Citation
Dominance, gender, and leadership emergence

Material Information

Title:
Dominance, gender, and leadership emergence
Creator:
Hegstrom, Jane Louise
Publication Date:
Language:
English
Physical Description:
viii, 160 leaves : illustrations, forms ; 29 cm

Subjects

Subjects / Keywords:
Leadership ( lcsh )
Dominance (Psychology) ( lcsh )
Sex role ( lcsh )
Dominance (Psychology) ( fast )
Leadership ( fast )
Sex role ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 152-160).
General Note:
Submitted in partial fulfillment of the requirements for the degree, Master of Arts, Department of Sociology.
Statement of Responsibility:
by Jane Louise Hegstrom.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
23458326 ( OCLC )
ocm23458326
Classification:
LD1190.L66 1990m .H43 ( lcc )

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

DOMINANCE, GENDER, AND LEADERSHIP EMERGENCE by Jane Louise Hegstrom B.A., Indiana University-Purdue University at Indianapolis, 1979 A thesis submitted to the Faculty of the Graduate School of the University of Colorado at Denver in partial fulfillment of the requirements for the degree of Master of Arts Department of Sociology 1990 .., ....

PAGE 2

This thesis for the Master of Arts degree by Jane Louise Hegstrom has been approved for the Department of Sociology by 9/ :Lb/tD I Date

PAGE 3

Hegstrom, Jane Louise (M.A., Sociology) Dominance, Gender, and Leadership Emergence Thesis directed by Associate Professor W.I. Griffith This research examined the extent to which dominance and/or gender were predictors of leadership emergence in ad hoc groups. The subjects were pretested using the California Psychological Inventory to determine relative levels of dominance and tha Bam Role Inventory to determine gender role identification. A 3 X 2 X 2 incomplete factorial design was employed to test the effect, if any, of dominance (high, medium, low), gender and dyad composition (same-sex, mixed-sex) on leadership emergence. There was a total of five treatments, each of which included ten dyads. The data revealed that dominance was a predictor of leadership emergence but only in samesex conditions. In such cases, high dominant individuals assumed leadership in much greater proportions than their low dominant partners. However, gender appeared to be a more potent

PAGE 4

predictor in mixed-sex dyads where males assumed leadership at levels greater than would be suspected given dominance levels. In addition to assumption of leadership, this work examined initial desires for leadership, initial offers of leadership, and initial nominations for leadership. Neither gender nor dominance was a predictor of who offered to be leader. Explanations of these findings from attribution and expectation states paradigms were offered. Implications for women in leadership roles and changing job requirements were discussed. The form and content of this abstract are approved. I recommend its publication. Signed __ W.I. iv

PAGE 5

To Dick, Michael, Nancy and Tom with love: and to Lou in appreciation for her generosity and patience.

PAGE 6

CONTENTS Figures. .ix Tables ...................... x CHAPTER I INTRODUCTION ............. 1 II. REVIEW OF THE LITERATURE ........ 13 Sex Roles and Leadership 14 Conclusions .......................... 24 Dominance ................. 27 Dominance and Situational Variables ............................ 45 Dominance and Stability Within Dyads.......... .... 51 Predictive Validity of the CPI and the Dominance Sea 1 e .... 54 Conclusions .. ........... 57 Attribution and Expectation 5 ta tes ........................... ...... 63 Mixed-Sex Dyads ........ 65 Same-Sex Dyads ................. 71 Conclusions .......................... 72 General Conclusions ............. 73 I I I. METHODOLOGY ................ ?? Instruments ................... 80 Procedures ...................... 83

PAGE 7

Hypotheses ........................... 85 IV. RESULTS ............. 91 Leader Gender and Gender 94 Leader Selection and Dominance ................... 95 Leader Gender: Gender Composition of Dyad and Dominance Distribution in the D y ad . . . . . . 9 6 CPI Dominance Scores and Bem Sex Role Inventory Scores ..... oo99 Initial Desires for Leadership and Gender ........................... 100 Initial Desires for Leadership and Dominance .. oo The Offer of Leadership and Gender .. 103 The Offer of Leadership and Dominance ...... 103 Nomination of Partner for Leader-ship and Gendero Nomination of Partner for Leader-ship and Dominance ... ooo V o SUMMARY AND CONCLUSIONS ............ o 113 Same-Sex Unequal Dominance Conditions ........................... 114 High Dominant Female/Low Dominant Male Condition .. 117 High Dominant Female/High Dominant Male and Middle Dominant Female/ Middle Dominant Male Conditions ........................ 122 Leadership Determination by High Dominant Females in Various Dyadic Conditions ............. 124 vii

PAGE 8

Leadership Determination by High Dominant Males in Various Dyadic Conditions ......... 127 Leadership Determination by Low Dominant Males in Various Dyadic Conditions ................. 129 CPI Dominance Scores and Bern Sex Role Inventory Scores ........... 131 Suggestions for Future Research .................... 133 ConclLlsions .................. 135 APPENDIX A. CONSENT FORM ...................... 141 B. CAL1FORNIA PSYCHOLOGICAL INVENTORY DOMINANCE SCALE (CPI) .............. 143 C. BEM SEX ROLE INVENTORY ............ 146 D. SCRIPT OF VIDEOTAPED MESSAGE ........... 148 E. LABORATORY QUESTIONNAIRE .......... 149 F. SCOPE CONDITION CHECKLIST ............. 151 BIBLIOGRAPHY .. 152 viii

PAGE 9

FIGURES Figure 4 .1. CPI and BSRI Scores and Rankings 101

PAGE 10

TABLES Tables 4.1. Number of High Dominant Subjects Assuming Leader and Follower Roles ........ 93 4.2. Leader Outcomes .................. 98 4.3a. Dominance of Subjects in Five Dyad Treatments Offering to Be Leader .... 107 4.3b. Dominance of Subjects in Five Dyad Treatments Initially Desiring to Be a Leader ................................ 108 4.3c. Dominance of Subjects in Five Dyad Treatments Who Nominated Their Partner to Be Leader ...... 109

PAGE 11

CHAPTER I INTRODUCTION Leaders have a significant role in creating the state of mind that is the society. They can serve as symbols of the moral unity of the society. They can express the values that hold the society together. Most important, they can conceive and articulate goals that lift people out of their petty preoccupations, carry them above the conflicts that tear a society apart, and unite them in the pursuit of objectives worthy of their best efforts. Annual Report of the Carnegie Corporation The integration of women into positions of leadership has been a slow, arduous journey. Research clearly suggests that the traditional sex-role stereotype for women is one that is incompatible with leadership and management roles (cf. Braverman, Vogel, Braverman, Clarkson, and Rosenkrantz 1972; Offerman 1986; Powell and Butterfield 1979; Schein 1973, 1975). Males consistently enjoy the attribution of leader while females languish in the attribution of follower. This attribution poses an immediate dilemma because as women continue to enter the work force in increasing numbers, with educational backgrounds comparable to their male peers', and expecting to

