful. Jolson (1974a) demonstrated the.ir existence, in varying degrees of importance, in fo ur widely different selling jobs. They also have face validity, being familiar to sales managers. Freedom From Uncontrollable Factors . The effecti.veness of salesperson's performance is influenced by situational factors beyond their control, such as fluctuations in economic condi. t .ion s , and differences in territory potential. Such factors cnn give r i. se to apparent differences in e(fecti.veness of s ;Jlcspersons ,1ernss industri.es, across firms within indus tri.e s , :111J even across territories wi.thi.n firms (Dnrden and French 1lJ70). Pe rfnn11:111 co evaluation s hould reflect onl y those f actor s that can be con tri)l l,~d --- by the individual being evaluated (Churchil 1 et ;:il. 1985). Good performance of the functions of selling can be described in terms of behaviors that are under the direct control of salespersons. Th us, t h e measure wi 1 1 b e re 1 a t i? ve 1 Y free f ?rom contamination by ext raneous fac tors . Freedom From Bias. The bias inherent in qualitati.ve evaluations can be reduced by basing the measure on readily identifiable behaviors and by clearly specifying criteria to be used for the evaluation of good performance (e.g., Cocanougher nnd I vance vich 1978). Lenny, Mitchell and Browning ( l 983) provided evidence that sex bias in ratings can be reduced, or eliminated entirely, by clear specification of evaluative cr1. ter1.. a . l2 Th e functions of selling a re a menable to description in terms of identifia bl e behaviors, a nd the criteria for good performnnce cn n also be clearly s pecif ied. Reliablitv. A good effectiveness measure should not fluctu~te with random events in the marketplace, but should be co nsistent over time (Jolson 1974b). Superv isor' s evaluation of the quali.ty n[ a salesperson' s performance of the functions should 1Je reL1t i.vely reliable over time . Validity. ? Th e f unctions of selli ng flO Ssess foce v.1lidity. They have been nppl i ed to the evc1luatio11 of salespersons in rour 86 d1" f ferent types of selling, including sel l ing in depa r t me n t s t ores , sales of industrial package manufacturers, s e lling fo r dis t r i butors of g?ift i.t ems, and selling life insurance (Jolson 19 74a) . Applicable Across Sales Jobs. Performance e va lua tion based on the performance of the functions lends itself t o measureme nt acr oss territories, firms, and industries. The importance o f each function in the performance of particular jobs varies cons i der a bly , however. In some selling situations, certa in fun c tions may not even be performed by a salesperson, but ma y be carrie d out by a n impersonal 13 promotional tool. The problem of diffe r e n tial importance of functions across sale s jobs can be cir c umve nted by a '"eighting scheme based on managerially de t e rmine d importa nc e weights. This device was suggested by Cotha m and Cr a ve ns (1969) for use in c ombining multiple quantita t ive me as ures fo r evalu3tion Purposes and was adapted by Jolson (1 974a) f o r use wi t h the functions. The procedure requires sales manage r s bo th to eva lua t e the quality of salespersons' s performanc e on the f unc t io ns a nd t o rate them on the importance of the functi ons to the sal espe r so n' s job. The results are then combine d into a s umma r y measure of overall selling e ffectiveness. Non-Sellin o Activities. In a ny sal es j ob the r e a r e some d uties that do not c ontribute t o the pe r fo rma nce of t he pe r s o nal s e l l i. n~ Process, but tha t a r e necessa r y pa rt s of the jo b. To be i nc lus i.v"' , 87 the se non-sellino activities were l?ncluded in the measure, but are 0 not of any theoretica l interest in the research. 2.wnmary_ Th e f unctions of selling were l?ntroduced as a basis for a performance tool that: (1) meets criteria for a good measure of selling effectiveness, and (2) is adaptable to measurement across territories, firms, and industries. The manne r in whi c h salespersons perform the functions of selling may be r e l ated to tile instrumental and expressive traits that the y possess in their s ex role identities. Sex Role Identity Sex role identity refe r s to the degree to wh ich persons possess personality traits that a re characteristic of their own ,rnd th e opposite gender. Bern (1974) rejected the concept of mascul -i.nity and femininity as opposite ends of a bi-polnr cont inu um. (For o discuss ion of prior masculinity-femininity scoles and the bi-polarity issue, see Constantinople l97 3 . ) Iler theor y pr o po s ed that all individuals possess traits associated wi.th both s exe s i.11 varying degrees . She distinguished bet1vee11 s e x-type d individu:1ls , who s trongly e nd o r se soc i e ty' s stereotypes of appropriaL e bchavi.o r <3 8 for a particular oender and non- se x-typed individua l s , who endor se 0 ' in appr oximate l y e qu;:il traits associated with both genders amounts. Bern developed a scale to measure sex role identity. The scale, currently in wide use, will be used in this study. A Scale to Measure Sex Role Identity of three subscales: masculi..nity, The 60-item BSRI consists masculinity a nd femininity, and social desirability. The femininity subscales each contain 20 personality trait s whic h we r e selected on the basis of their sex-typed social desir;:i bility . Traits on the masculinity scale were judged more socially desirnhle for men to possess than for women. Conversely, tho se o n the femininity scale were judged to be more socially desir;:ibl e [o r women to possess than for men. The social desira bilit y sec1le provides a neutral context for the masculine and fe mi..n ine i te111 s . The items included on the 13SRI are displayed in Figure 3-2 . Psychometric Prope r ties of the BSRL The BSRI ha s exhibited good internal consistency throughout its use. Be rn ' s ori.. g i..nal coefficient alphas for her two norma tive s ample s we r e : :v!asculini t y .86, . 86; Femininity . 80, .82; Social Desira bility . 75 , . 70 . Coefficient alphas for the original ;:indrogyny scale base d o n Nunnally ' s (1978) formula for linear combinations we r e . "3'5 cind . 86 . Original test-retest reliabilities 1,ere go od n t: ' la sc11l. i.11 i t :, 89 Figure 3-2 Items on the Masculinity, Femininity and Social Desirability Scales of the BSRI --------------------------------------------- ---- Masculine Items Feminine Items Neutral Items ---------------------------------------- ----- 49. Acts as leader 11. Affectionate 51. Adaptable 46. Aggressive 5. Cheerful 36. Conceited 58. Ambitious SO. Childlike 9. Conscientious 22. Analytical 32. Compassionate 60. Conventional 13. Assertive 53. Does not use harsh 45. Friendly language 10. Athletic JS. Eager to soothe 15. Happy hurt feelings 55. Competitive 20. Feminine 3 . Helpful 4. Defends own beliefs 14. Flatterable 48 . Inefficient 37. Dominent 59. Gentle 24. Jealous 19. Forceful 47. Gullible 39. Li.kable 25. Has leadership 56. Loves children 6 . Mood y abilities 7. Independent 17. Loyal 21. Reliable 52 . Individualistic 26. Sensitive t o the 30 . Secreti.ve need s of others 31. Makes decis ions 8 . Shy 33 . Sincen~ easily 40. Masculine 38. Soft spoken 42 . Solemn 1. Self-reliant 23. Sympathetic 57 . Tac tful 34. Self-sufficient 44. Tender 12 . Theatrical 16. Strong personality 29. Understanding 27 . Truthful 43. Willing to take a stand 41. Warm 18 . Unpredictable 28. Willing to take 2. Yielding 54. UnsystemRtic risks The nwnbe r preceeding each item r eClects the position of CJch a djective as it actually a ppears on the Inventory. Definitions of ambiguous terms will he included on the instrument, as was done by Comer a nd Jolson (19,'35) SOURCE: Bern, S . L. (1974) 90 ' ) r = . 90, Androgyny r = .93, Social Desirability .90, Femininity r r = . 89. Masculinity a nd Femininity Sub-Scales. The items on the masculinity and femininity subscales were originally selected to be consistent with Pa rsons and Bales' (1955) conceptualization of Such traits ;:is i ndependence, instrumental i. ty and expressiveness . wi.th masculine assertiveness, and dominance are associated such as sym pathy, compassion, rind instrumentality. Traits sensitivity to the needs of others a re associated with femi.ninc expressive or nurturing behavior (e.g., Bern 1974; l975; l976; Spence and Helmreich 1978). The traits on these s ubscoles have cl strong correspondence with the stereoty pes nssocioted masculinity and femininity in our cult ure (e . g ., Broverm.111 rt ,1 l. l972 ; Bern 1979). Construct Validity. Criticisms of the construct v.::ilidity of the masculinity and femininity ~u bscales (e . g ., Pedhaz ur ,nd Tetenbaum 1979; Locksley and Colten 1979) appear to be ori.gLnAte in a difference of opinion as to the intent of the scnle (13em 1977) . When the scale is vi.ewed as a n operationaliz.:1t Lon o[ instrument.J l and expressive qualities (Bern 1977; Spence i!nd lfclmreic:1 J 97') ) , t he construct validity is believed to be good (e . g ., T.:Jylur 198/4; Ramanai?h and ~? rtin 1984). 91 )) -- Classification of Subjects. As defined here, the term sex role type refers to the classification to which saleswomen are assigned on the basis of the strengths of the masculine instrumentality and feminine expressiveness in the sex role identities, as meAsured on the BSRI. Two methods of classification of subjects are currently in use. Bern originally used a subtractive technique to separate subjects into masculine, feminine, and androgynous categories based on the statistical sionificance of the difference between the means 0 of thei?r mascu 1 i? nity and femininity scores. This procedure has been criticized on statistical (Straham, 1975) a nd conceptual (Spence, IIelmreich and Stapp, 1975) grounds. An additive method, introduced by Spence, T-felmreich and Stapp (1975) uses a median split tech nique to divide the sample into four groups of masculine, feminine, androgynous, 3nd undif feren t ia ted indi vi.duals. This me th0d solves some of the difficultie s o[ th e subtrac t i.ve rnethod, b 14 ut results in a classification scheme that is s0mple b~1sed ? No netheless, Bem (1977), endorsed this method after its appUcut i.on strengthened her early results . This procedure is currently the classification method of choice , and was used in the current research. Evaluation of Sex Role Identity While sex role identity is nonnoll y rnensui-ecl by self-reports of tlH' i.nd .i.v?i.duol involver!, two i nstances of sex role i.dentity r :1ted bot-h -~- ----- ------ ----------=--- by self and by another have been identified. Spe nce and Helmrelch (1978) obtaine d ratings of the sex role i.dentities of the parents of college students bo th from the the pnrents themselves, and by 15 the students. They posed three questions: ( 1) are the measure s correlated , (2) a r e the resulting classifications the same , and (3) are the results diffe rent when the da t a are a nalyzed using classifications base d on the two sets of evaluations? They fo un d that correlations between the ratings by s tudents and by the parents themselves to be modest but positive and highly significant, which they interpreted a s evidence that there was some veridicality be tween students ' perceptions a nd the self-images of the parent s . They found only minor diffe r e nces between the mean mascu linity a n d mean femininity scores using the the two classification schemes. Differences wer,~ in the direction of the sex role stereotypes . 16 ,r,a l nssi. f1. cat1. ons into sex role types produced only minor differences i.n the distributions. Alperson and Fr iedman ( 19W3) evaluated th e degr<'P to which opposite sex partners in interperso nal dyads were nhle to vnlidate each o ther s sex r o l e identity, ;:is meosur ed by the 13SRT. (Validation entai l ed ag r eement about 1,hether or not e:1ch i.ndi.vi.dual tra it was characteristic o f the tnrget i.ndi.vidual.) They [n1ind that the masc uline cha r acte ristics of wome n, 1vere less I i.kely Lo be validated by their mal e partners tkin Wt're the i.r fem i.ni.11e <)J -- characteristics. The Relationship between Sex Role Identity and Behavior Bern (1975) theorized that an individual's sex role type would correspond to his or her behavior in sex-typed situations. Behavioral Assumptions. Several assumptions about the relationship between an individual's sex role identity ,-rnd his o r her behavior underlie Bern's (1974) theory. Bem held that sex-typ?~d individuals ~.vere strongly motivated to maintai n gender-appropriate behavior, consistent with society's culturnlly determined standards. Hence, she predicted that they 1wuld excel .:it instrumental or expressive activities corre sponding t o the i.r c;ex role type (Bern, 1975) . She f urther predicted that they 1,ould prefer activities that were congruent with their sex types, would 1Void those that were not, ;:ind would experience 6reater c omfort 1,hc11 performing such activities (Bern and Lenny, 1976). 1n co nlr:i.::;r: , non-sex-typed (androgynous) individuals would disp]ay flexibi li.ty in their behavior, perceiving a wider r.::Jn ge of opti.ons a nd performing activities identified with either sex c01 11 fo rt .?1bly , ,1S dictated by the situation. Behavioral Valid::ition . Be~1 and her co lle.:1g ucs c ondu c t,~d Lhr, 0 1, empirical tests of her theory. Over::il l, i.n t he l.:ibot, .ltory , 94 ------ masculine and androgynous subjects displayed more masc 11line (instrumental) behaviors than did feminine subjects, while feminine and androgynous subjects displayed more feminine (expressive) behaviors than did masculine subjects (Bern 1975; Bern, Martync1 and Watson 1976). Sex-typed individuals actively resisted performing sex-crossed behaviors, while non-sex-typed indivi..Ju?ls were equally comfortable performing activities regardless of their gender-appropriateness (Bern and Lenny 1976) . The se studie:, demonstrated a relationship between the possession of masc uline and feminine personality traits by individuals, and their perfocnn11ce of activities categorized as instrumental or expressive. They establish the theoretical basis for investigating the relations hip between the psychological sex type of saleswomen and their performance of the instrumental ancl expre:,"ive as!w,: ts o[ th -~ fu ncti.ons of selling . Gen der Schema Theory . Bern (1979; l98la; 1981b; 1982 ) strengthened her theory of behavioral correspondence by suggestin:_; that sex-typed individuals organize their thinking around ge nd ,~r lines, possessing cognitive structures known as gender schem.7 i.n their : ;elf L7 concepts. Gender schema theory pcrmi.t s s tron gC' r predictions ::ibout the relationship between psycholo gi.c. 11 scx-Lvp<' and behavior. It provides a theore ti.cal basis for inves t i_ ,,: 1 t i.n ,, ._.., t) the rel.1tionship between psychologi.c;.1l sex type ~1nd perf ,>1-,n.111c c or 95 sale s women, a bov.m1d beyond gender sterotypic behavior by instrumental and expressive activitie s. Challenges to the Behavior Asswnptions. Bern's theo r y of correspondence between psychological sex-type and sex-typed behavior was questioned by Spence and Helmreich (1978 ). Whil e they agreed that Bern's laboratory studies revealed a relations hip personality trait s and between instrumental and expressive behavior, they argued that the corresponde nce extende d onl y to behaviors that were classifiable as instrumental o r express ive . I n (1976) s tud y , a conceptual replication of Bern and Lenny's Helmreich, Spence, and Holahan (1978) confirmed this line of thinking by finding sex role identities t1f s ubjec t s ,rnl y weak l y related to sex-typed behaviors that were not c harr1cte ri znb l , cis instrumental or expressive activities. They ;:i r 3 ue y v:1ri o 11 s authors on dimensions that correspond to the in s trument n l a nd expressive dichotomy. Tnstrwnental qua lities s uch c1s domin a nce , ego drive, achievement motivation, and a ggress i ve ness (e . g ., Merenda and Clarke 1959; Dunnett,~ and Ki rchne r 1969 ; /hrr e lJ 1960 ; Greenberg and Mayer 1964; Howells 196S; Ghiselli Jl)7 3 ; a nd Mat theis et al. 1977) and expressive quc1 lities , s uch .:is empa t hy ( Men:>n d.1 and Clarke 1?~ 59 ?, !!o we 11 s 1C.J 68, ) h ave b e e n f o un d sig ni [ icn n t l y r e lated to i? nc re::ise d s e lling effec ti. venes s. l S Of 111 ten~,, t 1:o t hj s s t udy are the ana lyses o f Jolson (1984 ) :rn d Bbkc ;md ' lout,rn 103 - - - ---- Cl 980). Buzzotta, Lefton and She rberg (1982) classified salespersons on the basis of the relative amounts 0 f "dominance " and "warmth" they brought to the selling situation. Based on their ~odel, Jolson (1984) described three types of salespersons, th;:it he re f? erred to as "aggressive, " " su bm ?i ssi? ve, " and "ass er t _i. ve." A similar two- dimensional model of selling was introduced by Blake and Mouton (1980), who identified five fundnment::illy different orientations to the process of selling based on the relative nmoun:: of "concern for the sale" and "concern for the customer " that they demonstrated. The dimensions underlying the model s::i lespersons of Jolson and Blake and Mouton reflect their relative endorsement of instrumental and expressive traits, and cor respo11 Contacting I I HwA,ban? 1? I I Pro bing I 1 Relations I ( 1) I SCtl iom?iuolag t iog I Motivation I I I 1L __ I I Retaini ? I I ------------------1 I--------------------C RELATIONSHIPS TO BE EXPLORED: Perceived Gender stereotypic Behavior aod Effect iveness (l) Sex Role IdentitY and Per ceived Geoder St e r eotypic Behavior (2) Sex Role IdentitY and Selling Ef f ectiveness ( 3 ) 11 2 gender stereotypes. predictions about the makes no While gender schema theory cross-typed individuals, it is logical to predict thot behav i. or of ine-typed saleswomen, lacking feminine schema, would be less mascul' and behave in stereotypically feminine ways, motivated to less likely co be perceived as confor?inl to gender cons equently be from empirical this view comes stereotypes. s support ?or that have demonstrated chat masculine-typed women pecform findings comfortably in craditionallY masculine envirmc?ents than do more more feminine counterparts (e,I?? Bero and Lenny 1975 ; Rea a nd their Bern's ( I 974) hypothesis of beha? ioral flexibility, Strange 1983). predictions co androgynous fema les who are more extends the in craditionallY masculine environments tbnn arc their comfortable femini' ne-typed female counterparts, role identities of (3) T~ of their erformance he rel ati.o nshi between the sex .,,a 1 esw omen and the effectiveness 0 f til e f unctions of~ early theory of Bern (1974; J975) and cbe conservative The vi oral assumptions of Spence and Helmreich ( I 978 ; 1979) hold beha. ehavioral predictions based on ???en's psychological sex type that b ? made in situ? cio?? chat involve ins crumen ta 1 or ex pres,;i vc can b relationship the activiti' es. Based on?d ? ? to eff"rti.ve perform,111ce of ench Sal eswomen' s sex role 1 entitJ.e 5 ---- - ---- - -- -- - ---- nc ions of selling depends upon the particular combi nation of the fu t? qualities involved. Thus , of instrumental and expressive omen with strong masculine traits will excel at instrumental salesw g ehaviors, those with strong feminine traits will excel at sellin b and those with strengths in both expressive selling behaviors, me and fem in in e traits ( i. e. , androgynous) will excel at mascu1? both?in S t rumental and expressive behaviors, that are addressed in the research are (1) whe thec The key issues impact:, .'.J I\ i: he behavior percept? gender stereotypic eff ecti ion of funct1? 011s of 11 ? veness of sale????? in performing th? se thonise ?) h ? ? O f 1-'omen modera t e and ( - w ether the sex role ident1nes ? ew' ed the l Lter atuF ' and Three have revi - effect, Chapters Two framework the study and ha?? proposed a conceptual relevant to Chapter four presents the methodology underly?i ng the research, that was used in the study, Endnotes 1. The term qualifying prospects refers to the process of evaluating th ei? r SUl? ta bl" ll" t y f 0 r further development as c ustomers. A suitable prospect has a need for the product, is eligible to buy it, is able to pa y for it, has the necessa ry a uthorit y to make the buying decision, and is not committed to another supplier (You ng and Mondy 1982). Prospects may be qualified either i n advance of th e contact, during the contact , or at both times. 2 , Use of foot-in-the-d oor te c hniques , such as a telephon e pre-notification technique (Jolson 1986a), ca n reduce the tr:1um;i associated wi?th effort prospecti.n g. I n us ing Jolson's (19868) pre-notifica tion approach to effo rt prospecti ng , salespe r so ns use a series of low-pressure, te 1 ep h one contacts, t o sec ure comp 11? ;_1 n ce With a series of small requests befor e ac tually aski ng for ;i n appointment for a face-to-fa ce interview. This approc1ch us es foot-in-the door principles to reduce sa l es r esis t ance . 3 . For example, when prospecting by c old calling , snlespe r so ns call upon prospect s unannounced . The functions of pros pecti ng anu c ontacting a re, therefo r e , performed s imul taneo us l y . 115 - - - -- - 4. Since salespe rsons interrupt meaningful activi t y of the Prospect, i? t i? s necessary to re 1a x prospects ? es t a bl ish r a ppor t With them through small talk, and to redirect their thou ghts f rom previous concerns. 5 ? It should be noted that the re-qualification of prospec t s usually occurs after needs have been identified, hence i t ac tua lly overlaps with the probing process. 6. Four model customer types have been id entified, defin ed on the dimensions of responsiveness and assertiveness : amia bl e , expressive, ana 1 yti?c a 1 , and driving. By adjusting the i r beha vio r to these c us tamer types, salespersons are bel ieved t o i nc r ease their se11i? ng effectiveness. 7. Neuro-linguistic programming i s based on the preml sP t ha t all bel1avi? or provides information about interna l ne ura l process0s . Physical cues provide information about me ntal processes . 