PAGE 12

attain job equality (Stitt, Schmidt, and Price 1983:31), society will be forced to examine this issue. In 1987, for example, women held a 53.51. share of the work force and these numbers are steadily climbing. Many barriers prevent women from assuming leadership positions; however, the sheer numbers of women in the work force will exert great pressure for change. While women struggle to break through the traditional sex-role stereotyping that casts them as followers in this society, men and women alike question what the elusive role of leader actually entails. Decades of academic analysis have given us more than 350 definitions of leadership. Literally thousands of empirical investigations of leaders have been conducted in the last 75 years alone, but no clear and unequivocal understanding exists as to what distinguishes leaders from nonleaders (Bennis and Nanus 1985:4) Leadership research has not necessarily provided clarity on this subject. While all the leadership models have contributed to research, they also contributed to a proliferation of approaches and conflicted results. Stogdill (1974) concluded that: It is difficult to know what, if anything, has been convincingly demonstrated by 2

PAGE 13

replicated research. The endless accumulation of empirical data has not produced an integrated understanding of leadership. (p.vii) The pattern of leadership research has been one of discarding, extending, and introducing ideas as the limitations of existing ideas are realized (Calder 1977). Leadership research is criticized for its simplicity, its emphasis on personal attributes and personality traits, and its descriptive orientation which doesn't explain leadership processes. In its simplicity, the study of leadership offered the Great Man Theory which attempted to explain leadership on the basis of inheritance, survival of the fittest, and intermarriage which produced a class of leaders distinct from the lower classes (cf. Galton 1879; Woods 1913). A more current analogy to the Great Man Theory of leadership came from Jeffrey Pfeffer (1978) who discussed the attainment of leadership as nothing more than a phenomenon largely based on social class implications. While the Great Man theorists attributed leadership to individuals based on inheritance and class differences, Pfeffer's tenet was that attending the "right" schools in addition to an upper class family background provided a track 3

PAGE 14

for leaders which inevitably made the analysis of leadership a "mythological phenomenon." From the turn of the century to the 1940's, leadership studies turned their attention to the attempt to identify personal attributes and personality traits which characterized leader qualities. The trait approach focused on those characteristics or traits that separated "great leaders" from the masses (Hunt 1984:114) and examined areas such as intelligence, dominance, extroversion, verbal fluency, etc. Traits were measured in an attempt to describe a leader's personality and establish actual and/or potential leadership capabilities. The weaknesses of the trait approach were discussed by Gibb (1954), Hemphill (1949), Mann (1959), and Stogdill (1948). These authors found the trait approach to be unsuccessful in providing an understanding of universal personality qualities shared by leaders for a number of reasons: (1) the examination of a large number of traits did not provide a differentiation of leaders from followers; (2) the traits demanded in a leader varied from one situation to another; and (3) the trait approach ignored the interaction between the leader and the group (Stogdill 1974). 4

PAGE 15

With the trait approach in disrepute, the next orientation shift was to the actual behavior of leaders and their style of leadership (Calder 1977: 179) The leader behavior approach looked at how leader behavior impacted individuals and/or group outcomes. The behavioral approach was undertaken as a shift from the trait approach by those who were primarily interested in organizational leadership. The behavioral approach identified such issues as "task" and "maintenance" (interpersonal) behavior (Bales 1950), initiation of structure and consideration (Shartle 1952), and patterns of supervision related to productivity and morale (Kahn and Katz 1960). The behavioral approach addressed itself to the follower's evaluation of the leader and how that style affected behavior expectation and follower satisfaction. Although the behavioral approach pointed out some important aspects of leadership, its descriptive orientation did not provide an explanation of leadership processes (Calder 1977:180). During both the trait and behavioral eras, researchers sought to identify the "best" style of leadership. They had not yet recognized that no single style of leadership was universally best across all situations and environments (Chemers 5

PAGE 16

1984) Fiedler (1964, 1967, 1971a, 1971b) stated that the "contingency model," which attempted to address leadership situational demands, acknowledged that different behaviors would be manifested to meet situational variables. Research in organizational behavior welcomed a model that assisted in providing some prediction of effective and ineffective leaders. Contingency research was concerned primarily with follower-outcome such as performance and satisfaction and these outcomes were functions of "contingencies" facing the leader (Hunt 1984:114). While empirical support for the contingency model was mixed (cf. Chemers and Skrzypek 1972; Fielder 1971b; Graen, Alvares, Orris, and Martella 1970; and Hunt 1967), the contingency model has been the predominant leadership approach since the 1960's. A number of other contingency-oriented leadership approaches also addressed the relationship of leadership decision-making style to group performance and morale (Chemers 1984:97). Some of these "second generation" contingency models included Normative Decision Theory, Path-Goal Theory, and transactional approaches. The Normative Decision Theory (Vroom and Yetton 1973) looked primarily at leadership styles which 6

PAGE 17

could be broken down to autocratic, consultative, and group style decision-making. The leadership studies which dealt with authoritarian and democratic leadership (cf. Adorno 1950; Bell 1950; Kutner 1950; and Lewin 1948) were forerunners to the Normative Decision Making Theory. Path-Goal Theory studied the effects of the Leader Behavior Description Questionnaire categories of consideration and structuring behavior (Chemers 1984:99) While structuring and consideration grew out of the leader behavior approach, these variables were incorporated into the contingency model of Path-Goal Theory to help account for effects on subordinate psychol'ogical states when the subordinate's task is unclear and/or difficult, that is, unstructured. The structure provided by the leader helps to clarify the path to the goal for the subordinates. On the other hand, consideration behavior will have its most positive effects when subordinates have a boring or distasteful job to perform. (Chemers 1984:99) Essentially, Path-Goal Theory involved studies of follower satisfaction rather than group performance. Transactional exchange provided an even broader view of leadership, one which fully recognized the multiple contingencies of the leadership process. Hollander and Jullian (1968, 1969, 1970) developed a "transactional" approach to leadership which 7

PAGE 18

emphasized the exchange relations between leaders and followers. The leader provided resources in achieving group goals and in return received status, esteem, influence and legitimacy (Calder 1977:180), Hollander amd Jullians (1968) "idiosyncrasy credit" addressed the process of social exchange as a legitimation of leadership. Critics of leadership research have stated that the literature is in disarray, that researchers are entranced by a handful of models, that it is too complex, too simplistic, and covers a range of situations which do not require leadership. Lombardo and McCall (1978) criticized the quality of research and called it short range and atomistic, focused on leader-group relations and passing over leader-group system relationships with little study involved in what is actually done in leadership situations. Louis Pondy (1978) stated that there should be more effort given to the documentation of variety in leadership rather than uniformity. This study does not dispute McCall and Lombardo's (1978) claim that leadership theorists need to "rediscover the phenomenon of leadership" and perhaps one method may be to document more variety in leadership rather than uniformity (Pondy 1978) However, as we have seen in the brief 8