8 ? ObJ? ee t i? ons can occur at any point in the intervi ew. Si nc e they are most prevalent at the time of c losing ( Young and ,'1ond y 1982 ), they are included within the fun c tion o f clos in g . EffP-c livr~ s alespersons address objections thoroughly 3t the t ime t hey ~r i s e . If objections are allowed to go unanswered, they wdl beco me' barriers to t)1e s uccess ful comple tion of the sa l es cal L. 116 ? In reinforcing the sale , the sal esperson makes a br ief 9 to as s ure the new customer that he or she has made comment d esigned a wise d ec1.s.1 on. does not need to be the sal e 10. It should be noted that many selling si tuations, call? In completed on the initial periods . over considerable time negoti. ations t ake place uties that are necessary component ? of the sal es j ob ll. Some d The se activi t ies ,'1 r e selling functions- the f o e are not part of by incorporating a special ca t egory included in the measure non-se11i?n g activi t ies . ? While sex bias ha? been identified i n evnluations of the 12 ievements of women (e. g., Cline, Holm?? and Werner 1977; Etaugh ach ? ; Lins?????? and Worm?? 1979 ; Palud? and Bauer 1983 ; and Rose 1975 . . Wallston nncl Kiesler and Goldber g 1971 ; Peck 197S?, Pheterson, 81; Zic-nd, Hitt and Pickens 1978) , the f ind ings are by O'L eary ac tuallY r:1t cd 19 fo und women studies no means consistent. Other 1977; Jacobson and Effertz 1974). hio men (Abramson e t a l. c::, he r than likely to occur wh? ? women are evaluated io The effect is most ti on ally ma l e field rather t ha? t rad itio na 11 y fem? le ( Et au gh trad? 5 1 1983) or sex neutral (Etaugh and f ur? ? ? ? ? 1"83 fields . and R1? ley know the ac t uR l pe) rso) ns hc inS In t hese s tudies, eva luators di d not to dee rense ivhen bcha v L ,r ri l e va l Sex bi as has been fo und ua ted information about the individual being r ated is available (Locksley 1 et al. 1980), a nd when the evaluator is personally familiar wit 1 Elrnore and the target person (e.g.' Cullen and Perrewe 1981; LaPointe 1974; 1975; Ferber, Huber and Spitze 1979). Lott (1985) suggests that devaluation of women is ~ likely in situations where the woman is known to the evaluator. 13. retail selling, the functi.on of For example, in prospecting is usually carried out by advertising rather tha n by 3 Salesperson. 14. If saleswomen were assioned to sex role t ypes usin :s Bern' s 0 0 974) subtractive method, assignment would be mor e objective, but the analy s 1? s would b e o b scured since t J1 e proce d u:- e pr ovi? ,i es n o 11sy to separcite androgynous from undi f ferentiat ed fems l cs . .\ compromise method, advocated by Orlof sky, Aslin a nd Gins berg ( l 977) first separates individuals into sex-typed and non-sex-typed groups using the subtractive method, and then further separ~tes non-sex-typed group into androgynou s undi ffcrenti. 1tcd individuals b y me d ian split. TI1is st i1 l r equires the use of c1 sample media n, but i s somewhat mo r e objec tive. l5. Spence and Helmreic h (1978) measured sex role i.dcnti. t y using their own Personal At tri bu tcs Quest i.onn:1 i. re (P.\()). Th,, instnunent is an altern,~1 t1.? v e to t J1 e BSR, I. , bein5 s imilar in co 11 tPnt L l S and construction. The PAQ is also in wide use. 16. Students perceived fathers as higher in masculinity and mothers as higher in femininity than the parents, themselves, reported. When the statistical analyses were repeJted using Parents' ratings of students' sex role identity, the overall trends seemed to be similar. 17 ? A schema is a network of associations that organizes nnd guides an individual's perception. According t o Bern's view, sex-typed persons evaluate their adequacy as persons in terms of their gen d er schemas. 18. Not all studies have reported these relationships. For example, Lamont and Lundstrom (I 977) found no significant relationship between either empathy or ego drive and the scllin8 effectiveness o f i? ndustrial salespersons. 119 CHAPTER FOUR RESEARCH DESIGN This chapter describes the methodology that wa s used i n the st ud y' the na ture o f th !:! specifying the key variables, the hypotheses? the a na l yt i ca l sample, and the data collection procedures, techniques that were used. Variables The independent sales mana gers ' pe r ceptio ns of variables we re gender stereotypic behavior of saleswomen. Th e de pe nd ent v, 1rinb l es were the effectiveness of sale swomen in performi ng 8 se r ies of six functions of selling and non-selling activities . The e f fec t of o ne moderating variable, the sex role type of sales1vo111 e n, was investigated. /\ series of control variables were also in c lud ed . Independent Variables The independent variables we r e sales ma na13c rs' pe r ce p tio ns of gender stereotypic behavior of s aleswome n r e l ati ng t o "se ll i.n,, ,) a bility," "human r e lations , 11 a nd 11 moti v.:itio11, 11 a nd n s ummnry 120 Multidimensiona l sca l es we r e measure of these perce ptions . developed to measure the degree to whi? ch saleswomen we r e pe r ce ived to behave stereotypically in each of the three a r eas . Sal es 7, ho w managers fr om 1 t o were a sked to indicate, on a scale closely each item on a series of statements describe d fr om o ne to four target sal e swoman. Three individual stereotype scores were calcula t e d fo r eac h saleswoman: a "l11unn n a "selling ability" stereotype scor e , relations" stereotype score, and a "motivation" ste r e o t ype score . An unweighted three stereotype scales wos ;1lso average of the calculated to serve as an overall mea s ur e of the degree t o which a saleswoman is perceived to be stereotyped. Dependent Variables The depend e nt variable s were the e f fe c tiveness o f sa 1 es11ornen i.11 performing six functions of selling and non-selling ac tivi.ties , a nd a summary measure of overall effectivenes s. Sales mana ge r s r a t ed the effectiveness with which they belie ved the t a r ge t sa l es1v0me n performed each of the functions, us ing a scal e of O t o 10 . Soles managers also rated th e rela tive impo rtance of th e f unctio ns to tlw perf ormance of the sale s job. These irnpo rta ncP wc ~i. g ht s we r e summorl a nd normalized to conform to unit leng th. An o ve r.i l l se ll i ng e ffectiveness score was ca lculated ns an aver age ,Jf t he i.n div i.,lu,11 121 function effectiveness scores weighted by sales ma na3e r s ' r a tings of th e importance of each function to th e effective per fo rmance of the job. E?a ch saleswoman receive d a scar e 011 eac~1 individua l function of selling and on non-selling ac ti vi ties. Eac h suleswoman also received an overall selling effectiveness s core. !i._oderating Variable Sex role type of saleswomen was investigated as a vnriablc that moderates the relationship between the perceived gender ste r eo typic behaviors of saleswomen, and selling effectiveness. Sales mannge r s provided descriptions of their perceptions of the sex r o l e ident2? ti? es of saleswomen using a short forrn of the I3SRI . Ba se d 0 11 these ratings, saleswomen were classified into s e x r o l e t r pes Within each sample, using a median split technique as r ecomme nd e d by Spence, Helmreich and Stapp (1975). Those who were hi g h on both masculinity and femininity scores were classified ns :Jndrogyno u.s females . Those who were high on masculinity and low on femi nini ty were class1.?f-1?ed as masculine females. Those 1vh o we r e 1 0 1.,r on masculinity and high on ferninini ty were cl::issified 1s fe mi n i_ nc females, while those 1vho were low 0 11 both masc u hn i ty ;in d femininity were clossified Ds undi f f e ren t i;:ited femol cs . TI](' classif? t? I ? ica ion sc1eme is a convenient way of s eparati.n :.; in ch v i.du.1 1s on th e basis of the relative ::unounts of masc uli.n.i_ty and [Prnini_n i_t:y in their sex role identities and nl so r esult s i_n r e lotivcl_v e qu ,11 122 It is limited, sample sizes within cells for statLstical analyses. however, . of saleswomen to sex r o le types i s since classification only rela tive to other saleswomen in the same sample . .?2ntrol Variables A set f 1 d d t ai? d 2? n clarifyin 0? the 0 control variables we re lnc u e 0 relations hips of interest. ( A discussion of the manner ln whi c h control variables affec t or c larify relationships is '.5iven in Green and Tull 1979.) Control variables wer e used to assess : ( 1) whe the r a ny differences in the relationships between dependent vari a bles and sex role types resulting from addition of a ny of the c ontro l va riables, ( 2) whether the basic model could be Lmprnvc d by a ddi.t ion of the control varia bles , or (3) whether any o f th e control va riables co uld pr oviJe better explanation s of t he variability in the dependent variables. Control va ri.Ji.J lcs th,11: were includ e d were the age , education, fa mil y stat:u s , a nd tenur e o n the salesforce o f saleswomen, as we ll as the sex ratio of the s Jl0s force. ~- The age of saleswomen potentially rehit s to bo th perceived gend er s t ereo ty pi? c b e h avi? or an d sel1. 1?ns) c-f? f ec t L? ve ncss . Since younge r women are l ess likely to holJ tradit ionn1 ,1t t l t1 1d cs toward sex roles than are older women ( e . g ., Thornto n, t\lwin ,ind Cam burn 1983; Helmreich, Spenc e and Gibs on , 198:.2 ; Sl c vLn .111d 123 .k 1 t be perce ived as Wi ngrove 1983), they might be less l i e y o A rela tions hip behaving in accordance with gender stereotypes. . ng ef fee ti veness ( e ? g ? , Hae fe r might exist b etween age an d Selll variable in 1986) . control Age has been used previously as a research into gender differenc es in selling (Bus ch a nd Bus h 1978 ; Swan, Futrell and Todd 1978; Swan and Futrell 1978; Futrell 1980 ; Gibson c 1 980 - 81 ; Gi? bson and Swa n 1981-82 ; , ochran and Epley Skinner, Dubinsky and Cronin 1984), and has been found t o impact differen t i ally on job- related attitudes (Gibson, Cochra n and Epley 1980- 81) and values (Gibson and Swan 1981 - 82) of real e sta t e saleswomen. Education. The education of s ale s wome n po t e ntially r e lntes t o both perceived stereotypic be havior and s elling ef f ecti veness . Because persons with higher educations are l es s likely t o hold traditional u,. .,tt.L? tudes t owar d sex ro 1 es t h an t h osc 1, 1.? t /1 J-,-=,ss education (e . g, Thornton, Alwin and Camburn 1983 ; Helmrei c h, Spence ao d Gibson 1982) , saleswomen with higher educations might he l ess likely to be perceived as behaving in acc ordanc e wi th gend e r stereotypes. Because education develops important pe r s onal ski ll s that are needed for sa 1 es J?O b s, sales women 1. Lt h higher e d uc ,1 t i_ ons might also e.?ccel i? n se1 1_ i?n g. 1 ,"__', d ucati?o n was mea s ure(l b y yea r s ,J,r formal education. Educ3tion ha s previously be e n incl 11d ed ;1s ,1 c ontrol variable L!1 research into gend e r dif fe r e nce ::; i n se l li.n g l ~4 (Busch and Bush 1978 ; Swan, Futrell and Todd 1978 ; Swan a nd Fu trell 1984 1978; Futrell 1980; Skinner, Dubinsky and Cronin ). Marital Status . The marital status of saleswornen potentially behavior. Married relates to perceptions of gender stereotypic women have been criticized for permit ting family re s ponsibilities . h ? ? b (e g Bergmann t 0 interfere with the performance of t eir JO s ? ?' 1986 ,? K mar?ri? ed mi? ght appear to conform anuk 1978). Women who are more closely to gender stereotypes than women who are not. Tenure Tenure on current job potentially on Current Job. Saleswomen who have pe r fo rmed relates to selling effectiveness. their jobs for longer periods might be likely to perform their jobs more effectively. Length of time on job, dichotomiz ed into s hort - and extended-longevity groups, has been used used to clar i fy relationships in research into gender diffe rences in selling (Swan , Futrell and Todd 1978; Swan and Futrell 1978). The Sex Ratio of the Salesfo rce. The sex ratio of tile salesforce upon both gend er s t e reotyp i c potentially impacts behavior and selling effectiveness. Saleswomen mi ght be 1111>r <"' likely to be perceived as behaving according t o gend er stereotypes on sales jobs where women are statistically r a r e than on jobs 1,hl:ri~ they are plentiful (Kanter 1977; Taylor a nd Fiske 1978) . Becrn1s the pressures they encounter in these s ituat ions di ffe r, s:Jl c s1Mrn'11 125 No previous r esearc h may also differ in selling effec tiveness? the sex into gender effects in selling has explicitly considered ratio of the salesforce. Q_ther Variables A set f d 1 d d to use in testing the 0 emographic variables was inc u e samples for comparability. A set of variables was included for the purpose of testing for sex bias in the evaluation of sales1,romen . felt in These were: (1) the degree of comfort sales manager s supervising saleswomen, (2) the sex ratio of salesper.so nnel rep or t? (3) the sex of sales managers. ing to the sales manager, and for the purpose of t est ing [or A set of variables was also included differe n t?i a 1 response tendenc. These were self-perceptions of: ies. how ( 1 ) their leniency in evaluating saleswomen, ilnd (2) well-qualified they felt to perform the evaluations. Demographic Variables. The ciemographic voric.Jbles that ivere included nnd were the age, education, a nd sex of sales manrt8c r s , the age, education, marital status and tenure on the salcsforce of saleswomen. Comfort in Supervising Saleswomen . Th e de~ ree of com[1J 1?t felt by sales ma nagers in supervising saleswomen potentinlly 'lffect" tl!e ev~luations of saleswomen. Some sales managers hav e been report,'.! to feel uncomfortable when supervising wome n, pa rti c 11l;~rl y when 126 to take di s c iplina r y required to g ive constructive criticism or who feel ve ry comfort a ble action (Futrell 1980). Sale s managers the gend e r different perceptions of supervising women might have st of performance of saleswome n ereotypic behavior and the quality than those sales managers who do not. . f ? The sex ratio of sales pers ons The Sex Ra tl 0 0 the Supervisor . otentially impac t s upon under th e supervision of sales managers P who the way sales managers evaluate saleswomen. Sales s u pervise only a few women might have different perceptions of the gender stereotypic behavior and the quality of performa nce of saleswomen than those who supervise many. Sex of Sales Manager. TI1e sex of the sale s manager potenti.nl L:r impacts upon the evaluations of saleswomen. Male anJ fema l e sa l es managers may have differing percept i.ons of the ge11J e r s t e r eo ty pi c behavior and tl1e quu~ 1i?ty o f per f ormance o f sa 1 c swome n. . Tendencies toward Leniency . Some individuals have a t e nd e ncy To to rate sa 1 esperson s with e ither leniency or h ar s I1 nc ss . counteract this bias, sales managers were specifica lly r eques t e d t o be highly c ritical of sale s women in their pe rformance ratings rind r cm,1 i.n were assured that the evalua tions the y ga ve would confidential c1nd be used for researc h purposes only . The y 1-1 c r c asked to evoluate themselves as rate r s on :i lenie nc y sea.le. TL 1-1 c1 ' 127 of sa l es managers who cons ide r ed intended for the responses the mse 1 ves to be lenient raters t o b e compa red to those who considered themselves to be harsh raters . Degree of Qualification. Many sales managers are not in direct, daily contact with salespersonnel under their supervision, and might feel that their knowledge of the target saleswoman' s performance was inadeq uate. Sales managers were asked to indicate how well-qualified they felt themselves to evaluate the targe t saleswomen. Responses of sales ma nagers who felt themselves t o be very well-qualified to evaluate saleswomen were compared wLth those who d l? d not feel very well-qualified . Statement of Hypotheses The research hypotheses were constructed arounJ the mode1 that \v.'JS developed in th e precee d ing chapter. They were des i gned t o investigate tl1 e th ree 1 i?n k s in the mode 1 i?n varyi?n g d egr ees o f specificity . For each link, the most genernl case is prese nted first, followed by a more specific case . Because the applLcHhLLiL v of the concept of sex ro 1 e identity to the f? ie J_ d of se 11 i.11 g h:1--; been 1 arge 1 }' unexplored, the hypotheses :1re s t ci t ed i 11 th e nul L form. 128 fupotheses One, Two and Three investigate the Hypothesis to one, two, and three were included of gender relationship managers' perceptions between sales stereotypic behavior of saleswomen and selling effectiveness. Hypothesis the relationship between the over all one concerned perceived gender stereotypic b e h avi? or scores of saleswomen a nd the -overall se1 i ? Hypothesis two concerned the ing effecti.v eness score. relationship between each of the three individual perceived gender stereotypic behavior scores and their overall selling effectiveness scores. Hypothesis three extended the analysis t o the relationship between each of the three individual stereotype scores of saleswomen and eac J1 o f t h e si?x i?n d i? vi?. d ua 1 functions sco r es and non-selling activities three, scores. Hypotheses one, two, and expressed in the null form are: f!o(l) There will be no significant correlation between th e overall stereotype scores of saleswomen and their overall selling effectiveness scores. Ho( 2) There will be no significant co rrelation between e.1ch of ~hree individual stereotype scores of saleswomen nnd their overall selling effec tiveness sco res. Ho(3) There will be no sionificant correlation bet '.veen c.1ch 0 o[ three individual ster eo type scores of saleswomen nnd each of their individual function effec tiveness scores. Hypotheses Four and Five Hypotheses four and five were included to investi gate tlw l:?9 role types of saleswomen and sales relationship between sex behavior. Hypothesis manag e r s' perceptions of gender stereotypic between a summary measure of four evaluated the relationship perceived gender stereotypic behavior and the sex role types of 2 saleswomen. Gender schema theory (Bern 1979; 1981a; 1981b; 198 ) holds th 1a t sex-typed individuals organi? ze thei?r thi"n Ki?ng around gender lines. T11ey would therefore be expected to be more likely to behave ? Thi s suggests in accordance with gen d er s t ereo ty Pe s ? that sex-typed saleswomen, with strong feminine and weak masculine traits (feminine females), in their sex role identi.ties would be ~ likely to be perceived as behaving in accordunce wi.th gend er stereotypes. Hypothesis four, expressed in the null form, is: Ho(4) There will be no significant differences amon g the me? n of overall stereotype scores of saleswomen c l assified according to their sex role types . Hypotheses five extended the analysis to the relations hip between each indi.vidual s t ereotype scores and the sex role types of - saleswomen. Selling Abili tv Stereotype Scores . Since sex ro l e masc11 li.n i. ty in women has be? en foun-u' to be associate d wi? t J1 h 1? g I1 er cog ni? t L? V r-' ability Ofaccoby 1966) and with mathematica l (Eccles [(J13J) a nd mechani ca l ( Anti 11. an polarit v 0 f h - ? alf of the items on the individua.I s t ereotype scales to g uard against th e bias of acq uiescence . [/hen developing these items , it b ecame obvious that some ivere be tter ivorded ns ncg:.i ti. ,?es and others as Posit ives. For example, asking questions alio11t " se1 1 J?. ng ability,, i? n a nega tive way sounded demeanin g to so.IPs1v0mcn (e.g.' does not have good produc t kno1vle d ge , d ocs not rc:1 lI y I ; 11011 job, doe not ca tch ) onto new concepts quickly . On the other lw11cf, some of th e "huma n re.Iations " and "motivation'' items secmecl tn hr' better phrased as negatives in the interests of c l ar ity. Revcr:3i11,; their polarity d pro uced cumbersome items that wece confusi.n?t to 135 does not have respondents (e.g.' does not over-nurture customers, ove 1 ? ) Accordingly, each item r Y emotional reactions to reject.1ons ? The f inal was asked in the way that seemed most appropriate. the questionnaire, are shown in items, 3S they appeared on Figure 4-1. 