PAGE 19

description of leadership models in this chapter, there are many faces to the phenomenon of leadership. Leadership as a role examines leader behavior, traits, decision-making styles, effectiveness, management of meaning, and leader satisfaction. In turn, the role of follower examines follower satisfaction, performance, morale, and the follower's evaluation of the leader. These faces include the initial research on leadership which is most interested in who becomes a leader while later research focuses on productivity and satisfaction. While leadership studies may seem to be in a state of disarray, the importance of the phenomenon of leadership cannot be denied. Those that lead possess powers which can be translated into many different forms. It [power] carries with it a host of connotations incurred over thousands of years. These implications--including avarice, insensitivity, cruelty, corruption--have led in aggregate to the disregard and disintegration of power across the board. In other words, power is at once the most necessary and most distrusted element exigent to human progress. (Bennis and Nanus 1985:16) While this study does not deal directly with the issue of power, leadership is certainly tied into power and its use. In other words, conditions that 9

PAGE 20

allow some individuals to exercise power over others is a phenomenon well worth investigating. Equally well worth the investigation is the reality that one cannot lead without followers. Present and future problems will not be solved without successful leadership and an effective, motivated followership. Without this marriage of dynamic and influential leadership which has the capability to keep the followership aware of their shared values and goals, there will be little impetus for social change. The quality and direction of social change is what makes the study of leadership still a critical endeavor. The review of the literature in Chapter II will examine some of the basic works in the area of leadership and gender research. While this chapter will deal with variables that seem most relevant to gender and leadership studies, dominance and sex role identification will be analyzed as potent variables in leadership determination. Further, this chapter will present a cognitive approach to leadership research. A cognitive approach is a contingency model which offers a presentation of attribution/expectation states theory. Chapter III covers the methodology employed in this study. This chapter will present an outline of 10

PAGE 21

procedures, an operationalization of the variables and a presentation of the research hypotheses. This study chooses to keep its focus narrow enough to examine one area of leadership: emergent leadership in same-sex and mixed-sex dyads of unacquainted subjects who will be asked to choose a leader between themselves in order to perform a task. Chapter IV will present an analysis of the data. This presentation of findings will include the hypotheses of this study, a justification of the statistical tests that are used, and the statistical outcome of each hypothesis. The focus of Chapter V is to provide a summary and conclusion of the study. The findings from Chapter IV will be fully elaborated with suggestions for future research. It is the goal of this study's experimental treatment of gender and dominance to advance the state of knowledge in leadership research in two ways: (1) attribution/expectation states theory, rather than sex-role prescriptions, will frame the analysis of gender and leadership and (2) sex-role identification as measured by the Bern Sex Role Inventory and dominance as measured by the California Psychological Inventory Dominance Scale 11

PAGE 22

will be analyzed as competing complimentary in the of While the variables of gender, dominance, and sex-role identification will the state of knowledge of leadership and gender studies, the for is not merely what variables affect the of leadership. Research must be at the underlying of the leadership 1977:202). 12

PAGE 23

CHAPTER II REVIEW OF THE LITERATURE The review of literature will examine three areas that contribute to the understanding of gender and leadership research: sex roles; dominance; and attribution/expectation states theory. Each area will contain a general review of relevant literature. The general review will be followed by those studies which are closely related to the study at hand. The review will begin with sex-roles and leadership research. A general review of the sexroles and leadership research will frame the focus of this study which will be to replicate partially and extend Megargee's (1969) work entitled "Influence of Sex Roles on the Manifestation of Leadership." Review of the trait, dominance, will be examined next. Megargee's (1969) study, along with three replications, will examine dominance as a predictor of leadership. Other reviewed literature will explore how the trait dominance can provide either stability or instability for dyadic

PAGE 24

structures. The importance of situational variables in relationship to dominance will also be presented. The concluding section of surveyed literature will involve attribution/expectation states theory. This section of the review will provide a competing perspective to show how this approach can inform traditional leadership studies. Sex Roles and Leadership There was voluminous research on leadership and the extent to which sex-role stereotypes influenced the evaluation of leadership behavior, follower satisfaction, performance of task, and expectations of leaders and followers. A subject's expectation of leader competency was based on sex-role prescriptions and these perceptions and expectations informed the phenomenon of leader emergence. This study was only concerned with leader emergence in dyadic structures based on various levels of dominance, gender, and gender role. Some twenty years ago case histories of discussion groups in a series of Minnesota Studies examined groups that were all female, all male, and groups which contained one man or one woman. The Minnesota Studies found that women rarely emerged as 14

PAGE 25

the leader in a coeducational group which contained two or more men 1Bormann 1969:219). Bormann referred to "the woman contender" as a "special problem" in his publication and noted that not only did the woman rarely emerge as the leader in a coeducational group but men usually refused to follow directions given by a woman in the presence of other men. Some expressed the opinion in their interviews, questionnaires, and diaries that women should obey and men should lead. In some groups where a woman was in strong contention, the men even expressed doubts as to the advisability of women getting higher education. (Bormann 1969:219) A number of other studies have described societal expectations for male and female behavior. Women were expected to be dependent on others, sympathetic, compassionate, and humanitarian (Tyler 1965). Milner (1965) found that expectations for women also included spiritual values, concern for others, and artistic inclinations while male behavior was oriented toward power, prestige and initiative. Evaluations of leadership behavior provided a rich display of sexual stereotyping. Research suggested that women were not perceived by male managers as having the requisite temperament for 15

PAGE 26

leadership (Bowman, Worthy, and Greyser 1965) and Gilmer (1961) found that over 651. of male managers believed women would be inferior to men in supervisory positions. However, other studies indicated that women in leadership positions did not function differently than males in similar positions (cf. Bartol 1973; Doll 1965; and Martin 1972). Rosen and Jerdee (1973) also examined behavior in terms of evaluation of the following supervisory styles: threat, reward, friendly-dependent, and helpful. Evaluations of supervisory styles were determined by some subjects who were presented with a situation involving a female supervisor and some with a situation involving a male supervisor. Subjects were not aware that a comparison of male and female supervisors was involved. Rather, the task as it appeared to them was simply to evaluate the propriety and potential effectiveness of four alternative supervisory approaches that were being considered by the supervisor depicted in the case description. (Rosen and Jerdee 1973:45) Rosen and Jerdee (1973) examined the effects of a supervisor's sex on subordinate evaluation of the supervisor's effectiveness. The results indicated that while sex-role stereotyping impacted on supervisory evaluations, not all four styles were affected. 16