1_ength of Questionnaire The original questionnaire was pretested on a small 3roup of sales managers who volunteered to assis t with the r esearch . It w::is found the that the questionnaire was excessively long . Accordingl y , queS t ionnalre 1,as shortened: Function Effec tiveness Scal e . The original funct i.on sub-scales , each effectiveness scale consisted of seven multi-item single item composed se ven of 4-5 items. These were cond ensed int o scale s . In its revised form, only one eva luation was requi.red per fun c tion per saleswoman. Detailed descriptions of good performn11cc were provided for each function. The items from the functi.rrn effective ness scales, toge ther with the descriptive information, that a ppeared on the final questionnaire are given in Figure 4-~ - The 13cm Sex Role Inventory. The questionnaire wns further shortened by using a s hort form of the BS IU . ?n,e number of ite111s was red ucecJ from from 40 to 20 , 10 from the masculinity nnd l 'J fi- 0,11 the femininity sub-scales . The i terns from the mascu 1 i.n L y md l36 Figure 4-1 ltelllS intended fo, the Selling Ability,"-" Relations, and Motivation Stereotype Scales Ability Stereotype Items (1) Is not well organized. Has strong persuasive skills (RC) . (4) Displays unusually strong product knowledge (RC). (7) Catches on to new concepts quicklY (RC) ? (10) Adapts quickly t o non-routine situations (RC) . (13) Really knows job (i . e-, cleacly uodecstands how to pecfocm (16) her job including tasks to be performed, priorities of tasks , and how to allocate time aroons tas'") (RC) ? Human Relations Stereotype Items Tends to over-socialize with prospects / customers during (2) Asaclceesp tisn tperrovsipeewcsts. excuses and objections at face value. (5) Tends to relinqui sh control of sales interviews to customers. (8) Has overly emotional reaction to rejections on job- (9) Over-nurtures customers ( i.e., spends too much time on (11) providing extra services to present customers while neglecting other aspects of the j ob , such as prospecting Ffourn cnteiwo ncs u1s,troelml earss )". team player" (Le-, cooperates with others (14) Mona ijnotba in(Rs Cs)e. xually neutral manner on the job at all times ( RC). (17) Motivation Stereotype Scales Tends to avoid activities that she considers to be unpleasnnt. (3) (6) LNaecekdss ccoonntfiindueonuces . reinforcement to sustain performance . (12) Allows famly responsib ilities and / or personal considerations (15) Htoa s in"dteynrfaemriec wininthe r jodbr i. ve " (Le., makes every sacrifice of (18) time and effort that is necessar y t o succeed) (RC) . (RC) Reverse coded 137 Figure 4-2 Items intended for the . Function Effectiveness Scale with P ive Information about Effective Performance De scra1s tth? ey Appeared on the Final Questionnaire (1) Prosp- ecti? ng. good in prospecting involves generating sufficient numbers Eroeff qfuecirtei. vde. npersos sp. ects to be contacted for the firm's produc t as for the ~Good prospects are potential customers that have a need the purcr ? uct, the ability to buy it, and the authority to make calling ha0sing decision). Effective prospectors do not ovoid cold is necessary to the job. wen it . (2) Contacti_no ettino . s in contactino involves routing calls efficiently, Effect establi ? vshen a? ? es o.o rs easily, making eicellent first impressions, gth e como ming dr apport quickloy , and establishing th? legitimacy of pany and the credibili tY of herself as its representat ive. ( 3 ) Probi? ng for Needs . ldent. f . ness in probing for needs invol,es being adept at lE?f ffeecc~t?l'yV ien g the needs of probl?? areas of prospec t s /c ustomer s . Probin~ve probers are good listeners and ore abl'1.'0 to direct t he process through the use of discernin l queSC n?? ~(4~~) ~r Stimulatm? g Desir~-performance of stimulating desire involves g.l vinl a_rticulnhte ' add mative a d . presenta tions / deroons_trati o_ns t ::it oft the ne d ? t / ustorner s Effective s tiinul nto r s eren ss ' en s op ? erpsuasive eff use sale .. d ( rospec s. c 1 aids? ? produc t demon s tn n.o ns ) and ect?i vely ass ai s e. .g., t v,suthae prese' ntation, f"i e l1d ques t1? t) ns ab ob J. ecti.o , ns adp pd ro1p ria. te1 ho ffectivelY and are cno1v 1 eu. 1g ::i bl c th out compet1.? t.a n ea d ,it , t em?? e effective' compari.s ons be t1,een 1 Pr eermm ita n dtw ot he civome papnryo' su cptrso, dumcat, 1wnghe n apt prtohperiira te1 Jrteos endt oat1 :.on.s , Thb. ucyt 1 etai way co,nmunication thro?I1 ,oo ' n stro ng control of the intervJ?. e ?? Figure 4-2 (cont'd) Items Intended for the De Function Effectiveness Scale with scri?p ti?v e Information about Effective Performance As they Appeared on the Questionnaire (5) ~- Effect? o r d er J. aV eness J.. n c losin o involves asking for, and obtaining , the cl oser' s s a ppropriate too the job. Salespersons who a re effec tive of tirnin approach the close with confidence, and have a good sens e on the f~ when asking for the order. If the order is not obtained Sit Uatiol'lJ.. rst att empt, they folo1-1 up as appropr?i ate t o t h e selli?n g ( 5) Reta? ? ~-~----1-....:r.::n.; n.~0 th e Sa le and the Customer. Effe . f ct:i.vene . 0 11ow-u ss 1.n retaining the sale and the cust omer involves Product P after the sale to be sure that the customer receives the and . as promised (e g on ti? me in good condition, s erviced apPropJr.niastt a 1 led), a'n d ? pr?o' viding g' ood customer services , as Cust0 e. Effective retainers maintain regula r contac t wi th n eeds rners ' are responsive to chanoes in customers , pr o d uc t-re l ated ap Plic' ata. nd 1c eep customers inform?e d about new pro d uc t .", d? r1 d . ne 1, exPed. . J.ons. They handle dissatisfac tion and compla in t s J. t1.ous1 y. (7) ~Non-se11 ? - Good keepJ.. nPg erformance o f non-se1 1i ?n ? d u ti? e~ s i? nvolves s uc h _ things as PromPtl accurate and up-to-da~e records, filing call r eports Y, a nd controlling expenses. 139 - ( femi ? h final n1nity scales of the BSRI as they appeared in t e questi onna?i re are given in Figure 4-3. The Attitudes toward I/omen Scal1:.- It has originally been intend eel to ? the atti? tudes of include a 15 item scale to measure sa l es managers toward women as a test for sex bias among responding sales questl?onnai? re still managers. I n or d er to s h or t en the further a one-item measure of the degree of comfort in supervising saleswomen i,as s ubstituted. 2 Eight As a f"i nal pretest small pilot study 1,as performed. a sal es manager s were contacted by t elephone and asked to partic tp,ite in the Uter research. The questionnaire was transmitted by mail. the were sales morwgers had reviewed the questionnai,e , they interviewed by the reseorcher. No serious pr oblems 1{e,e ,eported in either l e ng th of questionnaire or comprehensibility of i tems . Overview of the Study Stage Tvo: Survey of Sales Managers .Sales managers were contacted by tel ephone over a three mont/J Period beginning in Augus t and ending in October- of 1987 . Th e researcher personally telephoned selected firms and nsked to spe:1k w-Lth sales manage r s . If sales managers were not i.mmcd i 1 te 1y 140 s~ Role Identity Itemfsi gIn~cleu4de~d on the Questionnaire A Short Form of the Bero Sex Role Inventory ~ ~culinity Items (2) Is affectionate (1) Defe n d sown beliefs (4) Is sympathetic (3) Is i?n dependent (6) Is sensitive to the needs of others (5) Is assertive (8) Is understand ing ( 7) Has a strong personality (IO) Is compassionate (9) Is forceful (12) Is eager to soothe hurt feelings 01) Ha s leadership abilities (14) Is 1;arm 03) Is wi.lli? no to take risks ( 16) Is render (ls) I .., s domina nt (18) Loves chilJren Cl 7) Is w.il.l inq to take a stand on i':'s sues (20) Is gentle 09) I s aggressive e to conta ct them, until they available , further attempts Were mad Were reached. In some cases contac t wa s made on the fi.r s t attempt, Sales ma nage rs while in other cases, many calls were necessary . represented of fJ.?rms, r a nging from regiona l offices a wide variety 3 of lar ? smaller local firms. In the ge national firms to offices o case of than one sales mana ger was some very large firms, more contacted, but in no case di d more than three sale s manage r s respond from any one firm. R?ta Collection Procedures Sales managers who were contacted were informed as to the L:ene r a l purpose of the research and asked whether or not they s upe rv i. sed saleswomen. h Superv].. sed s "les wornen 1,e r e Those sa 1 es managers w o '"' asked to participate in the survey. They were info r11~il 1-ha t (1) participation would involve c ompletion of a self- .:idmin is t err!d questionnaire describing some of the saleswomen who wo r ked f,)r them, (2) completing the questionnaire would take n pproximnte l y one-half hour of their time, and ( 3 ) the saleswomen t hJt they described would be comp 1 etely anonymous, only they would know t he identity of the women they chose to de s c ribe . All s ales ma n.Jg(!r.s ivere promised a c opy of the rnano gcment s; 1mmary when th e r esc,1rch was complet8. Questionnaires were mailed direc tly t o s J les managers together with a personal l e tt e r o f trans mi.t t :il from the res earcher. A formal cover l e tt e r was printed on th e fro n t or t he 142 quest? lonnaire. The cover letter related the purpose of the research , a nd gave detailed instructions to the sales manage r for completl.o n of the questionnaire. The questionnaire 1,as Self-ad . mJ.nistered. Completed questionnaires were returned directly to the researchers in a self-addressed, postage-paid envelope. After the completed questionnaires were received, sales manage rs were sent a personal thank you letter from the researcher . The letter of transmittal and thank you letter appear in Appendix A. 'I'he cover l etter and the final questionnaire appear in Appendix B. questionnaire 1,as designed in such a way that sales could respond about as few as one or as many as four sale swomen ? ,nc .ach sales manager ivas asked to se l ect sa 1 e s1-1omen I w 10 were maximally different from each other in their selling s t yle , Pe:rformance ? and results. sample of sales managers i,ho supervised sales1,omen i n different 11 male selling j obs ims types of traditiona Y assembled S b t ype of customer. This ales jobs were classified Y Classificat. d by Comer and Jolson ion was consistent 1,ith that use (I98S) . preferences fo r sel lin13 ln their investiP.., ation of st udent S sales managers 1,ho ample one consisted of responses from SuPe:rvised . al end use r s . Sample sa leswomen selling t o organiza tionc 143 who supervised managers from sales two consisted of responses Sample large ticket items co in-h??? con??????? saleswo men sell1' ng of responses from sale? managers who supervised thre either e consisted resellers, resale to Sal for eswomen selling products lvhol esalers or retailers, who were managers 195 sales Of the Of Response Statistic~-188 agreed co receive the questionnaire . 132 returned eligl? b le to respond' questionnaire, those who agreed to receive a response rate of 70.2% of those completed 'usable questionnaires, Of the??? 77 sal?? manager? represented sell ing W. ho were eliing viob 1lv e e.d se JI ing co ind us crial end users, 26 re pr esen te d Jobs tha t selling Jarge ticket it??? co in-hose 111 senl?n g J?o bs that involved 1 ?? 29 represented selli?I job? chat tn?? ? ?d ?? 9""0 co cons umers ' d These resulted in a cot?! of 202 ca??? of sol? " """ resell ers 63 1vho sold for the a nalysis: and 511 ff icientlY Sold larg to e-ti.c ket items ensured a ThiS businesses , Products for reale by group~ te sample, large Lor analysis within each separs 4 for performed were sam rests 1es , differences were Com arabilitv of no com parab?1? of the three l lty of of th found among the demographics the t hC e Sig n.i ficant differences emer~ed s;i1 ~ eswomen th at they descri bed. . _.. . ..? . .. __..;"'."' ..... .. ., : : '. . .. : ; ..?.-....- --- --- ut analyses 1vere conducted separa tely fo r the were not combined b for sal es with national averages three groups. In comparison proprietors ( Bureau of La bot Statistics I 986) , on supervi? sors and sales managers were older, better educated, and more the average, of national averages likely In comparison to be male. selling commodities , exclusi, e of retailing ( Bureau of Sal eswomen ics 1986), on the average, the saleswomen who were Labor St atist? younger and better educated. Details of the sample descrl?be d were char acterist?i cs are contained in Appendix C. ~uest?io nnai? re ire that was completed by sales managers consis ted of The questionna? nine sections: O ~ction On e. An estimate of the reJati?< of selli? d - act;V; t;es to the saleswoman' s funct.i ons ng an non-se1 1ing ~ ~ ~ A scale of O to 10 i,ra s used' job was requester! in Section One. to a function that was compJetelY irreleson t and Wh ere to"O " a rfeufne rred ,, . rtant to a sal eswoma n' s 0 10" ction that was criticallY imP concise d f. . . I f t. n 1vas provided. job. A e 1n1t1on of eac1 unc io of t a rget performance of sw~ocmtion Descriptions 5 en 0 ---=-WvTv uO ? 005 f eJJiog and sale on performance of the six functi -sellin>! Section f1,ro. A sca l e of 0 non was activities were requested in 1 of ef f ec t iveness . 0 1 to 10 used where "10" was the highest eve 145 i n concr ete described For each item, good performance was Criteria establishing good performance of each beha vi. oral terms. The criteria were determined based on a search of were provided. literature and oere confirmed by discussi ons wi th the sale smanship representing the three types of selling jobs , sales managers Sec tion Items from the BSRI w-ere included in Section Three. managers reported their perceptions of the sex role Three. Sales A scale of 1 to 7 1,ras used, w-hcre .i, dl"e ntit?i es of target saleswomen. never true" and "7" referred to " referred to "never or almost alwa ys or almos t always true." from the gender stereotype scal es ?.,ere Section Four. Items degree t o Section Four . Sal es managers in di ca te d the included . to be i n 10 perc eived the behavior of target sa 1e swomen Which they "sel li. ng 1tems f r om the accord ance gender sca l es 1,rere with "motivation" ab? and llity," "human relations" A scale of 1 to 7 was used' on which "1" refe re ed to alt ernated d referred to "al ways or almos t 11711 " 11 never or almost never true an alwa ys true " about t he sa l eswomen who Demographic i.nformation Info rmat ion Section Five . Section Were i n stn tUS, and described was requested mari.ta l education, re age , The _inf orma t i.on quested pertaininn0 to the s e 11 in g job of each targe t sole , woman ? tenure on was evaluate the three samples demographically. used to manage rs Demographic information about the sales Section S ix. was requested in Section Six . This information pertained themsel ves sex . These their education , and to sales managers' age , their t o t o describe the samples and va ri? abl es were included in order comparability among sales managers in the different types test f or Sales managers were asked the percentage of their of sales jobs. the percentage of female and company's was female . salesforce tha t that: ,.,ere su pervised sale spe rsons that they personallY be that tor effects T to test tO rate his inf ormation was included were asked managers attr ? Sales describe i b utable to sex ratio- how well qualified theY felt themselves to themselves on the relative leni ency of saleswornan, on the c har acte ris t ics of each other with the way their comparison .1nd theiY evaluat ?i?o ns in wi. t h their evaluate comparison same level might rest s degree in supervising to provide of comfort Toese i terns were included supe rvisi?n g salesmen. for the degree of bias in the responses - t h >'. ' -::?: '.";?:I; ' ' ?.. -----------~ Table 5-7 Test of Hypotheses Five Means of Selling Ability, Human Relations and Motivation Stereotype Scores of Saleswomen Classified according to Sex Role Types Sample one Sample two Sample three To Organ'l To In-home To End Users Consumers Resellers = 202) (n = 72) (n = 63) (n AF UF FF MF AF UF FF MF AF UF FF MF -------------------------------------------------------------------- Univariate Analysis: Means of Stereotype Scales: Abilitv 3 .71 3.00 2.48 2.03*. ... 3.83 3 .15 2.40 2 . 24*** 3 . 82 3 .29 2.28 1.64*** Human Re l a tions 4.10 3.77 3.13 2.54- 3 . 86 3 . 95 2.69 3.59* 3. 36 3.46 3. 51 2.61 Motivation 3.52 3.00 3 . 13 2.65-IHHI- 3 . 70 3.39 2.72 2 . 45*** 3 . 02 3 .06 2.67 1. 74*-IH- Effect sizes: Ability .35 . 35 . 39 Human Relations .25 .1 2 .1 2 Motivation .31 . 29 . 26 Multivariate Analysis: Wilks Lambda . 50*-IH- --------------------------------------------------------------------- NOTE: The more stereotyped the saleswoman, the higher the score * p < . OS ** p < . 01 *** p < . 001 Multiple Range Test (Scheffe Proced ure ) : Groups Significantly different at . 015 level of significance : Ability: AF < FF, UF AF < UF AF < FF , UF < MF < UF MF < UF MF UF FF < UF Human Relationships: AF < non e FF, UF none MF < UF General Motivation: < AF , UF , FF AF FF, UF AF < UF MF < UF UF Und ifferentia t ed females, FF= Feminine females , MF Mosculi ne females , AF= Androgynous femal e s 176 - ----=---_,..- Effect sizes were similar across samples: . 35, . 35 , .39 . Approximately one third of the va rianc e wa s expla ine d by the sex role type of saleswomen. This provides strong evidence t o rejec t Ho(4) in the case of "selling ability." Based on the review of the literature, it had been pred i cted that saleswomen with strong masculine traits in their s ex r ol e identities (masculine and androgynous females) would be perceive d as conforming least to gender stereotypes about "selling ability." Post hoc analysis shows that in all three samples, both androgynous and masculine females had significantly lower selling abil i t y stereotype scores than did undifferentiated females (AF, MF < UF). In samples one and three, androgynous femal e s a l s o had significantly lower overall stereotype scores than did femin i ne females (AF < FF). While the evidence ?or accepting th e e xpl oratory alternative hypothesis was strong, it was not conclus i ve . ~iuman Relations" Stereotype Scores. Inspection of the Univariate analysis of variance of "human relations" ste r eotype scores reveals statistically significant differences in sample one only? The difference between the means in sample s two and th ree did not a c hi. eve stati.s ti.c a 1 si.g ni.c ance. Comparison of effe c t s i zes revealed an effect of almost twice the size in sample one ( . 2.5 ) than Was seen in either sample two ( . 12) or sample thr ee ( .1 2) indicatino 0 that this phenomeno n 1vas no t an a rti f ac t of s :1.11p le 177 size. The null hypothesis could be rejec ted in the s pecial case of selling to industrial end users, but could not be r ejec t ed in the other two types of selling. Based on a review of the literature, it had be en pr e di cte d that that saleswomen with strong feminine traits and we ak ma scul i ne traits (feminine females) would be most likely to be pe rceive d as conforming to gender stereotypes concerning human rela ti onsh i ps . Post hoc analysis showed that in sample one f e minine f e ma l es did have significantly higher scores tha n did androgynous females , but also had significantly lower scores than did undifferentia t ed females. No differences were found between feminine fema l es .:md any other type females in either samples two or three . evidence, gives no basis for accepting the alterna tive explora t o r y hypothesis. "Motivation" Stereotype Scales. Inspection o f the re s ult s of the univariate analysis of variance of "motivation" s t e r eo t y pe scores shows the presence of statistically significa nt di f f e r e 11ccs among the means in all three samples. Effect sizes a r e l a r ge i n each of the three samples, ranging from . 26, in the case o f sample three, to . 31 in the case of sample one . This pr ovi des s tro ng evidence to reject the null hypothe si s of no di f f e rence a mont.: the means of "motivation" stereotype s c ores. Ba sed on the review of the lite rature , it ha d bee 11 pr edicted 178 ---- ------...:-___, tha t sa l eswome n with st r o ng masculine and weak feminine traits (mas culine f e ma l e s ) would be l east likely to be perceived as conforming to gender stereo t ypes i n "mo t ivation ." Post hoc a na lys is showed that the mea ns of the s t ereo t ype scores of masculine femal e s were significantly lower than tho s e of on ly one group (undifferentia ted f e males). The results do not sup port the pred i cted rel a t i ons hi ps. Hypothes es Six a nd Seven Hypothese s s i x a nd seven concerne d the r ela t ions h i p between s elling effectiveness and sex role t ypes of s aleswomen . Hypo t he s i. s six conce rned the differences among t he mea n s of the over all s ell i.ng effec tiveness scores of saleswome n c l ass i fied acco r din g to s ex r o l e type? Hypothesis seve n e x tend ed t h e a nalysi s t o the relat i nn s hip s a mong the a ve r ages of the individ ua l f unc tio n e f fective ness sco r es and the sex role types of s a l e swomen. Ho (6 ) There will be n o s i g ni f i cant differences amo ng the means of overall selling e f fec tive ness scores of sa l eswome n classified according t o the ir sex ro l e types . 111e r e sults o f the o ne-way a nalys i s of va ria nc e ( ANOVA) .ir e prese nted in Ta ble 5- 8 . St a t istically s ig ni f ican t e f fec t s ~e r e o bserved i n all three samples . The null hypothes i s of no rel a tions hip wa s t he r efo r e r ejec t ed . Effect s i zes were s imilar a mong the s amples rangin g fr om . 24 in s ampl e one to . 2'3 i.n srrn1ple 179 - ---- ---~__,_ Table '.>--8 Test of Hypothesis Six Means of Overall Selling Effectiveness Scores of Saleswomen Classified according to Sex Role Types Sample one Sample two Sample three To Organ'l To I n-Home To End Users Cons umers Resellers ( n = 202) ( n = 72) (n = 63) - - ------------------------------------------------------------------------- UF FF MF AF UF FF MF AF UF FF MF AF ------------------------------------------------------------------------ Means: 6.37 7.22 7.81 8 .36-l!-!HI- 5.69 6.49 7.79 7. 74*- 6.20 6 . 63 7 .70 8 . 37*- Effect sizes: 23.7 27.8 27.l NOTE: The more effective the saleswoman, the higher the score fiHI- Significant at p < . 001 Multiple Range Test (Scheffe Procedure) : Groups Sig nificantly different a t . 05 level of significance: AF > FF, UF AF > FF , UF AF > FF , UF MF > UF MF > UF MF > UF FF > UF FF > UF UF Undifferent iated females, FF= Feminine fe males , MF Masculine females, AF = Androgynous females 180 - -- ______ _JII- two. Approximately one fourth of the variance in overall selling effectiveness was explained by the sex role types of saleswomen in this analysis. The null hypothesis of no relationship was therefore rejected. Based on a review of the literature, it had been predicted that saleswomen with strong masculine and feminine traits in their sex role identities (androgynous females) would excel, overall, in their selling performance. Post hoc analysis of orthogonal contrasts, a Scheffe test with p = . 05 , revealed using statistically significant differences between the me;rn scores of the androgynous group and those of both the feminine and undifferentiated groups in all three samples. TI1is gave some support for the predicted relationships. The means of sa l eswomen classified as androgynous and masculine, however, c!Ld not diff,~r significantly from each other in any sample . Ho(7) There will be no significant differences a mon g the means of each of the individual function effec tiveness sco re s of saleswomen classified according to their sex role types. Table 5-9 shows the results of the one-way multivariate analysis of variance (MANOVA). Multivariate significance was estJhlished by an omnibus test in samples one and three, but not in snmple two. The univ3riate results were then ex::unined. Since seven relationsh Lps were estimated simultaneous ly, an error rnte of . 