PAGE 27

The reward style was rated as more effective for male supervisors than for female supervisors, while a friendly--dependent style is rated as more effective for supervisors of either when used with subordinates of the opposite evaluations of the threat and helping style did not differ for male and female supervisors. (Rosen and Jerdee 1973:47) Some questions arose from Rosen and Jerdee's (1973) assertion that these results would occur regardless of the sex of the evaluators and their occupational status. The authors drew their population of males and females primarily from undergraduate business students and female banking supervisors. The female business students could not be categorized as a typical sample; in fact, they would be fairly described as nontraditional women. In 1971 women majoring in business pursued a nontraditional occupation. A sample of more traditional women could clearly elicit a stronger, stereotypical view of the four supervisory styles based on men's and women's of behavioral roles. Bartol and Butterfield's (1976) study, like Rosen and Jerdee's (1973) study, stereotyping of supervisory behavioral styles. Bartol and Butterfield presented four leadership 17

PAGE 28

styles in a story which described a manager using a certain style in a particular situation. Each story had two versions, one in which the manager was a male and one in which the manager was a female. In every other respect the versions were identical. The sex of the subordinates in each story was purposely made ambiguous in order to avoid confounding effects On back of each story was a set of eight questions designed to reflect a variety of perspectives for evaluating the effectiveness of a particular leadership style (Bartol and Butterfield 1976:448-449) The four leadership styles (derived from the Leader Behavior Description Questionnaire) were initiating structure, production emphasis, consideration, and tolerance for freedom. The consideration subscale used to examine leader behavior took into account the extent to which a supervisor showed consideration for the personal needs and feelings of subordinates (Sheridan, Kerr and Abelson 1982:132). Initiating structure was that identifiable behavior which involved a clear definition of the leader's role, and let followers know what was expected. Leadership style which was high in production emphasis was one in which there was pressure for productive outcome. Tolerance of freedom allowed followers scope for initiative, decision, and action (Stogdill 1974). 18

PAGE 29

Bartol and Butterfield's (1976) research indicated that managers of both sexes experienced sexual stereotyping. The evaluators of leaders gave women managers higher scores than male managers when women used the consideration style. Male managers scored significantly higher on the initiating structure style than women managers using that same style. Tolerance for freedom and production emphasis found neither male nor female managers scoring significantly higher. The fact that identical leader behavior was evaluated differently depending upon whether the leader was male or female indicated the existence of strong sex stereotyping for managers (Bartol and Butterfield 1976). Jacobson and Effertz (1974) examined the effect sex-roles had on the perceptions that male and female leaders and followers have of themselves and of one another. The task was the same task used by Fleischer and Chertkoff (1986) and required that the group leader be given a diagram of an arrangement of dominoes and the followers were given a set of dominoes. The leader sat at a separate table in the same room with his back to the followers. The followers could not see the leader's diagram and the leader could not see the followers' dominoes. The followers were required to reproduce the arrangement in the diagram based solely on the leader's verbal 19

PAGE 30

instructions. (Jacobson and Effertz 1974:387) The subjects were divided into four types of three subject groups: male leader with male followers, male leader with female followers, female leader with female followers, and female leader with male followers. The perceptions that were analyzed were perceptions of performance, success, anticipated success, and enjoyment of the leadership role. Jacobson and Effertz (1974) found that leaders of both sexes tended to rate women followers' performance worse than male followers' performance. The authors concluded that cultural expectations dictated that women are followers and, as a result, there will be greater expectations for women in that role. For that reason, women will be rated more harshly than men in the role of follower. Followers of both sexes perceived leadership as a male role; however, the male leaders' performance was rated worse than the women leaders' performances. Again, the authors believed that cultural expectations ascribed leadership to men and, as a result, followers had greater expectations of males as leaders and rate male leaders more harshly than female leaders. 20

PAGE 31

There was no difference between male and female leaders in terms of enjoying the role of leader. Cultural expectations seemed to dictate that because males were ascribed the role of leader and have generally more experience in that role, men would enjoy the role of leader more than women. Two explanations were presented for this finding that there was no difference between male and female leaders in terms of enjoying the role of leader: (1) women so seldom get to be leaders that they enjoyed the privileged opportunity and/or (2) perhaps men assumed the role of leader because it was thrust upon them rather than being a leader for the pleasure of it (Jacobson and Effertz 1974). Finally, the study showed that male leaders were more likely to attribute greater hypothetical success to themselves than were female leaders (Jacobson and Effertz 1974). It should be noted that while the experimental task was presumed to be sex-neutral, it was a task deliberately designed so that it could not be completed in the allotted time. Because there were no differences in the amount and quality of output, the authors concluded that any perceived differences in performance stemmed from perceptions based on sex-role stereotyping rather 21

PAGE 32

than actual fact (Jacobson and Effertz 1974). The authors further concluded that while males anticipated being more successful in the role of leader than females anticipated success for themselves in that role, relative failure of the task may have caused males to underrate themselves in terms of actual success, and this may have explained the results of perceptions of performance and success. Gerber (1988) examined whether the leader-follower relationship with marital roles accounted for the stereotypic differences between the sexes. The methodology used by Gerber was to have subjects rate an adult married couple, Ann and Bob, according to information provided about the type of leadership in their marriage. The type of leadership varied over three conditions. In the "man is leader" condition, Bob was described as the leader in the marriage (there was no mention of Ann s role). In the "woman is leader" condition, Ann was described as leader (there was no mention of Bob's role). In the "both are leaders" condition, both Ann and Bob were described as equal leaders. (Gerber 1988:654) Gerber organized people according to their level of agentic and communal values. The "agentic" personality traits involve self-assertion, self-expansion, and the 22

PAGE 33

urge to master. Thus, they reflect the amount of influence that an individual exerts through his or her self-assertive acts. The "communal" personality traits involve concern for others, a desire to be at one with others, and accommodation. (Gerber 1988:650) Gerber operationally defined the leadership role in marriage to those who possessed power being the leader with the nonpowerful person being the follower. Gerber stated that regardless of sex, the powerful person was rated stronger on agency than on communion and the nonpowerful person was rated stronger on communion than agency. Further, cultural stereotypes have consistently viewed males as strong in agency and weak in communion while females are viewed as strong in communion and weak in agency (cf. Bakan 1966; Bern 1974; McKee and Sherriffs 1957; Rosenkrantz, Vogel, Bee, Broverman, and Braverman 1968; Spence, Helmreich and Stapp 1975) Gerber (1988) found in analyzing powerful and nonpowerful roles that men and women were perceived in very similar ways when they were assigned a leader or follower role, that is, there was no differentiation between them in commLtnal or traits. 23

PAGE 34

However two differences associated with female as compared with male, leadership were found. When the woman was the leader, there were fewer communal traits within the relationship as a whole, indicating that there was less mutual accommodation between the dyad members than when the man was the leader. In addition, the male follower was perceived as having fewer socially desirable characteristics than his female counterpart. (Gerber 1988:658) Conclusion Rosen and Jerdee (1973) found sex-role stereotyping impacted on the evaluation of supervisory styles of threat, reward, friendly-dependent, and helpful. Evaluations were not different for male and female supervisors with regard to threat or helping styles. The reward style, however, was rated as more effective for male supervisors than female superviors. The friendly-dependent style was rated effective for both male and female supervisors when this style was used with subordinates of the opposite sex. Bartol and Butterfield (1976) found sexual stereotyping occurred with regard to evaluation of leadership styles. Ev-aluators of leaders rated males higher on initiating of structure style than women and rated women higher on consideration than men. However, tolerance for freedom and production 24