05 1v:1s s,'t (,Jr 181 - _______ _:,r Table 5-9 Test of Hypotheses Seven Means of Individual Funct.j_on Effectiveness Scores of Saleswomen Classified according to Sex Role Types Sample one Sam pl e two Sample three To Organ'l To In-Home To End Users Cons umers Resellers (n = 184) ( n = 66) (n = 60) --------------------------------------------------------------------- MEANS: Univariate Analysis: UF FF MF AF UF FF MF AF UF FF MF AF Prospecting 5.81 6.68 7.48 7.89*"** 5.1 0 6 . 25 6 . 85 7. 45* Contac t i no 0 6.73 7.75 8.48 8.76*?-ll- 6. 52 7 . 00 8 . 00 8 . 15 6 . 10 7. 09 7 . 82 8 . 71**-ll- Probing 6.10 7.18 7.86 8.32-l!--lHI- 5. 86 6 .33 8 . 00 7. 75** 4 . 95 6 . 36 7 . 82 8 . 12~* Stimulating 5.98 6.95 8.05 7.98**-ll- 5.10 6 . 50 7 .77 7 . 35** 6 . 29 5 . 91 8 . 00 8 . 35** Closing 6.35 6.83 8.00 8.43*"** 5. 52 6 . 50 8 . 31 7 .45** 6 . 05 6 . 55 7 . 27 8 . 35* Retaining 6.85 8 .25 7.57 8.59**-ll- 6 .14 6. 33 7 .46 7 . 70 6 . 24 7 . 09 7.00 8 . 65** Nonselling 6.75 7.83 7.21 8.48*"** 5. 24 6 .17 7.00 7 . 75* 6 .1 0 6 . 82 6 . 91 8 . 35 ---------------------------------------------------------------------------- Multivariate Analysis: Wilks Lambda .63*'** . 61 EFFECT SIZES: (Explained Variance/Total Varian ce ) Univariate Analysis: Prospecting .15 .15 Contacting .19 .1 2 . 31 Probing .20 .1 8 . 38 Stimulating .19 . 25 . 24 Closing .18 . 21 . 14 Retaining .13 .1 0 . 1 7 Non-Selling .12 .1 6 . 12 ---------------------------------------------------------------- Multivariate Analy s is: Explained / Total Variance . 37 . 39 . 56 --------------------------------- ----- --- ----- ------------------------- 182 -- ----~--~ Table 5-9 (cont'd) Test of Hypothesis 7 Relationships between Individual Function Effectiveness Scores and Sex Role Types of Salesvomen Multiple Range Test (Scheff~ Procedure) Groups Significantly different at .007 level of signi ficance: ---------------------------------------------------------------------- Prospecting: AF > UF none MF > UF Contacting: AF > UF none AF > UF MF > UF Probing: AF > UF AF > UF AF '> UF MF > UF MF > UF MF > UF Stimulating: AF > FF, UF AF > UF none MF > FF, UF MF > UF Closing: AF > FF, UF MF > UF none MF > FF, UF Retaining : AF > UF none non e MF > UF Non-Selling: AF > UF none none MF > UF ----------------------------------------------------------------------- UF Undifferentiated females, FF= Femi ni ne females , MF= Masculine females, AF= Androgynous females NOTE: The more effective the saleswoman , the higher the score * p < .05 ** p < .01 -lHH!- p < .007 183 - -------_,,.. each analysis, requiring a p-value of .007 for establishing statistical significance. 5 Inspection of the results of the univariate analysis of variance shows statistically significant differe nces among the means of all the functions in sample one. Only the differences among the means of contacting and probing effectiveness scores achieved significance in sample three, while no differences a mo ng the mean s were sig nificant (at p < .007) in sample t?. . o . The null hypothesis of no difference among means wa s rejected in the case of sample one only. Comparison of the effect sizes ac ros s the three samples suggests that some of the apparent di ffe ren ces among the samples are a n artifact of the relatively small sample sizes in samples two (n = 72) and three (n = 63), r esultin g in low power to detect s mall effects . Based on a review o[ the literature, it had been predicted that saleswomen who had both strong mascu.Line und feminine tr~1 its in their sex role identities (androgynous females) would exce l ::it each of the individual fu nctions. No prediction 1v:1s m,:11le [or non-selling ac tivities. To investi3ate these relntinnsilips further, post-hoc orthogo nal co ntrasts we r e per forr,1ecl u:3 i. rq th ,' Scheffc procedure with p = . 007 . The results ,l!()W some support [or the superiority of androgynous fem::ilcs, but not to : 1 st -1ti.stL c.1ll.y significant extent . 184 The Function of Prospecting. The mean prospec ting effectiveness scores of androgynous females in sample one: (1) exceeded those of undifferentiated females, but (2) did not exceed those of feminine females or masculine females. No significant differences were observed in samples ~~o or three. The Function of Contacting . The mean contacting effectiveness scores of androgynous females: (1) exceeded those of undifferentiated females in both srunples one and three , but (2) dLJ not exceed those of feminine females or masculine females in ,my sample. The Function of Probino for Needs. The mean probin g effectiveness score of androgynous females exceeded those of undifferentiated females in all three samples , giving so11H' SLipport for the superiority of androgynous femal8s in per(orm ..r nce o( this function. The means of androgynous fe:11ciles were not, ho1vevPr, significantly higher than those of either masculine or fomin inP females. The Function of StimuL=iti.ng Desire . The me:111 ,~timuL1ti.n ~-; effectiveness sco res of androgynous fo mal8s : (1) exceeded thoso nr undifferentiated females in sc1mples one and two only, (2) excc>('ded those of feminine females in sample one only, but ( '3) d Ld not differ from those of masculine femcil0s Ln uny snmpl,~. 'l ,> 185 -- --- ---__,... significant diffe rences emerged between groups of saleswomen in mean stimulating effectiveness sco r es in s ample three. The Function of Closing. The mean c losing effectiveness score of androgynous female s : (1) exceeded those of undiffe r entiated females in s amples one and two, (2) exceeded those of feminine female s in sample one only, but ( 3) did not differ from those of masculine females in any s ample . There were no statistica lly significant diffe rences among groups of saleswomen i n sample three. The Function of Retaining. The mean retainin2 effec tive ne ss scores of androgynous females ( 1) exceeded those of undifferentiated females in sample one only , but (2) did not differ significantly from those of masculine or fe min i.n e femalP ,~ i.n rn y s ampl e . No signi ficant di ffe rences emerged be t1,een groups o[ saleswomen in mean retaining effec tiveness scores in e ither sampl 0 two or three. Non-selling Ac tivities . No pred ic tion had been m:::icte for the superiority of a ny grou p of females in non-sel li ng nct i vitiPs . Nonetheless, t he means of androgynous [emales were: (L) s ignificantly d i.fferent from those of und iffe r ent inted fem:::iles in sampl e one only, but (2 ) did not Ji [fe r f r om t hose of masc1 1U11c or feminine females in a ny sample . No stati.sti.cul l y s i g ni. f i.cn111 136 differences were found between groups of saleswomen in samples two or sample three. Additional Data Analysis In order to probe deeper into the relationship between the sex role identities of saleswomen and their selling effectlveness, the relationships between the raw sex rol e masculinity and femini ni ty scores were examined using multivariate regression techni[Jues. The procedure served to clarify the manner in which mascu Line am! feminine traits work together to explain effective performa nce of the individual functions. The r esults of the analysis are given Ln Table 5-10. After multivariate significance wns est,1hli sherl in ?111 three samples by an omnibus F-test, the univ,rinte r gressinn equations were examined. This a nalys i s s h owed that sex r0lc masculinity was statistically significantly related to effectlve performance of all six functions and non-sel I i.ng activities in all three samples. This was not true or sex role femininity. Femininity appe::ired to complement masculinity i.11 the effective performance of " co ntv:.:ti ng ," "probi11~ [or needs," "retaining " and "non-selling activities," but the arnou n l: to whicl1 it contributed t o the explanation of e[fectivc perrorm.:111ce dif"f 'red in the three sample s . 187 . ' . .,,.---- : -~~- -- . -=:..?.". - ?? _:,: ?? - -~?11r ........ .....~ ~~...:;-,~ ... -- ?:-,...u tr.n-::-~~~ ~~\~~--~~~ !"?tt1.?Jr,i:."' ..., ..,'1".7.er.._. ..:?~~\~?~ '''):'.~ --------- Table 5-10 Additional Analysis of Hypothesis Seven Multivariate Regression Analysis of Individual Function Effectiveness Scores vith Masculinity and Femininity Scores of Saleswomen Sample one Sampl e t wo Sample three To Orga n' l To In-Home To End User s Con s ume r s Reselle r s (n = 184 ) ( n = 65) (n = 59 ) Univariate Tes ts: b Beta b Beta b Be t a Prospe cting: Masc .86 .44-ll-ll-ll- . 92 .43*-l!-l!- Femn .31 .16* .44 . 15 Contacting : Masc .88 . 53*,H!- ? 77 . 44*** . 86 . 58*ln(- Femn . 32 .19-lHHE- . 32 . 13 . 44 . 2s~:~-:~ Probin g : Masc .96 . 56*** .99 . 54*** . 99 . 62- Femn .31 .1s - .18 .06 . 37 . 21* Stimulating : Masc 1.04 .58*-lHI- . 95 . 53*** 1.1 5 .70*** Femn .19 .11 . 35 . 14* . 17 ? 1,),1 Closing : Masc 1. 08 . 59*-lK- 1.13 . 60*-lK- 1. 00 . 53*** Femn .19 .10 - .1 6 - . 06 . 01 . 00 Retaini ng : Masc .50 .29*** . 75 . 41*** . 64 . 36- Femn . 68 . 39?--lH- . 08 . 03 . 69 . 36~,H; Non-Selling : Masc .34 ? 19*-''t-* . 98 . 46*** ? 72 . 24',HI?* Femn . 63 . 35*-H--l!- . 75 . 25* . 58 .1 6?:l- -------------------------------------------------------------------------- Adjusted R-Square : Prospecting . 23*-lHI- . 20*-l!-l!- Contacting . 33**?~ . 20- . 51*-lHI- Prob i ng . 35*** .29*?ll- . 50*** Stimulating . 35*-l!--l(- . 30*-ll-* . 53-lHH~ Closi n g .37*** . 33*** . 26*lH> Re taining . 25-)(-l'c* .14- . 3 ') ~~is~ ~ Non-Selling . 16*** .27*?ll- . 25-'i-** Multiva riate Tests : Wilkes Lambda : . 42*** . 40*-lH> * p , . 05 lK- p . 01 *.;H} p . 07 1 'l3 - - ---~--- ---- Sample One. In sample one , the f e mini ne compo nent of sex r ol e identity contributed to effective performance of the funct i.o ns of contacting and probing for needs, but not to as grea t a n extent as the masculine component. Femininity seeme d t o be of s lightly greater importance than masculinity in explaining the e f fe c tive performance of the function of retaining and non-selli.ng activities. Sample t?. . o. In sample two, the f e minine compo ne n t o f s ex r ol e identity does not appear to complement the masc uline compo ne n t i.n any function. Sample three . The relationships in s ample three we r e s i_milc1r to those in sample one. The f eminine c ompo ne n t of sex ro l P identity contributed significantly to the e EE ect ive pe r fo r mc1 11co ')l the functions of contacting and probing fo r :1eeris , but not t,l : 1s great an extend as the masculine c omponent. In pe r [o r mi. ng t he function of retaining and non-selling ac tivities , howev~r, femininity and masculinity appe a r t o be nppr oxi.ma t e l y eq u,11 i_ n importance. Tests for Control Variables The relationships between ove r a ll 3end e r stc r,;otypc sco r, 's ind overall selling e ffe c tiv e ness sco r es o( solcswu me n ?rnd ,1 se t- of variable s introduced f or c ontrol purp oses 11.:i. s testerl . <\ mode l compa r ison a pproach was used (e.g., Gr een and Carroll 1978) . Using f i r s t the ove r a ll stereot ype scores and then the ove rall effectiveness s cores a s depend ent va r iables, a basic model of the rela tionship with s ex r ol e t ypes of sale swomen was co ns tructed. 6 Then each c ontrol va riable was int roduced separately i nto the model a nd i ts e f f ects eva lua ted. Si x models were eval ua t ed f o r each dependent va riable: (I) the basic model wi t hou t c on trol variables , ( 2 ) the basic model controlling f or age , (3) the basic morl e l controlling for education, ( 4) the basic model c on trolling f or marital status, (5 ) the basi c mode l co ntrol l ing ?or tenure on t he sales force, a nd (6 ) the basic model con t r ol l i ng fo r the s e x ra tio of t he c ompany. Ea ch control va ria ble 1vas eval ua t ed to de te rmine whether it: (1) clarifi ed the rela t ionship betwee !1 s ex rol e ty pes a nd the depe ndent va ria ble, (2) enhanced t he expLrna t or y powe r of the model, o r (3) offe red a n a l te r nati ve expl.1na ti.on of t he va riability o f the depend e nt va r iable . Overall S tereotype Scor es . Ta ble 5- 11 i ll us t r ates that the addition of none of the control va r iables : (1) altered th e b::i s i c rel a tions hips between the over all s t e r eo t ype sco r es a nd th e sex r ol e t y pes of saleswomen, (2) enhanc ed th e e xpl a na t o r y po we r of th 0 mod el, or (3) offered an a lternative explanation for Lhe va ria bility in ove r all s t ereotype s co r es . Ove r all Selling Effec tiveness Sco r es . fob J c '5 - 12 i.l]ust r ;-1tes 190 - -----...:!""'_,,,,,... Table 5-11 Alternative Models of Hypothesis Four Overall Stereotype Scores of Saleswomen Classified According to Sex Role Types With and Without Control Variables Model 1: Model 2: Model 3: Model 4: Model 5 : Model 6 : No Marital Tenure Sex Ra tio Covariate Age Education Status on J ob on J ob ------------------------------------------------------------------------ Sample one: To Organ'l End Users (n = 202) (n = 202) (n = 201 ) ( n = 202 ) ( n = 199 ) ( n = 194) -------- R-Square(a) .371*** .37JH?* .375?-IHE- . 376ll+lf- . 373-lHHE- . 368*-H- R-Square change(b) .000 .004 .oos .002 ------ . 003 - ----------------------------------- - --------------------- Sample two: To In-Home Consumers (n = 72) (n = 72) (n = 67) ( n = 72) ( n = 69) ( n = 72) R-Square .282*** .308*-ll-lf- .276*** . 298H--lf- . 310*** . 300*"!?-l(- R-Square change .026 .006 . 016 . 028 . 018 ---------------------------------------------------------------------- Sample three: To Resellers (n = 63) ( n = 63 ) (n = 63 ) ( n = 63) ( n = 63) (n = 63) ------- R-Square .317**-ll- .340*** . 350-lHH- . 318*** . 330*** . 325?~** R-Square change .023 . 033 . 001 . 013 .. 008 ------------------------------------------------------------------------- Model 1: Basic model, regression a ppr oximation of ana l ys i s of va r iance Model 2-6: Basic model, with addition of control variable (a) Unadjusted R-Square for Model (b) Change in R-Square after addition of Control Va r iable p < . 001 ----....-.---- Table 5-12 Alternative Models for Hypothesis Six Overall Selling Effectiveness Scores of Saleswomen Classified According to Sex Role Types With and Without Control Variables Model 1: Model 2: Model 3: Model 4: Model 5: Model 6: No Marital Tenure Sex Ratio Covariate Age Education Status on Job on Job ------------------------------------------------------------------- Sample one: To Organ'l End Users (n = 202) (n = 202) (n = 201) (n = 202) (n = 199) (n = 194 R-Square .237-lf--H- .240-lf--H- . 241*- . 242*-l!-l(- .250-lHHt- .250*-lf-* R-Square change .003 .004 .005 .013 .01 3 - ---------------------------------------------------------------------- Sample two: To In-Home Consumers (n = 72) (n = 72) (n = 67) ( n = 72) (n = 69) ( n = 72) R-Square .278-ll---lH- .292*"*-lf- .272*"*"* .281*"*-lf- .298-lt-H . 327?<-lf-* R-Square change .014 - .006 .003 . 020 .049 --------------------------------------------------------------------- Sample three: To Resellers (n = 63) ( n = 63) (n = 63) (n = 63) (n = 63) ( n = 63) R-Square .271*- .286H?* .280H?-lt .284*-ll-* .275*-ll-* . 294*-lf-* R-Square change .015 .010 .013 .004 . 0:23 ----------------------------------- ---- ---------------------------------- Model 1: Basic model, regression approximation of analysis of variance Model 2-6: Basic Model, with addition of control variable (a) Unadjusted R-Square for Model (b) Change in R-Square after addi tion of Control Variable *"lHf p < . 001 19 :..:'. - ------------ that, as with the overall stereotype scores, none o f the c ontro l variables ( l) altered the basic relationship be twe e n the o ve r a L1 selling effectiveness scores and the sex role types o f s aleswome n, (2) enhanced the explanatory power of the model, or (3) offe r e d an alternative explanation for the variability in overall s elling effectiveness scores. Tests for Bias Tests were performed for sex bias and for other form s of re s ponse bias: Tests for Sex Blas . Three variables were ex.::unin e d t o det e rmin e whether there was evidence of sex bias in the respons e s : ( L) the relative degree of comfort in s u pervising saleswomen, ( 2) th e sex of the sale s manager, and the (3) sex ratio o f the s alesforce . Table 5-1 3 shows no statistically significant di f f e r e nc e s i n responses in ei ther samples two or three , but some i ndica ti on o f possible sex bias appeared i n sample one: (l) male s al es man nge r? s reported perceiving saleswomen in more s ter eotypic terms than cl i. d females sales managers, (2) sales manage r s who f e lt r e L1tive Ly " uncom f or ta bl e " supe r vising s aleswomen r e port e d sa 1 eswomcn as l) e i.n g significantly less effective, and ( 3 ) those s a les m.inngers 1,1 hn s upervised fewer than 401 females, :1lso rep ort e d the Li- sa ] cs1v0rncn to be signi.ficantly les s effec tive . 193 --------- Table S-13 Tests for Sex Bias Means of Overall Stereotype Scores and Overall Selling Effectiveness Scores of Saleswomen Classified according to: Sales Managers' Degree of Comfort in Supervising Saleswomen, Sales Managers' Sex, and Sex Ratio of the Salesforce Degree of Comfort Sex of Sex Ratio in Supervising Sales Ma nager of Salesfor ce --------------------------------------------------- Prefers No Prefers Means: Males Dif Females Male Female ( 40% > 40% Sample one To Organ'l End Users: (n=l2) ( n=l48) ( n=37) ( n=l49 )( n=53) ( n=69 ) ( n=l24 ) Overall Stereotypes (a ) 3.51 2.99 2 . 82 3 . 10 2 . 72* 3 . 15 2 . 89 Overall Selling E. (b) 6.21 7.47 7 . 66* 7 . 40 7 . 54 7 . l 7 7 . D6* ------------------------------------ ---- ---------- ------------- ----- - Sample two To In-Home Consumers: ( n=lO) ( n=60) ( n=2) ( n=66) ( n=6 ) ( n=4 3) ( n= '2 9 ) ---------------------------------------- - - ------------------- ---- Overall Stereotypes 3 . 43 3 . 20 2 . 33 3 . 27 2.54 3 . 25 3 . 16 Overall Selling E. 7.06 6. 76 8 . 28 6 . 75 7 . 95 6 . 55 ---- 7. 29 --------------------------------------------------- ------ - --------- --- Sample three To Resellers: ( n=6) ( n=4 7) ( n= lO) ( n=52 ) ( n=ll) ( n=32) ( n='29) ----- Overall Stereotypes 2.90 2 . 91 2. 72 2 . 85 3 . 01 2 . 90 2 . 84 Overall Selling E. 7.34 6.94 8 . 15 7 . 19 7 . 10 7 . 09 7 . 28 --------------------- ----------- -- - ----------------- --- ----- ------- ------- (a) The more ste reo t ype d the saleswoma n , the higher the sco r e (b) The more e ffective th e sa l eswoman, the highe r the s cor e * Si gni f i ca nt at , . 05 -------- Te sts fo r Resp onse Bias . Two va riahles were inc lud e d Ln the ques t ionna i re fo r the pur pose of evalua ting r e spo nse bi as . These we r e self-re po r ts of how we ll-qua l ified sale s ma nagers f e l t to eva lua t e the s aleswome n, a nd how l enien t t hey f elt they had he en i n their e va lua tions . It was no t po s s i ble to eva lua t e the bia s a ttributa ble to leniency, since nearl y all s a les ma na ge r s r e porte d themselves as being neither l e ni e nt nor hars h. Table 5-1 4 s hows no sig nifican t diffe rences in evalua tion s of s ales ma nD!3<: r s i n sample ,c:; one o r two, but significant differences were f o und in sn:npl c three. In s ample three, those sales ma nag e r s who r e port ed the mse l ves t o be ve ry well-qua lifie d to e va luute saleswome n gav~ saleswomen s i g ni f icantly l ower o ve r a ll stere otype s co r e s ,rnd h i g her overall s elling e ffe c t i veness scores than did the o the r s Jles ma nagers. Discussion of the General Findings The fo llowin g section d i scus ses t he findin gs i n th e co n text o r t h t~ fo ur broa d ques t ions t hat g uid e d the r ese:1r c h . Resear c h Question One The f irst r esea r c h q uestion addressed t he ques t ion of 1vheth<' r sal eswome n ac tm1 l ly beha ve i n s t e reo ty pi. c 1.ay s wh Ll e pe r fo r 11 i ng : 1?, 195 Table 5-14 Tests for Response Bias Means of Overal.l Stereotype Seal.es and Overal.l Selling Effectiveness &ores of Sal.eswomen classified according to Sales Managers' Degree of Qualification to Rate Saleswomen Sample one Sample two Sample thr ee To Organ'l To I n-Home To --------- End Users Cons umer s Reselle r s --------- Not No t Not Mean s : Very Ver y Ver y Very Very Qual. Qual. Qual. Ver v Qual. ;;------------ Qua l. Qua i. era11 Stereotype ----------- Scores (a ) 3 . 03 2. 99 3 . 41 3 . 11 Overall Selling 3 . 31 2. 66* --E-ff-ec-ti-v-en-e-ss- Scores ( b) 7 . 18 7.53 6. 31 7. 11 6 . 09 ----------- 7. 71 * JH.------- - (a ) The more s tereot yped t he saleswoman, the higher the scor e ( b) The more effective the sal eswoman, the higher th e scor e * p < . OS -!Hf p < .01 -JHHt p < . 001 196 salespersons. Are saleswomen in traditionally male-oriented areas of selling perceived by their sales managers to behave in accordance with gender stereotypes impacting on their selling ability, human relationships, and motivation? The answer to this question is that some gender stereotypic behavior is perceived to take place in the marketplace, although not as much as critics of saleswomen have contended. The findings suggest that many saleswomen may act in stereotypic ways from time to time, but that the behavior is sporadic . It does not occur frequently, in most cases, or on a regular basis. ~arch Question Two The second research question probed the relationship between Perceived stereotypic behavior and effective selling performance by saleswomen. Does the denree to which saleswomen in traditionally male-orient;d areas of selling are perce~ved by their S a 1 es managers to be1?1 ave in accordance with gender ? ? stereotypes correlate with the effectiveness of thetr selling performance? The results suggest that, under certain c_i.rc umst.rn ces, when saleswomen act in stereotypic ways selling effectiveness is reduced. Saleswomen who were Perceived as behavi ng Dcco t-.i in :~2? to ,, selling stereotypes were further perce ived ability" as l ess 197 - ------_.,-...