PAGE 35

emphasis found that neither male nor female leaders scored significantly higher. Jacobson and Effertz (1974) studied the perceptions that male and female leaders and followers had of themselves and one another in terms of perceptions of performance, success, anticipated success, and enjoyment of the leadership role. Leaders of both sexes rated the female follower's performance worse than the male follower's performance. Followers of both sexes perceived leadership as a male role but the male leader's performance was rated worse than the female leader's performance. The authors concluded that these findings stemmed from the belief that because followership was ascribed to women, leaders had greater expectations of women as followers. As a result, women followers were judged more harshly than male followers. Likewise, leadership was perceived as a male role, thus a male leader's performance was judged more harshly than a female leader's performance. Male and female leaders enjoyed the role of leader equally well. Jacobson and Effertz (1974) speculated that women enjoyed the privileged status 25

PAGE 36

while men assumed leadership because it was expected and not necessarily because it was desired. Finally, males attributed greater hypothetical success to themselves than did females. There was no difference in the amount or quality of output relative to the task so any perceived differences in performance was a result of sex-role stereotyping. Gerber (1988) examined leadership in marriages. She defined the powerful person as the leader and the nonpowerful person as the follower. Regardless of sex, the powerful person was rated stronger on agency and the nonpowerful person was rated stronger on communion. Men and women were perceived similarly when they were assigned a leader or follower role, that is, there was no differentiation between them in agentic or communal traits. When women were leaders, however, there were fewer perceived communal traits within the relationship which indicated less accommodation. FLirt her, when the male was follower, he was perceived as having fewer socially desirable characteristics than the female follower. 26

PAGE 37

Dominance The studies which follow will review the relationship of dominance and situational variables, stability within dyads, prediction of leadership, and the predictive validity of the Dominance scale of the California Psychological Inventory (CPI). Surveyed literature will examine how dominance interfaces with situational variables, stability of dyads, and leadership prediction. While the differentiation between a person with a high need for dominance and a person who scored high in dominance was not an issue this study examined, there were indications that a high need for dominance was related to leadership (Bartol 1974:226). Mann (1959) cited 12 studies which analyzed how leadership status was associated with a need for dominance and found that in 731. of the studies there was a positive relationship between need for dominance and leader status; and in 421. of the results the relationship was both positive and significant (Bartol 1974:226). Bormann (1969) also stated that people with high esteem needs will be the contenders for the leadership role. At the same time that Bormann (1969) looked at role emergence in groups that were same-sex and 27

PAGE 38

mixed-sex, Edwin I. (1969) examiNed role emergence in dyads with the additional variable of high or low dominance. This experiment required the administration of the California Psychological Inventory Dominance Scale (CPI) and subjects that scored high and low became eligible for the experimental portion of the study. The dyads consisted of high and low dominant men (Group I), high dominant men and low dominant women (Group 2), high dominant women and low dominant men (Group 3), high and low dominant women (Group 4). The subjects within the dyads were at least 20 T score points apart on the Dominance Scale. This study's interest paralleled Megargee's (1969) study which investigated how social sex-role prescriptions influenced the expression of leadership by high dominant men and women (Megargee 1969:377) when paired with low dominant men and women. The dyad consisting of the high dominant woman and the low dominant man held the greatest interest in the present study. It was this dyad that contained a probable sex-role conflict. It seemed likely that social role prescriptions would actually inhibit high dominant women from assuming leadership when paired with low dominant men 28

PAGE 39

(Megargee 1969:377). This is, in fact, what Megargee discovered when only 25/. of the high dominant women paired with low dominant men assumed the role of leader. Megargee (1969) found that in his intended sex neutral two person task (Study II), when high dominant women and low dominant men decided between themselves who should be the leader, high dominant women made the final decision more often than the high dominant subjects in any other group, and 91/. of the time she appointed her low dominant male partner leader. The high dominant woman appeared to show her dominance by making the appointment of leadership rather than assuming leadership. This was very different from the behavior of high dominant individuals in other dyads who never appointed their partner leader more than 33/. of the time (Megargee 1969). Except in the sex-role conflicted dyad of high dominant women paired with low dominant men, high dominant subjects usually appointed themselves leader. When the low dominant subjects made the decision, they usually appointed their partner leader (Fleischer and Chertkoff 1986). Megargee found that the decision-making process was not a result of greater assertiveness by low 29

PAGE 40

dominant men, rather, greater reluctance on the part of high dominant women to assume leadership. The low dominant men in Group I (high dominant men paired with low dominant men) and the low dominant men in Group III (high dominant women paired with low dominant men) were nearly identical in appointing themselves leader (431. and 401. respectively). Megargee (1969) made the claim that the prediction of leadership should not only consider the level of dominance that an individual possessed, but also the effect of social roles on the overt expression of dominance (Megargee 1969:381). It was also important to recognize that task selections could either inhibit or facilitate the expression of dominance (Fenelon 1966). Megargee's (1969) Study I employed a masculine task which instructed the subjects to decide who should be the leader and who should be the follower so that a machine might be repaired. Colored nuts and bolts were located, removed with a wrench, and the follower obeyed the leader's instructions by holding bolts in place while the leader removed bolts. Study II was a replication of Study I except that Megargee's intention for Study II was that the 30

PAGE 41

task be sex neutral. Like Study I, Study II required that dyad participants decide who would be the leader and who would be the follower in order to complete the Stroop Color-Word Test (Stroop 1935). Thus the leader must transmit as much information to his subordinate as possible in the time allowed, as if giving orders in a crisis. The follower must record what his leader tells him. (Megargee 1969:379) Megargees (1969) Study I and Study II produced nearly identical results in the high dominant female and low dominant male dyads. High dominant individuals generally assumed leadership over low dominant individuals in same sex dyads, but in the high dominant woman paired with the low dominant man, only 25f. of the high dominant women assumed leadership. Megargees conclusion concerning this dyad was that the poor showing of the high dominant woman in assuming leadership was not limited by the highly masculine tasks of Study I because nearly the same results were evidenced in the presumed sex-neutral task of Study II. There was a legitimate question surrounding the sex neutrality of Study II. The conclusion that the task was sexually neutral may be flawed if one looked at the act of dictation and/or transmission 31

PAGE 42

of information and the attendant recording of that information. This process had strong sex-role prescription implications. This clerical task closely resembles the situation of a boss dictating to a secretary, roles that in our society are still associated with sex role differences. (Carbonell 1984:46) Carbonell (1984) replicated Megargee's (1969) masculine task (Study !) and extended the study by introducing a feminine task. Carbonell was interested in whether the passage of 15 years would find significant differences in the behavior of high dominant women who had, in Megargee's study, failed to assume the leadership role when paired with low dominant men. (Carbonell did not replicate the traditional sex-role composition of high dominant men and low dominant women and, like Megargee, she did not require subjects to complete the task after making their choice for leader.) Carbonell (1984) found, in replicating Megargee's (1969) Study I (masculine study), that despite the passage of 15 years High Do women when paired with Low Do men are stili significantly less likely to assume a leadership role than High Do men or High Do women who were paired with partners of the same sex. (Carbonell 1984:45) 32

PAGE 43

In Carbonell's (1984) study, 301. of the high dominant women paired with low dominant men assumed leadership while in Megargee's (1969) study 251. of the high dominant women paired with low dominant men assumed leadership. Carbonell found, in contrast to Megargee's study, that low dominant men exhibited more assertiveness in their desire to be the leader. Low dominant men paired with high dominant women in Carbonell's study actually were more likely to nominate themselves as leader. Carbonell's (1984) Study II was an extension of Megargee's study in that it was designed around a task stereotypically associated with women. Megargee's (1969) machine-like structure was used, but the nuts and bolts were removed and a panel of 7 X 7 patterns of holes which were surrounded by colored rings (red, blue, yellow or green) were visible from only the outside of the box (Carbonell 1984). Two subjects were needed, a leader to stay outside the box to call out the location of the colored holes and a follower inside the box to attach the correct button color to the correct colored hole. The follower was instructed to obey the leader's commands and put the needle through the 33

PAGE 44

buttonhole as directed by the leader (Carbonell 1984:47). Carbonell (1984) found that when presented with a feminine task, a majority of high dominant subjects in all treatments except the dyad composed of the high dominant female paired with the low dominant male, emerged as the leader. In the dyad composition of high dominant women paired with low dominant men, high dominant women were more likely to assume leadership over low dominant men in the button-sewing task than in the nuts and bolts task but the differences were not statistically significant (Carbonell 1984). As with Megargee's (1969) study, dominance did not predict in the high dominant female paired with the low dominant male dyad. Carbonell (1984) made the claim that the results of Study II (feminine task) indicated that the "Megargee effect" may be specific to task. One could argue that the button-sewing task cannot be assumed to be a feminine oriented task. While the tools of needle and yarn are typically associated with female work activities, there was still the dictation (command) and subsequent following of instructions (pushing a needle through a hole). The 34

PAGE 45

significance of a leader's command and a follower's obedience to the command may take precedence over the feminine associated tools involved in the task. Carbonell conceded that the gender of most tasks probably was not perceived as clearly feminine (Fleischer and Chertkoff 1986). The overarching issue was not the gender orientation of the task, rather, an issue of who would give the command (leader) and who would carry out the command (follower). Nyquist and Spence (1986) also replicated and extended Megargee's (1969) study. Their presumed sex neutral task (for replication of Megargee's Study II) was a board game called MasterMind. The object of the game is to determine a sequence of four colored pegs hidden from view. Players propose a series of sequences, after which they are given information about the number of matches, until the hidden sequence is matched exactly. (Nyquist and Spence 1986:88) No particular dyad composition performed the MasterMind game better than another and none of the subjects had ever played the game. In an attempt to assess the sex neutrality of the MasterMind game, the authors asked 49 men and 43 women introductory psychology students to read the instructions and on a 9-point scale the subjects 35

PAGE 46

indicated how well they thought men and women would perform and whether men or women would play the game better. The men thought that men might perform slightly better on the task while the women believed that men and women would perform equally. well. Overall, there were no significant differences on male and female judgments of performance (Nyquist and Spence 1986). Nyquist and Spence's (1986) study differed from Megargee's (1969) and Carbonell's (1984) study in three ways: (1) the task was actually performed; (2) each subject completed a questionnaire which probed the participant's satisfaction with their partner; (3) and each subject also completed a 15-item Attitudes Toward Women Scale (AWS) (Spence and Helmreich 1978) which measured attitudes about the rights and _roles of women vis-a-vis men (Nyquist and Spence 1986). Not surprisingly, the results of the AWS found women assuming leadership roles in both types of mixed-sex dyads were more likely to subscribe to egalitarian views than were women who assumed follower roles Thus, nonconventionality in the women's (but not the men's) sex role attitudes attenuated the effects of implicit demands that the man be leader. (Nyquist and Spence 1986:90) 36

PAGE 47

The partner evaluations contained a number of questions referring to pretask and posttask position preference but the dyad of greatest interest, the high dominant female followers of low dominant male leaders found that the high dominant female followers were more dissatisfied with their position than the low dominant females paired with high dominant men. There were, however, no differences between the posttask position preferences of these high dominant followers in other types of dyads (Nyquist and Spence 1986:92). In terms of the actual performance of the task, the high dominant female followers of low dominant male leaders displayed more dominance in their behavior than did low dominant female followers of male leaders but, conversely, they were no less dominant than high dominant followers in same-sex dyads (Nyquist and Spence 1986). Except for the high dominant women paired with the low dominant men, Nyquist and Spence (1986) found results similar to Megargee's (1969) study when examining high dominant persons paired with low dominance persons; 73Y. of the high dominant men assumed leadership in same-sex dyads and 90Y. of the high dominant men assumed leadership in mixed-sex 37

PAGE 48

dyads. In the high dominant women paired with the low dominant men dyads, high dominant women assumed leadership 35/. of the time. Conversely, the man became leader in 65/. of the pairs, thus showing the greater contribution of implicit role demands than personality in the instance in which these two factors are in opposition high-dominant women in mixedsex dyads .. made the initial nomination more frequently than did high-dominant men and women in the other types of pairs, but also were far more likely directly or indirectly to appoint their partner as leader. (Nyquist and Spence 1986:90) Just as Megargee's and Carbonell's (1984) studies indicated, Nyquist and Spence found that when the trait dominance and role expectations came into conflict, the high dominant woman displayed her dominance by appointing the low dominant male as leader. In short, their dispositional characteristics leaked into the leadership negotiations, even if in the service of traditional expectations that they be in a subservient role. (Nyquist and Spence 1986:93) Fleischer and Chertkoff (1986) also performed a replication and extension of Megargee's (1969) study. They were interested in primarily four questions: (1) could Megargee's findings be replicated after the passage of 17 years; (2) would 38

PAGE 49

regional differences affect subject responses; (3) would feedback of better performance to the partner less likely to become leader influence leader selection and; (4) what would be the ramifications on the leadership decision if the subject's initial desire (or lack of desire) for leadership were known prior to the decision? Fleischer and Chertkoff's (1986) results replicated Megargee's (1969) results in all dyads except the high dominant women paired with the low dominant men. The high dominant women assumed the leadership position 501. of the time compared to Megargee's 251.. Fleischer and Chertkoff (1986) used the instructions for a dominoes task borrowed from Jacobson and Effertz (1974). The instructions involved the placement of dominoes into a predetermined pattern but, like Megargee (1969) and Carbonell's (1984) study, the subjects were not required to perform the task after a leader had been selected. Jacobson and Effertz and Fleischer and Chertkoff considered this task to be sex neutral. Again, there was the question of the actual sex neutrality of the task because the followers were required to reproduce the arrangement in a diagram 39