-__,,,.- effective than others in their performance of all of t~e functions of selling . Saleswomen who w-ere perceived to behave according to "human relations" stereotypes were also perceived as less effective than others in " closing," but this effect did not occur in the case of saleswomen who sell to organizational encl users . Saleswomen who behaved according to "motivation" stereotypes were perceived ;:is less ef fective than others in performin g two functions : prospecting and retaining. "Selling Ability" Stereotypes. One measure of t he valid i tv of a study is whethe r or not the results support c ommon se:1se . In this case, common sense would say that women who were perce ived as weak in bas i c "selling ability" (e . g., product :rnowlerlgc>, pe rsuas iveness , and a daptive behavior) would .:1lso be perce ived to sco r e r e lati vel y low o n performance of the sel ling fu nctions (e . ~ ., prospecting, co ntac ting, and closing). This study found thnt generally this was true, especially when selli ng to o r :;.:ini~n t i l)J1 ;1l end users. In direc t selling, however, moti.v=ition se-.rncd tt1 overcome limitations in selling ability among women when pcr[orrning pros pec ting activit ies . "Human Relations" Stereotvpes. Saleswomen who wen~ perceived as weak in " human relations " ability (e . g., over-soci ,1li.:1)d with c us tome r s , w-e r e dominated by customers, could no t rejections) also were pe r ceived as bein g r el;it in~ly les,; e [Pct i vc 198 - - --...- ....z-...----- at closing sales than were the others. Thi s effect a ppeared a mong saleswomen who sold to resell e rs and in-home c onsume r s , bu t not among saleswomen who sold to organiza tiona l end us ers whe r e c l os ing occurs less frequently. Why is this s o? It may be tha t sa l eswome n who over-emphasize the social aspects of the job a l so ten d to over-value their friendships with c ustomer s . The y may hesi t ate t o place demands on customers that mi g ht jeopa rdize these relationships. Furthermore, those sale swomen who allow t hei r customers to dominate the interview may never have the op po r t uni ty to close and those saleswomen who tend to acc ept their c us t ome r' s excuses and objections at face value may give up too eas i ly. Those saleswomen who react too emotionally wh e n they l ose sc1 l es , m;) y avoid asking for the order or may react emoti o na lly whe n the Lr closing attempt fails. For a ll these r easo ns , wh e n a st1-0 11 g c lose is important to the sale, as when m.xdn g on e - c al L sales to i n-home consumers, saleswomen who act in accordance wi t h 11 hur:1,:1 n r el:iti.on s " stereotypes are at a serious disadvnntage . In s i.tua ti. ons when::- t he close is the logical outcome of a well-s tructured se lling pr oc Pss , such as in inrlustri.al sales, this t ype o f be ha vio r does not scr ' :,J t ,J be dysfunc tional. "Motivation" S tereotypes . Saleswomen who we r e pcrc e i v,,d , i s weak in "motivation" (e. g ., lac k e d confi d ence , r equir ed co 11 s t 111:- encouragement, ?r e f e rred pl easant aspe c ts o[ the job, 199 ----~---- preoccupied with family responsibilities, or simply lacked dynamic inner drive), were also perceived t o be relatively ineffective in "prospecting" and, to a lesser extent, in "retaining." The research showed that low scores in prospect Ln3 correlate ::l highly with insufficient "motivation" o n the part of salcswo;nen. Why may this happen? 1t may be that those saleswomen 11ho are insecure, who require a good deal of reinforcement from their manage rs, or who tend to avoid activities that they consicler to he unpleasant, procrastinate about prospecting. Those sa l cswor:ien who have time constraints because of conflicts he tween L1mi l v ,rncl professional responsibilities, may also fall into the tr;:ip of "put ti? ng off" prospec t i? ng ac t i? vi? t i? es . 1-Jhen heavy prospecti n3 is required on the sales job, insufficient motiv-it Lon see,ns t ,> be a serious liability. Salespersons need a high level of internal drive to sustain prospec ting activities over time. Scores on the retaining f unction also ,? ,Jrrelate wLth insufficient motivation. 1-Jhy may thi.s be so'.' The fun ct i.on of retaining can be critically importa nt to the est;1blLshrnent , f long-term relationshLps i n selling, but the duy-to-day t:isk:, involved may not always be immediately ta nglb lv rew:1rded. Salespersons who are not strongly self-,n()ti.v .1ted m.::iy not be wi. l I in:~ to expend the extra time and effort that retentLon nct: ivi.ti.l~s requin~. ~00 - -- __ .-;__ ___ _ Research Question Three The third research question probed i nto the question of whether saleswomen who behave in ways that a r e co nsistent with these gen der stereotypes can be identified by classifying them into sex role types. Does the degree to which saleswomen in tradi tionally male-oriented a reas of selling are perceived by their sales manao e rs t o behave in accordance wi th gend e r stereotype; differ among saleswomen of four diffe rent sex role types? The answer is yes, with qualifications. "Selling Ability" and "Motivation" Stereotypes . The st11 rl y strongly s ugges t s that sale swomen 11ho are perceived as :1ct Ln g according to "selling a bility " a nd " ,notivntlon" ste r eo types , c.111 be identified by knowledge of their sex ro l e ty pe in ~ll three sc l li n g s ituations. Those sa l eswo me n who have strong masc11line tr:1 its (masculine and ndrogynou s fe males) , wer e desc ribe d Ln s i gnificantly l ess stereotypic terms by their s..1lcs rn ,111ag~rs thn n were the others. "Human Relatio n s " S tereo t vpes . IdentifyLng salcswome,1 1vho nre perceive d as acting accordi ng to " human relnt Lnns" s tcr-erJt y pes o n the basis of sex role ty pes is less s tr3ightf ,Jnmrd . Th e m:111n1'r in which gend e r s tereotypic behav Lo r in " humnn reL1r Lnn,; " reL1tr> s l:.J '.201 ----- --- the sex rol e type of saleswomen seems to be highly dependent upon the selling situation. While saleswomen with strong masculine traits were described as being less stereotyped than others by their sales managers in all three selling situations, those with strong feminine traits were described in ctifferent ways in the three different samples: (1) in the case of saleswomen who sold to organizational end users, saleswomen with strong feminine traits were desc.ri bed as being l ess stereotyped than others , ( 2) in the case o f saleswomen who sold to in-home cons umers , sales\vomen with these same strong feminine traits were desc.ribect as be i 11g ~ stereotyped than others, a nd (3) in the case of saleswomen who sold to resellers, there seemed to be ve r y little relationship between feminine t rai ts and the descriptions of stereotyp i c. behavior . WLth regard to stereotypic. behavior in " human re lat ic)ns," the femini .ne traits of scileswomen seem to be assets when selling to organizationa l e nd users , liAbilitics when selling to in home con s umers, and immaterial when selling to resellers . Research Question Four The fo urth r esearch question asks whether it i s pos ,;Lbl,, t:,) ident ify the most effective s:.1leswomcn hy knowled'.~c of th,'Lr s c!, role type . 202 - ---------- Does the effectiveness of selling performance in traditionally male-oriented areas of selling differ among saleswomen of four di fferent sex role types? The answer seems to be, yes . The study sugges t s that effecti.ve performance of the sales job relates to the sex role type of saleswomen. Saleswomen who were c lassi fied as androgynous or masculine outperformed those classified as feminine or undifferentiated in overall selling effectiveness in nll three selling situations. These findings support the applicability ,Jf the two-dimensional model of selling effectiveness to the situAti.on of selling to customers who are organizational end users and of sel l ing to customers who are resellers. Examining the way that masculine and femi.ni ne traits rel ?1te to the performance of each of the functions of selling seems ::o be more helpful in understand ing selling effectiveness than focusin g on the performance of saleswomen classified according to sex role types . Saleswomen' s masculine "instrumentality" ;:ind femLnine "expressiveness" both contributed to theic effective per[or111~111r: ," of the individual functions, but did it in Ji[ferent w.'.ly s . In .i/ l three snmples, sa leswomen with stro ng mascllline tra.i.t ::; excel !Pd in performance of all of the funct.i.ons of selling . r<)l c masculinity , withoul henefit of femininity , reL1ted to cff,,ct i ve perform,::rnce of the functions of prospec ting , stimulating drsLn', :203 and closing. Both sex role masc ulinity and femininity contribu ted to effective performance of the fun c tions of contac ting , !)r obing for needs, and retaining, but this was s een in two of the thr e e types of selling situations only: selling to organiza tional users and selling to in-home consumers. It was not true in the in the special case of selling to in-home consumers . Prospecting, Stimulating Desire, and Closing. Effe ctive performance of the functions of prospe cting, s timul.1tin3 desire , and closing was associated with the presence of strong masc ul i_nc "instrumentality" in saleswomen. lvhy may this be s o? Tt :na y be that when prospecting, the pre sence of strong masculine trnlt s , helps saleswomen sustain the necessary perseveran c e t o c ont i nuo us l y seek out new customers for the firms product. \v'hen pe r fo rmi ng t he function of stimulating desire , masc1 1l ine tra i ts , s uch ns forcefulne s s and assertivene s s, may help saleswome n del ive r t he i. r persuasive messages. When performin3 t he f unction o f c l os in g , masculine traits, such as domin;:ince and l e ader s hip abi.li ty , m:Jy help saleswomen control the interview, ask f or the o rd e r, n ncl successfully obtain it. While f eminine tr:1 i.t s do no t sc1=m to c ontribute to saleswomen's e ffe ct i ve performa nce 0 1 t he!,<' t hr ee functions, their pre senc e doe s no t seen to de t r a ct fro m i t . Contacting, Probing f o r Need s , and Re t:::i i n i_n_R. Effoc t i_ve performance of cont,1c ti.n 3 , rrobin g fo r needs , a nd r e t :1in i n~ H .i s associated with the presence of both s trong masculine " i.n strumentality" and strong feminine "expressiveness ." Why may this be so? It may be that when "contacting," saleswomen's masculine qualities, such as assertiveness and willingness to take risks, give them the drive to overcome any reluc tance they may have about initiating new relationships; their feminine traits , s uc h ? s warmth a nd understanding, probably help in the process of establishing rapport with customers. When "probing fo r needs, " saleswomen's masculine qu3llties, s uch as their l eader ship abil i ty, may help them to direct the progress of the conversation; their fe minine qualities, such as sensitivity to the needs of others , may make them more responsive to both the verbal and no n-ve rb? l communicr1tions of customers . When "retaining ," saleswomen ' s masculine qualities, may help them ma i ntain the t e nac Lty .:.111<1 the perseverance to perform necessa r y ser-,ices ; t heir [ ,2mininc qualities, such as sympathy a nd gentleness, may ennble them to perform these services in a way that is sa ti sfying t o the c ustomer, and to maintain a sensitivity to the c hnngi ng needs of c ustornt~rs throughout the duration of the relations hip. :20.5 - - - --------- Discussion of the Findings in the Three Individual Samples There a re differences in the findings in the diffe rent samples . This section discusses the findings as they pertain to the three different types of selling si tu a tions . Sample One: Selling to Organiza tional End Users This section discusses the findings that are s pecific to sit? 1c1 tinns in whic h customers a r e organizational buyers who arp ma king decisions about products to be used i n the co nduct of their bus iness or to a u gment the ir company' s product. Stereotypic Behavior a nd Sell in~ Effect i v<' ness . Tn ,his selling situation, saleswomen who were pe r ce i VL! ,I :-is c o n fn nni.n ::; to "selling ability" s t ereotypes we r e Glso perc eived :is l ess e[(e tive in performing all of the fu nc tions of selling than 1-1c r e otl lt'r s . Saleswomen who were perce ive d to co nfo r :n to "humnn rel ,1t i. ,111 ..., 11 stereotypes did not show ineffective pe r( o rm:rnce o( ~ l)f t '1 ,~ functions o[ s elling. Par tic ularly notc1ble 1ms the Lw t t ln t , i.n contrast to the findings Ln the o the r t?.-10 sampl L' s , sGleswomen 1,ho a ppear ed to conf orm t o "h uman rel,tions " sten'utypcs Wt'l-L' n oL ~06 perceived as less effective in performing the f unction of closing. It is possible that the consultative nature of this selling situation renders the "close" less challenging to salespersons than it is in other forms of selling. In this form of selling, the "close" may come close to being the ideal case where it serves as n natural outgrowth of a problem solving process of matching product benefits to consumers needs, rather than as the c limax of a one-shot sale. As in the other samples , saleswomen who showed stereotypic behavior in "motivation" showed reduced er f ec ti vene ss of performing both the functions of prospecting and retaining. Stereotypic Behavior and Sex Role Type. The sex role type of saleswomen was strongly related to the degree to which soleswornen ' s behavior appeared to conform to stereotypes abo11t "se l.ling ability, " "human relations, " .-ind "motivation " i.n th Ls sel lin g situation. Those saleswomen with strong masc11J.ine and fo rni.nine traits (androgynou s females), were perceived to be less stereotyped in their behavior than others. Selling Effec tiveness and Sex Role Types. The sex role types of saleswomen were also strongly relnted to effectLve per o r111.111c<' of a.Ll of the fun c tions. ,\nJrogynous females nppc.ired t o be 1,Hirc' effective than the others. Although there wcis littl e di[[Prenc,' between the effectiveness of nndrogy nou s and mDsc11 LL 11, [0rn,1J 0s, s trength in ma sculi ne traits was posit Lvely c1s so,?i?1tcd 11ith :2()7 ---- - - --:--=---- effective performance of all of the functions of selling and of non-selling a c tivit ies . Strength in both masculine and feminine traits was positively related to saleswomen's effectiveness in performing the functions of contacting, probing for needs, and retaining. Sex Bias. Despite the fact that the sal es managers in this group seemed very supportive of the women on their staffs, a possible problem with sex bias was identifie r:!. "!ale sales munagr~r-s stereotyped women more than did female sale. managers. In addition, those sales managers who fe lt less comfortable supervising women than men a nd those who supe r vised fewer than !,O':: women r a ted the overall performa nce of the saleswomen under their s upervision significantly lower than did others. Why may this be') It seems only natural that male sales managers would think o[ saleswomen in more s t ereotyped terms than would female sales managers. Those sales managers who felt less "comfor-table " supervising women probab l y have dHficulty ev:i lu3 ting the Lr saleswomen a nd some bias is probably present in theLr <;s ,)f probing for need s in thL s type of se lling . :2l3 ------~- Sample Three: Selling to Resell e rs This section discusses the findings as they pertain to the situation in which customers are organizational decision makers who are making decisions about products that they will purchase for profitable resale to others outside the organization . Stereotypic Behavior and Sex Role Type. As in the other two types of selling, saleswomen who were perceived to conform to gender stereotypes were also perceived to be l ess effec tive Ln performing all of the functions of selling , although the relationship between "selling ability" scores c losLnz effectivenes scores did not achieve signi fica nce. Saleswomen vrho were perceived to beha-,?c nccorrl ing to "hum,rn relations" stereotypes were .:J1so percei ved to be les.- 1-2 ..; ; i, effective at performing the functio,1 of stimulating desire. \v'hy may this be so? It may be that in this type of sellin:s, c uslorn f t h L; relatio nship ,1mong the three samp l es c m b<~ o nl y te nt ;1ti.v ,, . S . Several special limit.1tLons pe rta Ln to thL::; .rn :1ly . L:~ . Th(' SPSSx :v!ANOVA progr ,:un that 1vas used c ould not :, c c ommnr!.1t ,, 111L ~ sin ~; 218 ---?--- data. The results are therefore generalizable only to saleswomen who perform all of the functions on their jobs. Since so few saleswomen performed the function of prospec ting in sample three , "prospecting" was not included in the analysis. Therefore , i.n the case of sample three, the results are generallz:1ble to sale s women who perform all of the functions, exclusive of prospesting. 8. A basic analysis of variance model was developed using regression techniques with effect coding. The techni.que is mathematically equivalent to univariate analysis of v.1.ri.ance. 1n the basic model, the four sex role types were represented hy three zero-one dummary variables. Control va riables wer e sp l i. t at the median and also coded into 0-1 dummay variables. CHAPTER SIX SUMMARY, CONCWSIONS, IMPLICATIONS AND RF.COMMENDATIONS The objective of this study was to investigate whether saleswomen were perceived by their sales managers as behaving in accordance with gender stereotypes and, if they were, to determine whether the behavior related to their selling effectiveness. It further souBht to determine whether saleswomen, classified into sex role types, differed in the degree to which they were perceived to conform to gender stereotypes, or the degree to which they were perceived to be effective in their performance of the functions of selling . Summary of the Study In order to provide a basis for examining the job-related behavior of saleswomen, t he relevant literature was reviewed. Based on this review, a simple mod el was developed a nd hypot~cscs w re f,>rmubterl for testing. Background of the Research The literature reviewed three major conceptu.,l ,ireus: :J-::!nder no - -------- stereotypes, selling effectiveness, and sex role identity. Gender Stereotypes. The literature concerning gender stereotypes defined the nature of gender stereotypes: the damaging effects of stereotypes about women in business, the co nten t of gender stereotypes, and the issue of the stability of these stereotypes over time. It established a context for the research through a discussion of bias in existing resenrch. Ster eotypes about the behavior of women in selling wer e discussecl in three major areas: "selling ability," "human relati.ons," a nd "motivation. " Selling Effectiveness. The literature concern.1? 11 ,--, se 11 0 1? .ng effectiveness reviewed issues in the measurement of selli ng effectiveness and justified the use of the f unctions o[ se l Linu ,1s the basis for a behavioral measure of effectiveness . The salesmanshi.p literature was r eviewed to estnhlish criteria for good performance of the functions of selling . Effective selling was discussed in terms of a two-dimensional mode l of se l Ling effectiveness, consisting of "instrumental " ;rnd "exp r ess iv-=" dimensions . Sex Role Id entity. The literature co ncerning sex role identity discu ssed the a ppropriateness of t he use of the l3em Sex !folc Inventory as a bnsis for understnnding sex reL1tcd bch.:1vior, ,.lilt! the a ppli cAhili ty of the concept of sex role! iJen tity: (l) to 221 - --~----- behavior in laboratory studies, (2) to behavior in occupations, -'.lnd (3) to behavior of salespersons. A Conceptual Model. These three clusters of liter:::iture were brought together to develop a correlati.onal model of the relationships of interest, which was used to guide the subsequent research. Methodoloqy A survey was conducted of sales managers who supervised saleswomen in three different types of selling situations . Respondin g sc1les managers described from one to four S3leswomen und e r t he .i.r supervision, on items pertaining to: (1) their degree of conformity to gender stereotypes, (2) their effectiveness in performi ng s ix functions of selling and non-selling activities, 3nd (1) thei.r sex role identities. Sales manalje rs also evalu3ted the L'llportance nf each of six functions of selling and no n-selling r1 ct ivitlcs to good sal esmanship on t he i r particular job . Sales manal._\crs prov l ,led demographic info rmatio n about the saleswomen they desc r i.be,i a nd about themselves. Gender S tereotypes . items r e lnti.ng to s ;:ile swomcn ' .;; conformity to gender s t erotypic behavior were cornbin,~cl into three scales representing stereotypes about "se l Ling ohillty ," "' I :Jc!, 111 relnti.ons," and "motivution ." Saleswomen recei.vcJ ?JC- ?I -----~'"'=---- masculine traits (masculine and androgyn ous female s) were leas t likely to appear to behave according to gender s t e r eo t ypes in "human relations." In the special case of s elling to in-home consumers, feminine traits were associated with a dy s functi onnl pattern of interpersonal relations that detracted from s elling effectiveness, but the effect did not occur when s e lling to business customers. The sex role types of saleswomen related to the ef fec ti veness of selling performance. Saleswomen with s trong masculine nnd feminine traits (androgynous females) were perceive d t o outpe r fo r m others in the functions of contacting, probing f or needs , und retaining. Saleswomen with strong mas culine tra its (ma sc uline a nd androgynous females), seemed to outperfo rm othe r s in pe rformin g t he functions of prospecting, s timulating desire, c l os in g . Soleswomen with strong feminine traLt s wer e nt n di.sodv;rnt.i gc wh e n selling to Ln- home consumers . Implications The findings of this stud y ha ve both co nceptual ;:i nd prn c L i c.1 L implications, as discu s sed be l ow: _. __ ,___?~--- Theoretical Implications While the purpose of thi s research was not to test theory, the findings have implications for increasing our und e rstanding of: (l) the relationship between sex role identity ~nd sex-typed be havior, (2) the relationship between sex role identity and occupational behavior, and (3) the two-dimensional model of selling effectiveness. Sex- Typed Behavior. The conceptual framework for the stud y wns suggested by a debate in the literature co ncerning the nature of the relationship betwee n sex rol e identity ( ?s mensured hy the BSRI) and behavior. Proponents of gend e r schema theory (e . g ., Bern , 1977; 1979 ; 198 1) argue tha t the prediction s aboul gend,:,r stereotypic behavior of sex-typed individuals can be made [lboiir the entire spec trum o E gender-appr opriate beha vior. Those th.1 t hold 11 mo r e co nserv,:;i t i ve i nterpretation (e . g ., Spence and llelmrciclt 1978) c ont. e nd that suc h prcdic ions , pply tn i.n. ' L!'um ntu1 IHI ('.'f)Ffl. ' 8 iv, b h vlors on y , TI1 s h ma h ry xp1 n ton . WUUl .r\ Wl1 W I [ll l"t I ,] ll feminine - typed by t:h ' iC' l!IIJ p ' r: L ? u I ~ behaving i n acco rdance with gender sLer-co y p '1-> ? ( ll11t.1 i I ( 'l'l ' I \ l L1 l \d female s emerged as t h e more stereotyped group ) ? was seen, h owever, between perceived qu.Jlity of perform.Jncc o( thl' instrument:::il a n d expressLve ,spects of the sa l es jnh c11Hl tlw masculine and femini ne traits of saleswomen. Saleswomen were perceived as performing those behaviors best that were congruent with their sex role types. Vocational Behavior. The findings of this study confirm that the concept o.f sex role identity relates to occ upational behavior in the area of personal selling. Both masculine and feminine traits were .found to be related to effective performance of the functions of selling. Personal selling may represent !'l 11 niquc opportunity for androgynous females to excel. The Two-Dimensional Model. The findings provide some Slip port for the two-dimensional model of selling effect i.vene ss (Blake .'.lnd Mouton 1980; Jolson 1984) . In the case o.f three of the six function s o.f selling, the presence of both ma sc uline and femini.ne traits was associated with .selling effective ness in two of the three sellino situations. Effective performance of the functions 0 of contacting, probing for needs, and retaining were assoc i.c1tccl with both masculine instrumentality nnd fe min ine expressiveness. Effective performance of the functions of prospecting, stimulating desire, and closin?> was associa t ed with masculine i.nstrumcnt.11 i.ty 0 only. Overall, the study provides stronJ support importance of masculine instrumentality to effecti.ve se lling performance , a nd demonstrates ,1 1,caker, but significa nt, contribution of feminine expressiveness . The importance of the '.!.'27 contribution of the femi ni ne compone nt is highly impacted upon by the pa rticular selling s ituation. Mana gerial Implications The findings of the s tud y have i mpl ications for recruitment :=md selection of sal eswomen, fo r training , for supervi..sLon a nd evaluation, and for incentive sys tems . Recruitment. Ideally, sa l es ma nagers would l ike t0 seek out those f emale applica nt s who ha ve the grea t est potential for success on the sales job. Re sult s of this s t udy s uggest that these are androgynou s a nd / or ma sculine fema l es . Sa l es managers who supervise saleswomen who sell t o or ganizationa l e nd users and to rcselle,-s, s hould seek out androgynous females since ~ c omhLnation of masculine and feminine tra i ts seems to be a n asset . Sales manager s who s upervise sal e swome n ,~10 sell lar ge-ticket items t o Ln-homc c onsumer s s hould seek out mascu l i ne femoles , since the prc s r11cc of feminine tr a i t s seems t o de t ract from effective performa nce . The BSRI as a Se l ection Tool. While Lt Ls tcmpti..ng to rec omme nd t he use of the Bern Sex Ro l e Inventory ,1 S ,1 s,1, :t i on t ool, l to do so on the basis of this rese ;1rch wouhl be pre111,1t 11re . Two problems s t and i n t he way : (1) the 2redir'tLvc n:1turc of the ') in s trument hns no t been established,- and pot e ntial " sexism" neerl to be countered . 3 Holh i s su,' s 1w ,t1 [() lit' resolved before the BSRI could effectively be used as a selection tool, and both are beyond the scope of the present research. Self-Selection. The BSRI might be useful in helping female applicants for selling positions select those types of jobs for which they are best suited. Some female applicants may express interest in jobs that are not appropriate to their sex role types. 4 In order to help these women determine which jobs are most suitable to them, recruiters ( as well as vocational ad visors) could make information available to them about the correlates of success on the job. Information about the rela tionship between BSRT scores and selling effectiveness in the particular type of sell i.ng would be helpful. Information about the type of behavior that i s perceived to be dysfunctional on the job would also be useful. Female applicants would then be in a better position to make a n informed choice a bout the job. Sales Training. The findings of the research suggest that the single most important asset of saleswomen is their "selling ability." Perceived problems in "selling abLlity " co rreL1tcrl negatively with the performance of eacl1 functi.on of se lling . Problems in "selling ability" nre direc tly correct::ible thro 11 gh training programs. n i.stori.cally, saleswomen huv foced harriers to acquiring important knowled ge a nd selling skills throu gh [urmnl training programs and informal networks . Thi s w,1s part i.c11lnrly 1 ----------- Problem ,-rhen women l?'er r? o 1 u I: ? y L V n WCO!ll r t u u fore ma nag r s 1-rh o perceive thelr- sal swom n 0 b Wi>-1]? knowledge i n a nd selling skills, should ex mine t:h rnunn F,' 1.1'1 wh?' h knowledge and skills a re trans mi tted wi t h i n their compo nie ,'l n d o certa in that the training needs of their sa leswomen are being adequately met. Training in Interpe r sonal Relations. The dysfunct i onal pattern of interpersonal relations that was noted in saleswomen who sell to in-home consumers may be correctable. Sales managers who pe r ce ive this problem among the saleswomen on their staffs need to (1) make the women aware of the problem, and (2) develop programs fo r the purpose of training them in alternative interpers onal s kill s . S Assertiveness and Sensitivity Tra ining. Those women whose sex role types are not congruent with the job environment may need special help. Saleswomen who a r e weak in masculine trai.ts may be able to s trengthen them by receiving asse rtiveness trJining. Saleswomen who are weak in feminine trai. t s may be ab l e to strengthen them by receiving s ensitivity training . S uperv1? s1? on an d Eval ua t ion. The study s 11 goested 0 t hat the performa nce of saleswomen was pe r cei ved t o be less effective when communication difficulties existed between sales 111c1nage r s and saleswomen ( e. g ., sales ma nagers felt uncomforL.Jb l c i n supe r visin g them, were not familiar wi th the work of s.Jlr>swomP n ) . S.J lcs 230 managers should try to establish on-going dialogues with the women in their employ. The BSRI could be used as the basis of such a dialogue. 6 Such dialogues could serve to open the channels of making it easier for sales managers to give communication, 7 constructive criticism, and for saleswomen to receive it. Incentive Systems. Based on the findings of this research, special incentive programs would be most usefully directed toward Up gra . f II d 1ng performance o prospec t i.n g II an d II re t a1. n1. ng fl activities, since these are the functions that appear to suffer most ivhen motivation is insufficient. Sex Bias on Salesforces. Because of t he bias uncovered in this research, those sales managers who are male, who feel less comfortable supervising women than men, and who s uperv ise [ewe r than 40% women, as well as those who have only infrequent c ont,icts with the women they supervise, should be alert to the possibilit i es that they may, unconsciously, be mentally s t e r eotypi ng their Since the behavior of "stereotypers" cn n nctuallf saleswomen . elicit stereotypic behavior on the pnrt of the " stereotypces," it is possible that, by their very manner, these sales mannge ,-s .ire inadvertently contributing to the problem of stereotypic hehavi r and reduced performance effectiveness of their saleswomen. 231 Managerial Recommendations The findings of the research suggest tha t s.:iles ma nage r s s ho continue their efforts to recruit wome n o nto the i r sa l es fo r e While some loss in selling effec tivene ss is proba bl y a ttributa to gender stereotypic behavior, the findin gs s uggest that the r e ways to minimize such problems. Specific reco~nend ation s addressed toward each of the three diffe rent selling sit ua t ions : Selling to Organizational End Users Sales managers who supervise s ales forces t ha t sell organizational end users should f oc us th e ir r?~r. rui ti ng selection activities on women who have bo t h s tr o ng masc ul i.n c feminine traits (androgynous fem3l e s). Trainin g pr ogr:1ms she focus on improving job-related knowle dge .:ind s ki ll s , .1nd s h< include both assertiveness and sensi. Livlty tr.:iinin g . Assert ive 1 and sensitivity training s hould both be i.n c ludcd becau se! masc t1 and feminine traits were both stro ngl y :1 ssoc i.,1tc d wi_L11 sc> L effectivenes s. No remedi a l trai nin ;~ i n intPq>e r s onn l s k appears nec essary, s inc e s t e r eo typi c be h,'1vi. o r in "h uman r 0 l ?i ti. did not relate t o reduced ef f ec t i.v cness of pc r fo rm:111c ,~ of , lll ) ---- --. --- --- the functions. Incentive programs should be directed toward improving performance of "prospecting" and "retaining," since performance of both of these functions appeared to suffer when motivation was perceived to be insufficient . .?..elling to In Home Consumers Sales managers who supervise salesforces that sell large ticket items to in-home consumers should focus their recruiting and selection activities on saleswomen with stro ng masculine traits (masculine females). Training programs should focus on developi13 job-related knowledge and selling skills. Training programs should include assertiveness training as well as remedial training to help saleswomen develop alternative interpersonal ski lls. Assertiveness training is important because masculi ne traits were strongly related to successful performance in this selling situ1tion. Remedial training in alternative interpersonal skills should be included since stereotypic behavior in "human relati.ons " was associated with reduced effectiveness. Sensitivity training should be given only sparingly, because an excess of femi nin e tralts appears with a p::ittern of interpersonnl to be associated relationships that detracts from effectiveness in "probing " and "closing." The feminine traits of saleswomen only appe,ired to be assets when saleswomen delivered their sales present:itions. :233 Special incentive programs should be directed toward improving performance of the function of prospecting since it appeared to suffer when saleswomen had insufficient motivation. Selling to Resellers Sales :nanagers who supervise salesforces that sell to resel lers should focus their recruiting and selecting activi.ties on saleswomen who have both strong masculine and fe mini.ne traits (androgynous females). Training programs should ir1<:l11de trai.ni.ng in job-related knowledge and selling skills. Trainin 3 programs should include both assert iveness an ry setting. '.239 Other Bia s Other possible biases in the s tudy i nc lud e : (1 ) sex bii:is a nd (2) the bias of acquiescence. Sex Ilias. There wa s an indication o f s e x bias i n t he res po nses of some of the sales managers who supervised sale s wome n sell i ng t o organizational end users. Simila r e rror may ha ve bee n i ntrodu ced by responding sales managers who were not as fa mili a r wi t h the performance of saleswomen as were others . The e rro r tha t wns introduced is probably in the directio n o f confirm i ng t he stereotypes. Acquiescence. The degree of "yea sayin g " or ac qui escence a mo ng respondin g sales ma nag e rs was not test e d. Post hoc e xamina t ion of the items on the individual stereotype s c ales r e ve::i l ed t hat the reversal of polarity of items was unba l a nc e d. Mo s t ite ms we r e reversed in the "selling ability" ste r e otype s ca les wh Ll e mos t items were not reversed on the "human relation s " .:ind "motiva t ion" stereotype scales. In future res ea r c h, 50% o f th e i terns 0 11 each scale should be reverse c oded. (The pr ob l ems e nco unt e r erl whe n attempting to reverse cod e the s e it e ms we r e disc ussed 0 11 page 135 . ) In future research, two ve rsions o f the ques t ionn :-ii r e !nigh t be used, each with a different se t of items r e ve r se cod e d. Res po ns es to th e two ve r s ion s o f th e ques tio nn::i ir e cou ] d the n be comr ,ir e tl .:i s 240 a test of the degree of acquiescence. Scoring of BSRI The study was limited because of the manner in whic h sal eswome n were classified into sex role types. The median split procedure that was used is in common use in research using the BSRI, but i s limited because it is sample ba sed. Several problems r e l a ting t o interpretation resulted: (1) the classificatio n of a ny partic ula r saleswomen was only relative, which could be mi s leadin g s inc e, in comparison with the general population, these s al eswo me n sco r ed very high in both masculine and feminine traits (e. g ., a s aleswome n who was classified as "undifferentiated" in this re s e a r c h mig ht be classified as "androgynous" in compari s on with women in the ge nernl public), ( 2) since saleswomen were c lassi f ied in to s ex rol e types bas ed on the medians of their own s amples, the defini tio n of "androgynous," "masculine," "fe minine," a nd "undi f f e r e ntia t e d," females was different in the three different s ample s . \vhil c the r e is precedent for this in the sex role id e ntity lite rature (e . g ., Hamby and Shapiro 1983), conclus i o ns involvin g compa ri so ns nmo ng the three samples a re thereby we ake ned . Norms are Needed. In future r esea r c h invo lving sex r o l r ty pes of saleswomen, the values used to se para t e sa le s women i nt o sex r o l r types should be pre-determined. /\ p pro pri. ,1t e no r111 s need to be '.: 41 ------ - established, both for research purposes and for use by sales managers who may want to apply the results. One suggestion would be to adopt the overall sample median of the masculinity and femininity scores determined by this research: 5. 3 (M), 5. 2 ( F). Another possibility would be to use an average of the medians weighted by the representativeness of each type of salesperson in In this research, these values were: 5. 31 (M) , 5. 29 the sample. (F) ? It may be that different norms actually apply to different selling situations. Sales managers might want to use the medians of the sample in which their saleswomen were c lassified as norms. These were: (1) selling to organizational end users: 5. 25 (M), 5 . 05 (F); (2) selling to in-home consumers: 5.30 (M), 5.50 (F); (3) selling to resellers: 5.50 (M), 5.20 (F). Further research should be conducted using standardization samples for the specific purpose of developing norms for use in selling situations. Standardization samples might include saleswomen from traditionally female areas of selling as well as some from traditionally male areas of selling. Re lative Importance of "T ypes " versus "Tr 31? t s " ? Ca uti?o n should be used ? ? ? 1? ndi? vi? dual saleswomen to sex role types . in assigning There is much room for error, and those Lndividunl s whose scores [t Ls lie close to the median are likely to be misclassi Eied . strongly reconnnended that further research Ln thls area not focus on the assignment of individuals t o categorLes , but rather on the 142 strength of their masculine and feminine traits. The results could then be analyzed using multiple regression techniques. Comparison of results across samples would be greatly facilitated and the problem of classification to sex role types would be eliminated. Terminology of the BSRI The Bern Sex Role Inventory is somewhat dated. Masculine and feminine items were developed from traditional cultural values that have been changing rapidly. At the time the scale was developed (circa 1974), people were not as sensitive to " sexism " in lang ucige as they are today . There is presently some sensitivity to the use of the terms "masculinity" and "femininity," and many researchers now substitute the terms "instrumentality" and "expressiveness (e.g., Jagacinski 1987). Because it was felt that the older terms ("masculinity" and "femininity") would be more me:mi.ngful to sales managers, they were used in the writin g of thi s report. Tt should be understood that the terms "masculinity" and " feminin i.ty" ,ire used synonomously with "instrumentality" and "expressive ness " in the discussion. Measurement Issues Certain aspects of measurement impact upon the confi.d ence 1vc may place in the conclusions and suggest further ureas for reseHrc h. ~43 The Validity of the Scales. The construct validity of the function effectiveness and stereotype scales was not established. Also, the issue of the dimensionality of the stereotype scales needs to be resolved. These issues should be resolved before any further use of these scales is a ttempted . .?.tatistical Limitations Some limitations result from the manner in which the statistical tests were applied to the da ta. Statistical Assumptions. The data did not conform to all of the assumptions underlying the derivation of the statistical tests (See Appendix F). Tue results probably over-estimate the true significance levels and effect sizes slightly. Requirements of Computer Software. The ge ne r alizability of the results of hypothesis seven (the r e l a tions hip between sex ro le types and individual function effectiveness scores), is limited to saleswomen who perform all of the functions of selling , because of the requirements of the computer program that w? s used (SPSSx MANOVA). In the case of those saleswomen who sold to r esellers, the function of prospecting ls exempt from this requirement. 244 An Agenda f or Future Research Based on the foregoing limitations, an agenda for future research is proposed: (1) A study should be performed to determine the extent of agreement between the perceptions of sales managl:!rs and saleswomen of the sex role identities of saleswomen on the BSRI. (2) Research should be performed to establish norms for scoring the BSRI in different t ypes of selling situations. (3) Research should be performed to validate the gender stereotype scales and the function effectiveness scale. (4) A longitudinal study should be performed to determine whether sex role masculinity anJ femininity scores change over time in response to conditions on the job. (5) Another longitudinal study should be performed t 0 establish the predictiveness of the BSRI. The masculinity and femininity scores of applicants for sales positions should be evaluated a nd compared with the selling effectiveness of the same women on the job. ( 6) The study should be re plica t ed with a s.-:implc of men. (7) The study s hould be replicated in a different nrcH 0[ the country and/or nationally. '.245 I .... Endnotes 1. See Leigh 1987 for a recent review of the us e of pe r s ona lity tests in the sales selection process. 2. The relationship that was establi s hed in thl s r esea r c h between sex role type and selling effectiveness i s c orre luti o na l only. There is no ba sis to determine whether sale swomen a r e mor e effective performers because of their sex rol e t yp e , o r whe the r these traits develop as a by-product of effec tive perfo rma nce . 3. The "sexism" issue seems to relate more t o the use o f "sexist" terminology than to di s crimina t or y pra c tices tha t mi g ht result from the us e of the scale. Chan ging the names of t he scn l Ps from "masculinity" and "femininity : t o "in s trument a lity " .:rn d "expressiveness," as has been done by a numbe r o f c urr e n t researchers (e.g., Jagacinski 198 7), would he lp. De-emphas i z ing classification into sex role types and e mpha s i zi ng the r e lnti ve streng th of the "Instrumental" and "express ive" compo nent s of sex role identity would also help. 4. While research in o the r di s c iplines has [ou nd individual s tend to prefe r and c hoose vocntio ns tha t n r e cu ng ru r n t 246 with their sex role types (e.g., Cl arey an d Sanford 1982), Comer and Jolson (1985) found some indication that fe male students of sales management were interested in selling j o bs that were i_nappropriate to their sex role types. 5. Script-based training combined with role playing exercise is a promising approach that might be useful in such behaviora l training programs (See Leigh 1987 for a recent discussio n of the use of script-based learning in sales training programs). 6. Sales managers and saleswomen could comp a t~c their perceptions of the saleswomen's sex role identity. They could discuss the importance of the traits in their selling areas nnd h ow improvement in the traits might enhance their selling perfo rmance . 7. Sales managers might a l so consider developin g cv;:i luatio n systems a round performance of the function s of selling, s uc h ns wns used in this study. The procedure focu ses on s pec i fie behav j oral aspects of performance. Constructive cri.t i c i s ;ns b:1sect on s 11 c h an evaluations would be helpful to saleswomen nnd might also contribute to the on-going dialog ue be twe e n saleswomen and the i_r managers. 