PAGE 50

based solely on the leaders verbal instructions (Jacobson and Effertz 1974). There was still the dictation (command) and subsequent following of instructions which have, as discussed above, strong sex-role prescription implications. Not only does the dictation of instructions and subsequent following of instructions have sex-role implications, but a number of studies indicated that there was a sex-linked perception of expectations of task success. Feather (1968, 1969) found females less confident than males about their ability to solve anagrams; Crandall (1969) had shown that elementary school girls have lower expectancies of success than boys on tasks involving mazes, memory, estimations of quantity, logical relations, spatial relations and numerical skill. The task of arranging dominoes on a grid involved spatial relations, numerical skill, memory, estimations of quantity and logical relations. If females perceived themselves as being less successful than boys in the areas mentioned above, then the dominoes grid task was not sex neutral as claimed. Fleischer and Chertkoff (1986) go on to suggest that the increase in high dominant females paired with low dominant males in assumption of leadership 40

PAGE 51

versus Megargee's (1969), Carbonell's (1984), and Nyquist and Spence's (1986) results may be due to changes in attitudes toward women over time and/or the more liberally perceived environment of the Midwest versus Carbonell's southern university setting and Megargees, Nyquist and Spence's southwestern university environment. Fleischer and Chertkoff stated that Carbonell's, Nyquist and Spence's results did not differ significantly from the 221. result obtained by Megargee, but 301. to 351. is about midway between Megargee's 221. and [our] 501. figure. It may be, then, that about half of the difference between Megargee's results and [ours] is due to a general increase in acceptance of women as leaders across the country and about half is due to regional differences in such attitudes. (Fleischer and Chertkoff 1986:98) The difficulty with stating that half the differences between Megargee's (1969) studies may be due to attitudes based on regional differences pointed to the dilemma of external validity which is concerned with the generalizability of research finding to and across populations of subjects and settings (Kirk 1982:21). There was an empirical question of whether people from different regions of the United States displayed distinct or differing attitudes concerning women. Differing 41

PAGE 52

attitudes towards women may be explained more by class and/or ethnic group value systems than by geography. People from other social classes, nations, and ethnic groups have a strong expectation for male dominance, for example, working-class people may have difficulty conceptualizing an equal relationship. They may assume that "equality" means the wife is really dominant (Gerber 1988:663) The third question Fleischer and Chertkoff (1986) raised was whether feedback of a better performance by the person usually designated as follower in Megargee's (1969) study would make a difference regarding leader choice. The high dominant subjects in all dyad compositions except the high dominant female paired with the low dominant male received feedback of poorer performance than their partner while all low dominant subjects in. addition to the high dominant females in the high dominant female paired with the low dominant male dyads received feedback of better performance than their partners (Fleischer and Chertkoff 1986). All subjects had knowledge of each person's scores. High dominant subjects who received negative feedback were still more likely 42

PAGE 53

than their low dominant partners to become leaders but the likelihood was reduced. Perhaps the most instructive finding was that High Do women in mixed-sex dyads given feedback became the leader with approximately the same frequency as did High Do individuals in same-sex dyads given no feedback. (Nyquist and Spence 1986:93) In the high dominant female and low dominant male dyads, where the high dominant female received feedback which indicated that she was superior at the task, the low dominant males changed their preferences and wanted their partners to be the leader. The feedback presumably legitimated the woman's selection as leader (Fleischer and Chertkoff 1986:98). It may be that the feedback which legitimated the woman as the leader actually deactivated gender as a consideration and performance became activated so that the task would have a better chance for a successful outcome. Subjects deactivated gender and chose to use performance as the basis of interaction. Finally, Fleischer and Chertkoff (1986) examined the subject's initial leadership preferences prior to the formal decision of leadership determination. The subjects who initially were strongest in desiring leadership were 43

PAGE 54

the high dominant men with either low dominant men or low dominant women. The high dominant subjects who were weakest in initial preference for leadership were the women paired with either the low dominant women or the low dominant men. The low dominant subjects who wanted most to be leaders were the low dominant men paired with the high dominant women while the subjects who showed least initial preference for leadership were the low dominant women paired with the high dominant men. While high dominant subjects, except in the case of the high dominant woman paired with the low dominant man, emerged as leaders in substantially greater numbers, the high dominant woman when paired with the low dominant man was weakest in desiring leadership when initial leadership preferences were examined. Prior studies explained the high dominant women's lack of assumption of leadership when paired with low dominant men as the women's adherence to sex-role prescriptions which dictated that males were the most appropriate gender to assume the role of leader. These studies stated that in the high dominant woman/low dominant man dyad, sex-role prescriptions overrode the trait of dominance in the assumption of leadership. Sex-role prescriptions 44

PAGE 55

may not override the trait of dominance if one remembered that in the studies of Megargee (1969), Carbonell (1984) and Nyquist and Spence (1986), many high dominant women displayed their dominance by appointing the low dominant male partner leader. The critical question was whether the high dominant woman displayed dominance or actual leadership? The question of whether there was a difference between dominance and leadership only invited a second question which demanded a clearer definition of leadership. Perhaps the person who appointed another person leader exhibited leadership in a behavioral sense. The definition of leadership may need to be extended further than the simple identification of the individual who is the leader (positional leadership) to perhaps the person who displayed actual leader behavior by appointing another person leader. Dominance and Situational Variables In Aries, Gold, and Weigel's (1983) study, the authors discussed personality research and how the goal of examining a trait such as dominance was done so as to develop an understanding of how differences in human dispositions contributed to the patterning 45

PAGE 56

of individual action. have that the of not but situational (cf. Fiske 1974; Mischel 1968; 1968; 1975). et al. (1983) conducted small discussions to dominance and The subjects the Psychological Dominance Scale and assigned to five and given 40 minutes to discuss an ethical dilemma and produce an answer. The measures of verbal dominance based on the research of Bales (1970) and Zimmerman and West (1975). They included (a) total talking time ... (b) acts initiated .. (c) interruptions (d) interrupter continues .. (e) continues (f) continues after (Aries et al. 1983:782) The situational variable was the sex composition of the group in which seven groups all six groups all and eight groups were mixed-sex groups. It was the second of the two hypotheses that was relevant to this study. It was hypothesized that altering the situation so as to invoke expectations would diminish personality46