8 . Chonko, Howell and Be llinge r (1986) suggest t:ha t self-ratings a re more commonly used in r esearc h involving j,)i> satisfac tion than in research invo lvi.n g selling e f fec ti_v e ness . ?.47 9. While some degree of corresponde n ce be t i..ree n pe r cep ti a n s o E sex role identity by the self and by an o the r h ;JS been reported (e.g., Spence and Helmreich 1978; Alpe rson and Friedma n 1983) , these studi?es i?n vestigate perceptions o f pers ons w h o a r e i ? n socia l_ relationships ( parents/children, heterosex c ouple s ) not in e.g., b . usiness relationships. 248 BIBLIOGRAPHY Abramson, Paul R., Philip A. Goldberg, Judith H. Goldberg and Linda M. Abramson ( 1977) 1'The Talking Plata pus Phenomenon: Competency Ratings as a Function of Sex and Professional Status," Psychology of Women Quarterly, 2 (Winter), 114-124. "A S urvey of Sales," (1985) Marketing Co mmunicati.o ns, (November), 71. Adams, Kathrynn A. (1980) "Who Has the Final Word? Sex, Race, and Dominance Behavior," Journal of Personality and Social Psychology, 38 (no. 1), 1-8. Alagna, Sheryle W. (1982) '1Sex Role Identity, Peer Evolut1tion of Competition, and the Responses of Women and Men in a Competitive Situation," Journal of Personalitv and Social Psychology, 43 (no. 3), 546-554. Alessandra, Anthony J., Philip. S . Wexler with Jerry D. Deen (1979) Non-MRnioulative Sellino. Reston, VA : Reston Publishing Company. Alperson, Burton L. a nd Will J. Friedman (1983) "Some Aspects of the Interpersonal Phenomenology o[ lleterosex Dyads \v Lth Respect to Sex-Role Stereotypes, " Sex Roles, <) (no. 4), 453- 474. Anderson, D. E. (1987) '1Most Women flold Low Paying Jobs," The Washington Post, (March 4), G .1. Anderson, Rolph E. and Joseph F. Hair, Jr. (1983) ~o les ~bnn?'emen t. NY: Random House. Antill, John K. (1983) "Sex Role Complement;1rity Versus S Lmilarity in Married Couples," Journal of Perso nalLty and Sr,cL1 l Psycholo~v, 45 (no. 1), 145-155. Antill , John K. and John D. Cunningham (1982) "Sex Differences in Performance on Ability Tests as a Fu ncti n o[ ~bscuLL11iLy, Femininity a nd Androgyny," Journal of Perso nnlity ind Soc j J I Psychology, 42 (no. 4), 718-728 . 249 Ashmore, Richard D. and Francis K. Del Boca (1979) "Sex Stereotypes and Implicit Personality 111eory: Toward a Cogniti.ve--Social Psychological Conceptualization," Sex Roles, 5 (no. 2), 219-248. Baer, Earl E. (1972) Salesmanship. St. Louis: McGraw-Hill. Bagozzi, Richard P. (1978) "Salesforce Performance and Satisfaction as a Function of Individual Differences, Interpersonal, and Situational Factors," Journal of Marketing Research, 15 (November), 517-531. Bar-Tal, Daniel and Irene IL Frieze (1976) "Attributio ns for Success and Failure for Actors and Observers," J o urnal of Research in Personality, 10 (no. 3), 256-265. Bar-Tal, Daniel and Irene II. Frieze (1977) "Achievement ~1oti v.:1 ti.on Formula for Males and Females as a Determi.mrnt of Attributions for Success and Failure," Sex Roles, 3 (no. 3), 301-313. Bartol, Katherin e M. and Philip J. Manhardt (1979) "Sex Differences in J ob Outcome Preference- -Trends among Newly '. fired College Graduates," Journal of Applied Psyc hology, 64 (no. 5), 477- 482. Behrman, Douglas N. and lvilliam D. Perreault ( 1982) " l?leasuring the Performance of Industrial Salespersons," Journal o[ Dusi.110s.s Research, 10 (September), 355-370. Bern, Sandra L. (1974) "The Measurement of Psycholo'.ii.caJ ,\ndrogyny ," Journal of Consulting and Clinical Psychology, 42 (no. 2), 155-162. Bern, Sandra L. (1975) "Sex Role Adaptability; One Consequence o[ Psychological Androgyny," Journal of Person,.1li.ty and SocL1l Psychology, 31 (no. 4), 634-643. Bern, Sandra L. (1976) "Probing the Promise of Androgyny," i.n Beyond Sex Role Stereoty[)es: Readings Toward A Psychology of Androgyny, Alexandra G. Kaplan and Joan P. Benn, eds ., Boston: Little Brown, 48-62. Bern, Sandra L. (1977) "On the Utility of Alternative Procedures for Assessing Psychologi.cal Androgy ny," Journ,1 I of Consnl tin,, ,111_=- =r--?:..:..s.:::.on.:.:.:::acl:::'----'S.:..:P_.l:::.l.:.c::::i.:.:n....,_?' and Sales Management, 5 (Ma y), 15- 22 . Concoran, Ma ry and Gregory J. Dunca n ( 1979 ) " \fo r k Ii i s t o r y , Labor-Force Attac hment, ond Eorn i ngs Di ffe r e nces l3c t 1vecn 1: lw Rae es and Sexe s , " -'-J-'-o_u_r_n....ac__l _o_ f_ _l l_um_a_n_ _R_ r>_s:....0:....1_1_r c=-~ .:::.ec.::..s , l!~ ( \vi n te r ) , 25J 3-20. Constantinople, Anne (1973) "Masculinity and Femininity: An Exception to a Fameous Dictum," PsycholoP,i.cal Bulletin, 80 (no. 5), 389-407 . Cooper, Harris M. (1979) "Statistially Combining Independent Studies: A Meta-analysis of Sex Differences in Conformity Resea rch," Journal of Personality and Social Psychology, 37 (no. 1), 131-146. "Corporate Woman The Indus tria l Salesman Becomes a Salesperson" (1979) Busi ness Week, (Feb ruary 19), 104-11 0. Cothan, .J. C. a nd D. W. Cr avens (1969) "Improving Measurement of Salesperson Performanc e ," Business Horizens, L2 (June), 79-83. Crave ns, David W. and Robe rt B. Woodruff (1973) " An Approach for Determining Criteria of Sales Performance," Journal of Appli_ed Psychology, 57 (no. 3) , 242-24 7. Crissy, W. S. E., Willi a m H. Cunni ngham a nd Isabella M. Cunningham (1977) Selling: The Personal Force in Marke ting . San ta Barba ra, CA: Wiley. Cullen, John B. and Perr ewe , Pamela L. (1981) "Super i.or' s , ind Subordi nates Gende r: Does it Really Matter?" Psychologi..: :iJ Reports, 48 (no. 2), 435-438. Darden, Bill R. a nd Wa rre n French ( 1970) "An Investig:1t ion i.:1to tli~ Salesman Evaluation Practices of Sales '.'1anagers," So uthen1 Journal of Business , 5 (July), 47-56 . Deaux, Kay (1984) "From Individual Differences to Sociol Categories," America n Psychologist, 39 ( no. 2), L05-Ll6. Dea ux, Kay (1 985) "Sex a nd Gender," Annual Review of Psychologv, 37, 1017-1031. Deaux, Kay a nd Tim Emswiller (1974) ;'Expbnntion.s of Successful Performanc e on Sex-Li nked Tusks : ~1~ t is Ski ll for the Male i . Luck for the Female," Journal of Perso nnli_ty .md Soci,11 Psychology, 29 (no. 1), 80-85. Deaux, Kay a nd Eliz:ibeth farrls (1979) "r\ttri.butinJ C:it1ses of One ' s Own Performance: The Effect of Sex Norms and 011tco111es, 11 ,Journal of Research in Personality, 11 (no . 1 ), 59-7~ . 254 Derlega, Valeria n J., Bonnie Durha m, Barbara Gockel, and Daniel Sholis (1981) " Sex Differences i n Sel f - disclosure: Effects of Topic Content, Friendship a nd Par t ne r' s Sex, " Sex Roles, 7 (no. 4), 433-447. Deutscher, Terry, Judi th Marshall, a nd Da vid Burgoyne (1982) "The Process of Obtaining New Ac c ount s ," Indus trial Ma r keti no Management, 11 (July), 173-181. Dion, Kenneth (1985) "Sex, Gender, and Groups : Sel ected Issues ," i n Women, Gender, and Social Psychology, Vi r g inia E. O'Lea r y , Rhoda K. Unger, Barbara S. Wall s t o n, eds., Hillsda l e , NJ: Erlbaum. Dipboye , Robert L., R. D. Arvey and D. E. Terps t ra , ( 1977) "Sex a nd Physical Attractivene s s of Ra t e r s a nd Applica nts as Determinants of Resume Evaluations ," Journ a l of Appl ied Psychology, 62 (no. 33), 288-294. Dipboye, Robert L., Howard L. Fromkin, a nd Kent Wi bac k ( 1975) "Relative Importance of Applica nt Sex, Atr r1c tive ness , a nd Scholastic S tanding in Eva luation of Job Appl ican ts ," Jo urnal of Applied Psychology, 60 (no. 1), 39-43 . DiPietro, Janet A. (1981) "Rough and Tumbl e Pl a y: A Func t ion of Gender," Developmental Ps yc holo~v, 17 ( no. l), 50- 58 . Donnelly, J a mes H. and John M. Ivan c evich (1975) " Role CLiri_ty a nd the Sa lesman," Journal of ~1a r ke ting, 39 ( J a nua r y) , 71- 74 . Doubman, J. R. (1939) Salesma ns hi p a nd Ty pes of Sel li ng . New Yor:c Appleton-Century-Crofts. Dubinsky, Alan J. (1980) "Sale s Tra in i ng : An Ana l ys i s of Fie] d Sales Techniques. Ann Arbor, MT: UMI Pre ss. Dubinsky, Alan J. , Roy D. Howell, Thomas N. Ingr am , a nd Da nn y N. Bellenger ( 1986) "Salesfo rce Socia l iza t io n," J o ur nn l of Marke tino, 50 (Oc tober), 192- 207. Dunnette, Marvin D. and Wa yne K. Kirc hm~r (1960) " Psyc holog i.cnl Test Differences be twee n Tndu s t riul S:.J l esrnc n oml RP. t- ,1i I Sale s men," Journal of Appli ed Psycho l o~;Y , 44 ( Apr il) , 12 1-12'i . Eag ly, Alice ![. (19 78 ) "Se x Diffe r e nces in Tnflue ncenb i.li ty ," Psychologic al Bulle tin, 85 (n o . 1), 86- ll 6 . Eagly, Alice !I. a nd L. L. Ca rl i ( l981) "Sex of Rese,irc hers .rnil :255 Sex-Typed Communications as De t erminant s of Influenceabilitv: A Meta-analysis of Social Influence Studies," Psychologic~l Bulletin, 90 (no. 1), 1-20. Eagly, Alice H. a nd Wendy Wood ( 1982) "Inferred Sex Differences i n Status as a Determinant of Gender Stereotypes abo ut Social Influence," Journal of Personality and Social Psyc hology, 43 (no. 5), 915-928. Eagly, Alice H. and Wendy Wood (1985) "Gender and Inflenceab ility: Stereotype Versus Behavior," in: Women, Gender a nd Social Psychology, Virginia E. 0' Leary, Rhoda K. Unger, and Barbara S. Wallston, eds., Hillsdale, NJ: Erl baum. Eccles, Jacqueline (Parsons) (1983 ) "Expectancies, Val ues, :rnd Academic Behaviors," in Achievement and Achievement Motivation, Janet T. Spence, ed., San Francisco: Freeman, 75-146. Elmore, Patrice B. and Karen A. LaPo.inte (1974) "Effec t s of Teacher Sex and Student Sex on the Evaluation of College Instructo r s ," Journal of Educational Psychology, 6 (June ), 386- 389 . Elmore, Patrice B. and Karen A. LaPointe (1975) "Effect of Teacher Sex, Student Sex, and Teacher Warmth on the Evalu~ti.on of Colle g e Instructors," Journal of Educational Ps yc hol o<;y , 67 (June), 368-374. Epstein, Cynthia Fuchs (1971) \.Joman 's Place. Berke ley, Cnl L[orn L,1: University of California Press. Erkut, Samru (1983) "Exploring Sex Differences in Ex pecta ncy , Attribution, and Academic Ac hievement," Sex Roles , 9 ( no. 2) , 217-231. Etaugh, Claire and Barry Brown (1975) " Perceiving th e C1uses of Success and Failure of Male and female Per(,>rnmnce," Developmental Psychology, 11 (no. 1), 103. Etaugh, Claire and Ethel Foresman (1983) "Evaluat ion s of Compete nce ns a Function of Sex and Marital Sta tus ," Sex R l es , <) (no. 7), 759-765. Etaugh, Claire and Sue Riley (1983) " Evaluat ing Competenc,~ o[ \vomcn and Men: Effects of naritnl and P,irenta l St .1t11s c1 r1d Occupational Sex-Typin g ," Sex Ro les , () ( no. G), 94J-l)52 . Etaugh, Claire and Suzanne Rose (1975) "Ad o l escents ' Sex f3i.:.1s jn '.256 the Evalua tion of Performance," Developmental Psychology, 11 (no. 5), 663- 664. Evans, Franklyn (1963) "Selling as a Dyadic Relationship--A New Approach," American Behavioral Scientist, 6 (May), 76-79 . Feather, N. T. (1975) "Positive and Negative Reactions to :vlale and Female Success and Failure in Relation to the Perceived Status and Sex-Typed Appropriateness of Occupations," Journal of Personality and Social Psychology, 31 ( no . 3), 536-548 . Feather, N. T. and John A. Said (1983) "Preference fo r Occupations In Relati.on to Masculinity, Femininity, a nd Gender," Br itish Journal of Social Psychology, 22 (June), 113-127. Feather, N. T. a nd J. G. Simon (197 5) "React i.ons to Male and Female Success and Failure in Sex-Lin ke d Occupations: Impressions of Personality, Causal Attributions , and Perceived Likelihood of Different Consequences," Journal of Pe r so na lity and Socicil Psychology, 31 ( no. 1), 20-31. Feldman-Summers, Shirley a nd Sar a I3. Kiesler ( 1978) "Those lvho ,\ re Number Two Try Harder: The Effect of Ca usal Attri but ions of Causality," Jour nal of Pe rsona lity and Soc i a l Psyscholo~w , 30 (no. 6), 846-855. Ferber, Mari.annc J. Hube r a nd Gle nd a Spitze (1979) " Preference for Men a s Bosses and Professionals ," Soc Lal Forces , 58 ( 110. 2), 466-476 . Fetler, M. (1985) " Sex Diffe r ences in the Callforn .L1 StatewL.ie Assessment of Computer Literacy, " Sex Roles , 13 ( no. 3-4) , 181-191. Figel, Mar ge (1983) "Women in the Winner's Circle, " The r\mer-L cc111 Salesman, 28 (June), 3-6. Finn, David W. and lvilliam C . Mo ncrief ( L985) "Investi.~:1t-Lng Salesforce Sex Discrimina tion," i n: ProceedinP s of the American ~farketing Association, Robert F . Lusch et al ., eus ., Ch ica go : America n ~farketing, 192-197. Fishman, P. M. (1978) " \./hat Do Couples Ta1k .\bout when They ' re Alone? " in lvomen ' s Lang uage and Style , D. Battarett , ed., Akron, OJI : University of Akron. Fox, L. 1-f., D. Tobin nnd L. Brody (L979) "Sex-role SocL,1li z:1tio11 a nd Achievement in Mi..lthematics, " in Sex-rel1ted Di.f'[cr 11c 0s i.n l.57 Cognitive Functioning, M. A. Wittig and A. C. Petersen, eds., New York: Academic Press , 303-334 . Futrell, Charles M. (1980) "Salesmen nnd Saleswomen Job Satisfaction," Industrial Marketing Mangement, 9 (February), 27- 30 . Gable , Myron and B. J. Reed (1987) "The Current Status of Wo men in Professional Selling ," The Journal of Pe rsona l Sellino and Sales Management, 7 (May), 33-40. Gaski, John F. and Michael J. Etzel (1986) "The Index of Con sumer Sentiment Toward Marketing," Journa l of Marke ting, 50 (July), 71-81 . Gaudreau, Patrice (1975) "Factor Analysis of the Bern Sex Role Inventory," Journal of Consul ting and Clini_ca l Psychologv, 45 (no. 2), 299-302. Gayton, William P., Gertrude Havu, John G. Baird, a nd Kenneth Ozman (1983) "Psychological Andro gyny an d Asse r ti ve 1wss in females," Psychologica l Reports, 52 (no . 1), 283- 285 . Gerdes, E. 0. and D. M. Garber (1983) "Sex Bias in Hiring : Effects of Job Demands and Applicant Compe tenc e ," Sex Roles, g (no. 3), 307-319. Gh iselli, Edwin E. (1973) "The Validi t y of Aptitude Tests in Personnel Se l ection," Personnel Psychololuring r i.;J l and Clerical Occupntions: Differe nces in Person,1li y 267 Charcteristics and Role Orientations." Ph.D. dissertation, University of Cincinnati. Myers, Jerome L. (1979) Fundamenta ls of Experime ntal Design . Boston: Allyn and Bacon. Nelson, Richard (1987) "Maybe It's Time to Take Anothe r Look at Tests as a Sales Selection Tool," Journal of Personal Sellino and Sales Management, 7 (August), 33-38. Neter, John and William Wasserman (1974) Applied Linear Statistical Models. Homewood, IL: Irwin. New Profit Opportunities in Hiring Women (1977) New York: Resea r c h Institute of America. Newton, Derek A. (1967) "A Marketing Communica tion Mod e l for Sa l es Management," in Risk Taking and Information Ha ndling in Consumer Research. D. F. Cox, ed., Boston: Harvard University. Nicholls, John G. (1975) "Causal Attributions nnd Other Achiev~nent-related Cognitions: Effects of Task, Outco~e Attainment Value, and Sex," Journal of Personality and Socia i Psychology, 31 (no. 3), 379-389. Nickels, William G. (1984) Marlcetino Communica tion. New Yor!c: 1viley. Nickels, William G., Robert F. Everett and Ronald Kle i n ( 1983) "Rapport Building by Salespeople; A Neuro-Li ng u Ls ti c Approach," Journal of Personal Selling a nd Sa les Man agement, 3 (November), 1-7. Nunnally, Jerome c. (1978) Psychometr -i. c Theory, 2nd ed., New York: McGraw-Hill. Nyquest, Linda v. and Janet T. Spence (1986) "Effects of Dispositional Dominance and Sex role Expectations on Lead er ship Behaviors , " .'.::'.J~o~u7r_:_:n~a~l'----=o:....:f'---_P_e:::...r::...cs:::...o' ~nc?=1-=-l=i.-.::t...._v_~m:..d:: :__.:S~o~c~i~D~l Psychology, 50 (no. 1), 87-93. "Oh, What a Feelin'--Selling l,'300 Cars a Year," (1984) ~1arketing News, (September 14), 5. O'Leary, Virginia A. and Rnna ld D. Ha nsen (1982) "Tryi ng llurt s Women, Helps Men: The Meaning of Effort," in Women in the Workforce, IL John Berna rdin, ed., New York: Pc1eger, LOO- L~3. 268 Olds, Debra E . a nd Phillip Shaver (1980) "Masculinity, Femininit:1 , Academic Performance and Health: Furthe r Evidenc e conc ernino II o the Androgyny Controversy, Journal of Persona lity, 48 (no. 3), 323-341. Olsc havsky, R. M. (1973) "Customer-Salesperson Inte r a ction in Appliance Retailing," Journal of Marketing Researr::h, 10 (May), 208-212. Orlovsky, Jacob L., Alice L. Aslin, a nd Sheila D. Ginsbur g (1977) "Differential Effectiveness of Two Classification Procedures on the Bern Sex Role Inventory," Journal of Personalitv Assessment, 41 (no. 4), 414-416. O'Shaughne ssy, J. (1971-72) "Selling as an Interpersonal 1nflue ncc Process," Journal of Retailing, 47 (Winter), 32-46. Osen, Lynn M. (198 4 ) Women I n Mathematics. Cambridge, '1A: The Massachusetts Institute of Technology . Parsons, Talcott a nd R. F. Bales (1955) Familv Sociali ~.:1tion ,11-1ct Interaction Process. Gl e ncoe , IL: Free Press . Paludi, Michele A. and William D. Bauer (1983) "Goldbe r g Revi.sted: What' s in a n Author' s Name," Sex Roles, 9 (no . 3), 38 7-390. Paludi, Michele A. and Lisa A. Strayer (L98 5) "Wlwt' s i n an Author's Name? Differential Ev;,luati.ons of Per[orma nc:e 18 .1 Function of Author's Name," Sex Roles, 12 (no . 3), '353-36 1. Peck, Teresa (1978) "Whe n \vomen Evaluate ',fo:nen, N0th i ng Suc c eeds Like Success: The Differential Effects oE S t a tu s lJpo n Evaluation of Xale and Female Professional AbLlLty ," Sex Roles, 4 (no. 2) , 205- 213. Pederson, Coulton A. a nd Melburn D. Wri gh t (1971) Sale s mon s hip. Homewood, IL: Invin. Pedhazur, Elazar J . a nd Toby J . Tetenbaum (1 9 79) "13e m S,~x R,J lc Inventory: A Theoreticol a nd Methodologi_cal Cri_ti.que, " J o urn ?1 L of Personality and Social Psychology, 37 (no . 6), ()96- ll)I6. Ph e t e r son, Gai l I., Sara 13 . Kiesle r and Phi.Li.p A. r.o l Jberg ( L'J?l) "Eva luation of the Performance of \./omen a s ,1 Func ti.,)n of he i 1- Sex , Achievement, and Pe rsonal l-1 Ls tory ," J o urn n I o f Pe r s ona lity a nd Social Psycholo~v, 19 (no . l), ll4-ll 8 . Pl.eek, Joseph l-1. (l Among College Freshmen ," Journal of Vocc1tional f\o h ,-1 \? i or, 1 '.2 (no. 3), 261-269. Yoder, J. D. (1985) "An Academic lfoman as a Token: A Case Study, " Journal of Social Issues, 41. 61-72. Yoder, J. D., J. Adams, and H. T. Prince ( 1983) "The Price of a Token," Journal of Political and Military Socioloc:sv, 11, 325-337. Young, James R. and R. Wayne Mondy ( 1982) Personal Selling: Function, Theory, and Practice. New York: Dryden. Zickmund, W. C., M. A. Hitt and B. A. Pickens (1978) "Influence of Sex and Scholastic Performance on Reactions to Job Appl i.cant Resumes," Journal of Applied Psychology, 63 (no. 2), 252-255. Zimmerman, D. H. and C. West (1975) " Sex Roles, Inter rupti.on s and Silences in Conversation," in Language and Sex: Differences and Dominance, B. Thomas and N. Henley, eds., Rowley, ~L\: Newbury House. '279 APPENDIX A LE'ITER OF TRANSKITIAL AND TI:IAN1C YOU LETI'ER 00000 THE UNIVERSl1Y OF MARYLAND COLLEGE OF BUSINESS AND MANAGEMENT College Pork . Maryland 20742 LE'.ITER OF TRANSMITI'AL September_, 1988 Mr. John Doe Sales Manager go?n Doe Corporation aithersburg, Maryland 20877 Dear Mr. Doe: I greatly appreciated the nice conversation I had with you today and am enclosing the questionnaire that I discussed with you together with a postage-paid envelope f or your response. I may be telephoning you in a few days to find out whether you have any questions. Thank you in advance for your help with this project. I will be s ure to send you a copy o f t h e summary repor t as soon as it becomes available. Sincerely yours, Lucette B. Comer Doctoral Candidate ~8 1 -- - ?~ODO THE UNIVERSl1Y OF MARYLAND COLLEGE OF BUSINESS AND MANAGEMENT College Pork . Maryland 20742 September_, 1988 Ms. Jane Doe Sales Manager XYZ Corporation Baltimore, Maryland Dear Ms. Doe: ~. would like to thank you for your _res_pons i veness to my sur vey oout women in selling and for your WJ.llingness to t ake t ime f r om your busy schedule to complete the questionnaire. Please be assured that your help is greatly appreciated. I will send you a copy of the summary report of the s tudy as soon as it becomes available. This will probably be s ome time i n the spring. . Meanwhile , thanks for your interest a nd your ve r y v a 1 ua b l e assistance. Sincerel y your s , Lucette B. Come r Doc t or a l Cand ida t e 282 .. APPENDIX B COVER LEITER AND QUF.STIONNAIRE 233 00000 H IL UNIVERSITY or MARYLAND COLLfGf or BUSINESS AND MANAGEMENT r:-011~-, Par* Mo,y?ont1 701.t"l ~ Sales Manager: Thank You for your poeitive respoose to our racent telepoone cx:nversat.ion about woren in 9ell.ing. rt is appu-ent that "?""" are appearing non, frequently en sales fo~ that ~ originally staffed only by males. My doctoral dissertation "1.11 at~ to rneoasun, O!l!ectiveness of various types of !!!111!les in performing a variety of selling tasks. The attached questionni,i.re n,quests }'OUr de9cript1ons of sane of the sa le9'oO!l?n wtn t'ep:,rt to }'OJ. Please select _S2 to 4 of ~ sale,,,,onen, woo yo., feel .u-e l!BXWally d.iffe~t fron eadl other in their selling style, performance .url results. Yru may flru it 8 ~ler to cx:mpire 4 people than to limit ~ description to T or 2. The questio,nain, is not brief. ~leting it 1111y 11M:>lve as l1UCh as one half hcur of }'OUr time. lole realize that th.is is a heavy dEm1nd , but we be llev.. that the in?,rt.ance of t:h1a r:-esea.rch justifies the t1me expendlture. As a particii:ent, YC>J "1.11 r..oelve a 0Jpy of the final report. Yoo nay also derive 9IJTl9 personal benefits thruugh the insights YC>J gain when a::t1')leting the thC>Jght-prCMJking it.ems. S.lnae our ~ 15 to pinpoint dlffen,naes dff'Cn9 sale""'?""'1, please be ~ critical in your responses. Cnly yu, "1.11 Janw the identity of the irdlvidual(s), an:J )'l'.XJ.r l:"eSpcnsa "1.11 be kept ~letely ccnfidential. Reports 1o1ill reflect only sum,,cy data. ~ .!_s rarticularly inp?rtant that .l'.Q!! ~ all of the questions. Pl&-i,.., return YOJr leted questionnaire in the enclooed seH--..cl:iresse::, post:age-p, ld envelope w! thin ~~- Thank )'OU in advance for p,r ccoperation. If you have any questlcns ,:ibe>Jt th.is l"eSearch, or any other related 1111 tter, please do rot hesitate to call oollect. ~~ (Mo. I wcette B. Curer Ph.D. candidate ( JOT I 871 - 3896 284 I. The per90n11l selling proce5s consial:3 of a seri8ft of tasks that sales people !7'rform. 1heoo are: ( 1 ) Prospecting: Identifying prospective custcrrers for the firm -anci/or its pro:lu:: t. (2) Oontacting: n.e first one--0n--<:ne relationship bati.een the s.a l ~sper90rl am trn prospect. (3) ?rc:tiing for Needs: Oet:e.rm.ination of the pruspect's needs/problem a """9. ( 4 ) Stinulating Desire for the Pro:lt.ct: 'Il>e actual sale:, pn,sentation/ di9CI.J s sion. (5) Clooing: Asking for, and obtaining, the order. (6) RetaWng: Retaining the order and t:.he aistcrrer inclu:ling the rendering o f p:,at-transactional 9erVioes. 17) !'on-selling Activities: Tasks that nust ba performoacl by trn salesp-,r90rl rut -,filch ,u,, not i:art of the actual selling duties (e.g., rea:ml keeplnslgna t,? tho>-" as SdleSWQlan A fl Can:! D (Salesworan A ? SW-A). Respc,nd to t i_. lt?Jn9 ,,tout ~,ch o f t MTI In the appropriate' 00'lUTtllS. If yw describe c:n1yo ne ,9aleSWCJT\:ln, use t111e cr, i..,.,s for SW- A ar,1 leave t:.he rest of t:.he =lunns blank; lf you Jes=ibe t:.oo, u"" t~ tJ,-, o:, l1rnr1s fo r 2'_".::_,\ an1 ~- Be sure .i'.Q!!. =nsls tenU,y use ti,e same =lu~ fo r ~ ~ -:!~J. ",!],::~~~ t;Jlr_~ .Jf~-!"t ~ ~estl_~l-~ - 285 I;f On a scale of O to 10 (where 10 is the hlgt-.,st level of effectivmess l, irdi ci te how :.,, ectively each sa leS'oOlrul perlorms the seven tasiwleclgable a.tout cunpetitive ixooucts, maJting '!ffective ~risons bet--, them and the ~y?s pr:odu::t ..oon appn;vriate to do so. They pe.nn.i t t--o way comun.i ca ticn thr:oughou t the.! r i:>r:esen ta tions , but retain strong Cttltrol of the ln tervi""'. ---- -- (S) Closin:q. Effectiveness in closing lnvolve,i asking for:, and obtaining, the order, ,19 appr:opr:iate to the j ob. Sa lesp,-rsons who ar,, effective c loser,i approad, the close "'ith ocnficienoe, uid have a goo:J sense of timing when asking for the o rder? If t1lP Order is ~t ootained en the first attE!!1)t, they follow up ,1s appropriate to the selling_s ituation. ------ (6) Retaining the Sale and the aistauer. Effectlveness in retaining the sale and the cust:oiP_r involve" follow- up after the sale to t:e sure that the oustorer receives the produ::t ~" pronisPd, (e.g. , on u,-. , in good ccnd i tion, serv1 ced an::l l ns t.-1110'.l I , vYl providing goo:l custorer services, ._,,,. appropriate. F:(fPcti VP retainer,, rraintdin regular: oontact with custcm,ars, are rosponslve to changes in custaners' produ:=t - relatoo net'ds, and ke,,p cu ~toner., lnforn-ed a.tout new ixoouct.s and new appllcat1ons. TI1ey tlal\'.:llc> rU ssa tis faction and ~la in ts '!Xj)frli tious l Y ? (?) ~~ Duti,,,.. Good performance o f ncn-,.,....lllnq rlutiPs involves 3 uch tlilngs ,19 keeping a=1rate anci ,,p-~o- not necess.arlly a ssoc i,1ted "'1th one's job p?rformance. Ind icatP. ha,, clos,,ly yoo f,-,e l "3Ch of these tr~its descrU:,,, "3ch sale swaran, using the f o llowing srn l e : 6 ?----- - - - -- +------------+------- -- --+- -- - - - - -----+---- ------- ?------ - -- -- + Never or Usually Soretimes Occasionally Of t en Usually Always or alnost not rut true true true alnost neve.r true true infrequently ,1lway" true true ( 1 ) Deferds own beliefs (2) rs affectionate (J) rs 1.rrlepen-ient (e . g., prefers to stand on own feet in thl t ls capable of acting free of the influence o r guidance o f o ther~ ! ( 4) Is sympathetic (5) rs asserttv,, (e.g., ts characterized by det.ecminatioo am l::ol.clre ss in stating opinions, or in o thent1se rraking her presence o r l n flu-,,>ae felt) - ----- -- (6) Is sensitive to the needs of otters ( 7 ) Has a strong pe r,aona, li ty ( 8) Is urderstan:J ing (9) Is f o r:-o,,ful ( 10) Is cx:rnp~ssionate (11) l!as l e ack,rship -ilii lities ( 1 2) Is eag~ to ~the hurt feelings ( 1 J ) ls "'llling t o take ri s ks (14) Is warn, (l5) Is cbni.nant (e .g . , ls forceful, masterful, iUld authortr,,. t! VP I ( 16) Is terder ( is soft:J-euted) (17) ls "'llling t o t..Jce a st;md on I s sues ( l 8) r,.,..,.,,. children (l 9 ) rs arJryr~ stv~ ( e . g ., l s /l\3rkecl by tx:>ld d PtPrinu\.., t l l")f'l <1r !'-r,ii-., rl v r> ro,ac! l rlPSS) ( 20) l s 'lPll t \-, 2S7 ~- Bel?"' is a list of characteristics that Me associated with one's s=ss ease indicate how closely you feel each st;,t.,,_.,t rlescribes &'!ch sales...-man. in s<>lli"'J . 2 J 4 5 6 ?----------+----------+-- --------+------------+---------- 7 + Never or Usually Scrnet.ilres Occasionally Of t en u;~ii;----N.:;;ys or almost not but true true true ,. lirost never true true infrequently "lway~ tru,, true 11 ) Is not ...,11 organized. ---- - - (l) Tends to over-socialize ..,1th prospects /c:us t:oie rn during sale" intervi~. ------ IJ) Tends to avoid activities that ,i1-ie oons iden to be unpleasan t . ---- -- l 4 ) Has strong persua,ilve sk.ill.9. ------ IS) Accepts prospect.,,' exa.r"""' and objections at faoe va lue . - ----- ( 5 ) L>cks o:nfidence. ------ (7) Displays unusually fine prod.rt Juvwledge. ------ (6) Ten:Js to relirquish control of ,iales inte rvi~ to c:ust:arers . ---- -- 19) Hae overly 81Dticnal r011ctioo to rejections oo j ob. ------ I 10) Catch,e,, 01 to new o::r,cept:.'I quickly. fl l) Over- nurb.Jres a.rst:oier.1 f Le. , ,iperd, too ouch t ime oo provid ing extra services to present c:us~ while neglect ing o ther aspects of the jcb, s uch as prospecting for new a.r9tarers I. ------ I 1 2 ) Needs continual rein forcenent to sustain !"' rfo rmanoe ? ---- -- ITJ) Adapts quidc!y t o non- ra>t.lne s ituations. ------ 114 ) F'\Jnctions ""'11 as "team player" (i.e., ax,.-,rate" "1th o thers on Job i . ____ ---- ( 1 5 ) Jll 1"""9 family respcns lb !lit ie9 Md/o r .-,r.iora 1 cons ide ra t ions to interfere with j ob. ------ I 16 ) Really know,, the J ob (i.e., clearly lfflers tands t,o.., to .-,rtom her Job including tasks to be ..,ctormed, pr!orl tie9 of tasks , -trrl Ix,., to alloo.te ti.me """""I tasks/ . IT7) Maintains 9eXUJlly neutral n,anflP_r en the job ;it all ti,,...,. ll8) Hit s "rlynam_t. c Inner dr ive" fi. e. , maJce., evPry sacrifioo of r1.,n,> ~ e ffo r t that ls ~ssary to suca,ed J ? ~88 '., . ~i We nee:! S<:ma lnform3t.lon that will allow us l:D evaluate the corpamblllty o f sa l e ""?"'-"? ease re"P()nd l:D the followl.ng quest!Qfl!I ii.bout ~ ch saleswanen yOJ hav-a Jes c d bed , by P 1a cing a check ln the appropr late sp3ce.: (1, What ls her age? 16-19 - --- - --- 20-24 -------- 25-34 ------ -- 35-44 45-54 - ----- -- 55 or older o::,n't ------ --knoW ( 2 > What is her e:luc.1tlon7 Hlgh school rot (Xff()leted Hl gh gchool graduate Serre c:nlleJt yr.,JrsPl fl your =~ny: (11 What ls your age? 25-34 20-24 <; <; or o lder -- ya,rqer than 20 - - 45-54 35-44 (2 , Bache l or' q df'I fr Pe Wh;, t ls yooir erluc.1tion7 o:>11-'le ted .:"'ICJTIP r. r-,"11 Ju, 1 tP loo("} rk, - - High schoOl rv.,t - - lllgh gchool gr,rluate Grr1du,-. t.P dA-Jt"eP Sore co [ l eg<> I e vel work f ,...-l\1 le nole DI Wh,t i s your sP..J< 7 , n c"N m m-:- is the typical cu.sterner in your business? (e.g., in- t>::rn!! O'.JnSurlo?r, ,....seller, industrial end-user). Typical O.ls toner: (10) On a scale o f O tD 100, evaluate t.ha CJVP_r.,11 selling effectiveness o f ~ch ""le""'?"'n Consider 100 tD represent sOJper1or perfor:manoe, itrld O to represent LIMo::?pt:,,t,[p perform.mce. Ol.-e.ra 11 Selling Et f ecti veness ScDre: ------ -- 'llWGC \UJ KR CDIP{Eril'C nus ~ll l APPENDIX C CHARACTERISTICS OF TIIE SAMPLE 291 APPENDIX C CHARACTERISTICS OF THE SAMPLE Three samples were defined by the type of c ustomer normally contacted. The samples are def ined on three levels: (1) that of the s a les J? Ob ' (2) that of the sales manager, (3) that of the saleswoman. The Sample of Selling Jobs The three samples a ll involve types of sales jobs th ,1 t traditionally have been perfo rmed primarily by male s . These jobs were classified according to the c haracteristics of the c ustomer normally contac ted: (1) or ganiza tional end users, (2) in-home consumers, and (3) resell e r s . ~ .2.tlling to Organizational End Users In thi s type of selling, customers are or~ani z<1ti.01wl d<>cls i 011 make r s who purchase products fo r use in their busLnesses . The Products that they buy include equipment .1nd .s11ppl Lc> s to bP used 292 -- in: (1) conducting the business (e.g., copy machines , telephone equipment), (2) manufacturing the firm's product (e. g ., electronics equipment, steel), or (3) augmenting of the firm's produc t (e.g., advertising). Also included in this customer group, are dec ision makers who do not actually purchase the products, but decide what Products will be purchased by others (e . g , physicians / pharmaceuticals, professors/ textbooks). 2.?mple Two ~ l ling to In-Home Consumers. In this type of selling, customers are in-home consumers who Purchase relatively expensive products for co nsump tion by themselves or by their families. In many instances , these customers do not seek out sellers, but are contacted directly by salespersons, in their own homes. Customers make major dec.isi.ons about things that are important to their persona l lives (e . g ., homes, insurance, financial planning, home protection sys t ems ,rncl encyclopaedias). ~le Three ~ling to Resellers In th:i.?~s t- ype o f se1 11. ng, c ustomer?s a re bus iness decision ma~crs who Purchase proc.lucts to be resold for profit. The products that they buy include such th.in os as fooc.l products, beverages , wearing 0 ~93 - apparel, books, greeting cards, cosmetics, phonograph records, and furniturr> The Sample of Sales Managers The actual respondents to the survey were sales managers. As seen in Table C-1, the demographic characteristics of the sales managers did not differ significantly among the three samples. These demographic characteristics were compared with 1986 national averages of: (I) "sales s upervisors and proprie t or s, " awl (2) "t otal executive, administra tive, and manage r ial personnel" (Bureau of Labor Statistics 1986). Overall, responding sales managers were Were older, better educated, with a higher percentage of males, th an the national aver ages . ~ ~ng to Or~aniza tional End Users In comparison with na t ional averages of "sales supe r visors and Proprietors," sales ma nagers ,,ho s upervised snleswomen selling to or ganizational end users, were ,nor e likely to be 1niddle-agcd (chi-squarc (2)=7 . 268 , p < . 026) and we ll-ed11cC1tcd (chi-squa re(2 ) =107. 808 , p < . 001) . They dicJ not differ Ln the dist ri?o, ution of the sexes (chi-sq uar2(l)=l . 240 , p In 294 - Table C-1 Demographic Characteristics of Responding Sales Managers Percentages Sample one Sample two Sample three To Organ'l To In-Home To End Users Consmers Resel l e r s ----- (n = 77) (n = 26) (n = 29) AGE: ---------------------------------------------------------- To 34 years 35 32 . 5 19.2 26 . 0 to 44 years 45 41.6 42.3 38 . S Years or older 20.0 38.5 27. 6 ---------------------------------------------------------------- ----- Chi-Square= 3.89163, df = 4, p > .4237 EDUCATIO;~-- --------------------------- ------ - ------- High School and Some College 27.3 38.5 31. 0 College Graduate 39.0 42.3 44. 8 Some Graduate Work 33.8 19.2 24.l -------------------------------------------- -- ------- - ---------- Chi-Square= 2.70791, df = 4 , P > . 6078 S-E-X-:- ------------------------------------------------------ ---- --- ,"1ALE 75.3 92. 3 89 . 7 - Female 24.7 7. 7 20 . 3 --------------------------------------------------- - ----- -- Chi-Square= 5.30294, df = 2 , p ' . 0705 ------------------------------------------------------------ :!95 - --- ----- comparison with national averages of "tota l exec utive , administrative ' and managerial personnel," respondents we r e more l"k J. ely to be older (chi-square(3)=36.272, P < .001), well-educated (chi- square(3)=13.223, p < .004), and male (chi-square(l) =l0.993 , P < . 001). ~le Two ~ing to In-Home Consumers. In comparison with national averages of "s al e s s uperviso r s a nd Proprietors," sales managers who supervi sed sales wome n se ll i ng large-ticket i terns to in-home consumers were more l i kel y t o be middle aged (chi-square(2)=14.002, p < .003 ), well-e ducoted (chi-square(2) = 26.930, p < .001), and male (c hi-s riuo r e(l ) =6 . 388 , P < .0ll). In comparison with national averages of "toto l executive' administrative' a nd managerial pe r s onne l," r espond en t s were more likely to be olde r ( chi-squa r e(3 )=1 5. 719 , P < . 00 I ) , College graduates (chi-square(3)=8. BSS, p / ' . 03l) , ':I nd ma l e (chi-square(l)=l3.497, p < .001). ~Ple To-.-~~ ~Sneg t~o Re sellers In comparison with national a verages of "sales s upe r viso r s ,i nd Proprietors," sales managers ,,ho supe rvi. s ed s .:Jl eswome n se ll i ng pr 0 d ucts to resellers were ,nore Ii.kely to be : (1 ) colle:;c graJ uates (chi-square( 2 )=38 _8 77 , p < .001) and (2) ma l e (c hi-sq ua r e(l)=S . 566 , 296 - p < ? 018). They did no t diffe r from na t iona l a ve r ages i n age (chi-square(2)=.185, p > .912). In compa ri s on with na tiona l averages of "total executive, administra tive, and manager ia l personnel," these sales managers were more likely to be : (1) male (chi-square(l)=l2.901, p < .001) and college graduates (chi-square(3)=10.345, p < .016), but did not differ from na tiona l averages in age (chi-square(3)=7.432, p > .059). The Sample of Sales Women While the data were contributed by sales managers, the foc us of th e analysis was on the saleswomen who they described. As s e e n in Table C-2, the demographic cha r acte rist ics of the s .:ileswo;nen were found to differ significantly. As a result, the samples were not combined, but were analyzed separately and the r esult s comp~ired . The demographic characteristics of the saleswomen i n th e t hr ee samples are summarized in Ta ble C- 2 . ProporU.ona t ely m()rc saleswomen who sold to in-home consumers were in the J'.i yea r a nd older samples, whi le group than were in those i n th e o th e r the proportionately more s ales wome n who sold to r ese lle r s we r e i n 24 Propor t iona t e ly mor e s ;:il c s 1vo111c 11 1v ho year and younger group. sold to i? n d ustria 1 en d Li se r s were c olle'.iP. grod u:1 t e s 1vh.i l e r ese l l e rs had hL g h Proportionately more saleswomen who so ld t o 297 Table C-2 Demographic Characteristics of Saleswomen Percentages Sample one Sample t wo Sample three To Organ'l I n-Home To ----------- End Users Cnsumer s Resellers AGE: ---------------------------------------------------- To 24 Years 25 9. 9 .g ? .3 20 . 6 -34 years 35 63.9 20. 8 42 . 9 ------Y-e-ars or older 26.2 70. 8 36 . 5 ------------------------------------ - ---------------- ---- Chi-Square= 53.96865 , df = 4, p < . 0001 EDuc;:ri;;~------------------------------------------------------- H. Slgh School only 12. 6 18 . 3 35 . 0 Come College 17. 5 31. 7 15 . 0 ollege Grads 69.9 50. 0 50 . 0 ---------------------------------------------------- ------------ ------ Chi-Square= 22. 63343 , df = 4, p < . 0001 MA - ------------------ --------------- --------------------- . RITAL STATUS: Married 40.1 55 . 6 30 . 2 Single ------ 59.9 44 . 4 69 . 8 ------------------------------------------------------------- ____ Chi-Square= 9.35707, df = 2, P < . 0093 TE - ------- ------------------ ------------------------------- 'NURE ON JOB: ~ess than one year 17 . 1 26 . 1 20 . 6 M- 5 Years 65 . 8 52 . 2 58 . 7 -- ore than five years 17 . 1 21. 7 20 . 6 --------- ------------------------------------- Chi~Squa re = 4. 55278 ~-df = 4 , p > . 3363 (ns) ----------- ---------------- ------------- ----------- 298 school educations only. Proportiona t ely more saleswomen who sol d two to in-home consumers wer e married that those in the othe r groups, These differences were significant by chi-square t es t a t P < ?0 5, No significant differences were found among the three samples in length of time on the job, The demographic characteristics of the three samp les we r e compared with 1986 national averages of saleswomen of "commodi ties , exclusive of retail" (Bureau of Labor Statistics 1986 . Saleswome n in samples one and three were more likely to be young (samp le one : .001, sample three: chi-square(3)=115.744, p < one : and college gradua t es (sample chi-square(3)=11.089, p < . 011) .000); s am pl e t hr ee : chi-square(3)=83.780, p < a ven 1ge . T~c < .001) than the na tiona l chi-square(3)=40.579, p demographic characteristics of the s ale swomen in snmplc t~o did not differ from the national a verage (age : chi-sq uare (4)=8 . 155, 77 p > .0866; education: chi-square(3)=1.091, P > - 9) . 299 APPENDIX D ESTIMATION OF flfISSING DATA 300 APPENDIX D ESTIMATION OF MISSING DATA Missing data in the three individual stereotype scales and in the masculinity and femininity sub-scales of the BSRI were estimated using regression techniques. The technique was c hosen b1~ca use it made use of information about both the individual case anrl the other items in the scale. The use of the procedure is justified if: Cl) good predictors are available within the data set and (2) linearity a mon g the predictors can be assumed. It requLres th;it the variables be correlated with each other. Th e s tron g correlations among the items of each each scale supp<)rted the use of this procedure. Tabachniclc a nd Fidell (l98J) recommend t he use of the procedure with large sample sizes and with very llttle missing data. The number of cases that were missing from each item, ,nci therefore estimated, for the Individual Stereotype sculcs a11d the BSRI scales, are given in Table D-1 and D-'.!, respcct Lve l y . .\ s c-Jn be seen, only a few cnses were missing from any Ltem . T1w exception wn s the Lte:n "loves chi.l---------------------------------------------------------- - -- p~genvalue 6 . 7 5.4 1. 0 rcent f 60.5 05 . 7 Ex _0 Variance 33 , 5 P1a ined ------------------- --------F = Item from Feminini~;-s~~~;-;i-BSRI M = Item from Masculinity Scale of BSRT 316 APPENDIX F EVALUATION OF TIIE QUALITY OF STATISTICAL ANALYSES 317 APPENDIX F EVALUATION OF THE QUALITY OF STATISTICAL ANALYSES This that resu1 t from the i-.ray section discusses limitations statistica1 tests were applied in the testing of hypotheses . The ana1ysis of variance techniques that were used to test hypotheses four through The extent seven are subject to certain assumpt.Lons. to which the data conform to these assumptions indicates the deg ree of contJ?. d ence that can be placed in the precision of the resu1 ts. Violations of these assumption s can result in distorted readings of the true sio n1". fi.c ance leve1 and effect sizes (Tabachnic'c and Fidell 0 1983). Issues that relate to all four analyses nre discu.c;sed first , folloived by analyses of the extent to 1vhich the sol11tio11:::; con furm to the underlying assumptions of the (1) univnriate DTl : (1) r-he 318 use of "parametric" procedures with ordinal-scaled data, (2) the lack of "orthogonality" ("unequal n"), and (3) the presence of extreme observations ("outliers"). Parametric Procedures. Analysis of variance techniques are of interval- or parametric procedures and require the use ratio-scaled data in these analyses were only data. The ord 1? .nal-scaled. While it is not strictly correct to do so , social for science frequently use parametric techni17 ues researchers analyses 1. .nvolving ordinal data (e.g., Kerlinger 1973). Authoriti. es seem to agree that it is mathematically "safe" to apply parametd.c methods I(erlinger 1973). to ordinal data (e . g . , Churchil I 1983; Under certain conditions, beca use of their greater versatility, parametric methods may even be the better c hoice (i'1yers 1979). In this case the highly correlated dependent variabl es mc1ndated the ' use of through multivariate techniques that were only nva ilab Ie Parametric methods. Or th0gonality. The techniques assUllle orthogonality (L.e., that the samples cell are eq uJl, the assumption of sizes within each ,, equal n"). sex role types, The classification of salcs1-1omen in to resulted in unequal nU111be rs of s aleswomen assig ne d to each ee l L. ,, The SPSS x usinJ ,"1ANOVA progra111 corrected for "un equol n ,, regression" tee h nique which adjusts each term for every other tcn7 in the model. The most serio us limitation cesulting [rom the J /0 Presence of "unequal n" was an impact on violations of the other assumptions. Analysis of variance techniques are known to be very robust to violations of the assumptions when sample sizes are ~. but are less robust when sample sizes a re unequal. Extreme Observations. Analysis of variance is extremely sensitive to the presence of extreme observations ("outliers " ) Within each cell. The presence of extreme observations can cause clistortion of the solution and introduce error into the analysis . The data were checked for "outliers" and several were identi fieci. These extreme values were not the result of error, but were legitimate parts of the sample. These cases were, therefore, retained in the analysis (as recommended by Neter ;ind \?hlssermc1n l974). In order to make certain that the presence of these extreme observations did not distort the mathematical solution, the analyses were performed both with and without them and the results compared. No major differences were noted . Analysis of Variance Hypotheses four and six were tested 11s i.n g uni variate nnaly s l s of variance (!\NOVA). The assumptions that were involved were : ( 1) independ ence, (2) normality of distribution of error terms, and ( 3) J .:?O homoge ne ity of variance. It should be no t ed that a violation of the assumption of independence was inherent i n the design of the research since more than one observation was provided by a given respondent. Problems caused by this violation were tested by use of the Durban Watson statistic. The assumption of norrnali ty of cti? stribution of error terms wa s te s ted by the chi-sq uare statistic, while the assumption of homogeneity of variance was tested usin g Bartlett's Band Cochran's C Statistics. the ass umpt ion of Hypothesis Four The results of the residual s a na lysis for hypothesis four ,1re summarized in Table F-1. A Durbin Wat son test r eve;1le,i no s ignificant autocorrelation. A chi-sq ua re test detected no se ri