PAGE 57

consistency (Aries et al. 1983:781). It was found that the personality-behavior correlation that emerged in the single-sex discussion groups was diminished substantially when the discussion groups comprised both men and women. Interestingly, the situational influence of the sex composition of the group was not sex specific. Both men and women in mixed-sex groups invoked sex-role expectations that inhibited the manifestation of dominance-oriented behaviors consistent with the subject's dispositional tendencies. Aries et al. (1983) took an interactionist perspective and concluded that an understanding of both the situational context and personal dispositions was necessary in order to predict behavior. Proponents of this perspective [interactionist perspective] argue that behavior is a joint function of the personality-determined dispositions of the individual and the characteristics of the situation within which the behavior occurs. In this view, personality eventuates in tendencies to behave in certain ways. The actual behavior that obtains is construed as resulting from both these internal dispositions and the social constraints embedded in the confronting situations (Aries et al. 1983:780-781). 47

PAGE 58

et al. continued discussion of the interactionist perspective and stated that it was meaningless to talk about of individual without specifying the of the situation in which the was exhibited. et al. (1983) study had implications small, mixed-sex discussion indicated that The dominance scale was not significantly with any of the nine with the pattern of dominance-oriented behaviors displayed by either men or women in the mixed groups .. et al. 1983:784). However, Aries et al. (1983) reported that while scores on the CPI Dominance Scale exhibited only limited capacity to predict the frequency of occurrence of single measures of dominance related displayed within the single-sex group discussion ... there was an exhibition of a pronounced relation with the overall pattern of dominance-related behaviors displayed within the situational context. Aries et al. (1983) did not find that the CPI Dominance Scale was of dominance displayed by either men or women in mixed-48

PAGE 59

groups. Their study looked at how consensus was reached while Megargee's (1969) and the replicated studies, along with the present study, examined how leadership was determined before a task was undertaken. There were, as a result, major differences between a mixed-sex group which needed to establish a consensus and a dyad where one participant was high dominant and one participant was low dominant with a requirement to choose a leader before the task began. In the leaderless discussion group, the process by which dominance determined leadership must emerge carefully and more slowly, simultaneous with task performance (i.e. reaching consensus on an ethical dilemma). As a result, if dominance was behaviorally expressed in the mixed-sex group situation, it took an extremely veiled form (Aries et al. 1983:784). The implications of Aries et al.'s (1983) study was that for life outside the laboratory where leaders were not appointed but can, rather, surface "naturally" through the performance of the task, perhaps the situational influence of sex would be attenuated. While this logic conflicts with the Minnesota Studies which found that women rarely emerged as the leader in a coeducational group 49

PAGE 60

containing two or more men (Bormann 1969:219), Bormann's coeducational group participants were not administered the CPI Dominance Scale with high and low dominant individuals strategically situated within the various groups. Megargee, Bogart, and Anderson (1966) introduced another situational variable in their study which was the forerunner of Megargee's (1969) study. The situational variable involved language in the task instructions. The authors used the same nuts-and-bolts machine that was described in Megargee's study; however, in Megargee et al.'s study, the subjects were all males. Each pair had one man high in dominance and one man low in dominance. The authors found that when the instructions emphasized the task, the likelihood of high or low dominant individuals who assumed leadership was left to chance. When leadership was emphasized, however, the High Do Ss assumed the leader role in 901. of the pairs. It is concluded that the CPI scale has predictive validity when leadership is made salient (Megargee et al. 1966:292). In summary, Megargee et als'. study indicated that instructions which emphasized leadership over task was another situational variable which must be 50

PAGE 61

considered when looking at dominance as a predictor of leadership. Dominance and Stability Within Dyads It will be remembered that Gerber (1988) organized the relationships between married couples into a schemata of interdyad matching which matched one person's level of agency and another person's level of communion. This schemata arrangement was related to dyadic solidarity which depicted a balance between the total amount of communion (woman's and man's communion) and the total amount of agency (woman's and man's agency) correlated with the overall amount of satisfaction or solidarity in the relationship as a whole. (Gerber 1988:664) Gerber's (1988) solidarity or satisfaction which resulted in a balance between the total amount of communion (nonpowerful, follower role) and agentic (powerful, leadership role) qualities within a dyad was similar to French and Raven's (1968) findings which stated that the stable dyad was one in which both members accepted the legitimacy of the leader's power through the "right" to influence the follower. The follower was "obligated" to accept that influence. 51

PAGE 62

Nyquist and Spence (1986) had not found that a personality-role mismatch affected the outcome of the task performed, but there was some dissatisfaction with certain dyadic compositions in all dyad types both partners tended to be less satisfied with their interaction and with their role when there was a lack of congruence between their level of dominance and their leader/follower position than when the two factors were congruent. (Nyquist and Spence 1986:93) Smelser (1961) examined dyadic solidarity by measuring group achievement rather than the amqunt of leadership actually displayed by the subjects in the task (p.38). Smelser's subjects (all male) were administered the CPI Dominance Scale and based on the results were assigned to seven different dyads with the following assignments: A. Dominant subject assigned dominant role; submissive subject assigned submissive role B. Dominant and a submissive subject interact with no role assignment C. One dominant subject assigned dominant role; other dominant subject assigned submissive role D. One submissive subject assigned dominant role; other submissive subject assigned submissive role E. Two submissive subjects interact with no role assignment G. Dominant subject assigned submissive role; submissive subject assigned dominant role (Smelser 1961:536). 52

PAGE 63

The study involved the relative achievement of 2-man groups in a cooperative task. Smelser (1961) utilized personality theory which involved the assumption that a person's mode of relating to others has, as its purpose, keeping anxiety to a minimum. Personality theory set forth the assumption that anxiety will disrupt cognitive functioning and lead to a less effective performance in a task (Smelser 1961). Smelser advanced this thesis and stated that different combinations of dominant and/or submissive individuals achieve more or less successfully according to the pair-combinations as well as the conditions of the assignment of dominant or submissive roles. (p.535) The different combinations of dominant and submissive individuals whose achievements were most successful in terms of production were the pairs in which the dominant subjects were assigned the dominant role and the submissive partner the submissive role (Group A). The least productive group was composed of the pairings with the roles reversed (Group G) (Smelser 1961:541). 53

PAGE 64

Predictive Validity of the CPI and the Dominance Scale The California Psychological Inventory (CPI) was developed by Harrison Gough and contains 480 true-false items. The inventory was scaled and profiled for 18 variables. The purpose of each scale was to predict what an individual would do in a specified context, and/or to identify individuals who could be described in a certain way. The theoretical basis for scaling the CPI may be described as folk concepts. A folk concept must transcend a given era or a particular society. The trait dominance for example, was as relevant to Plutarch in his description of Julius Caesar in the first century A.D. as it was to Arthur Schlesinger, Jr., in his proposal of John Kennedy in the twentieth. (Megargee 1972:12) Further, folk concepts contained properties and attributes of interpersonal behavior that were found in all cultures and societies and possessed a direct relationship to all forms of social interaction. The CPI Dominance Scale was developed to identify individuals who manifested dominant, ascendant manners and who would take the initiative and exercise leadership in interpersonal situations. They would also be seen as forceful, self-confident, and capable of influencing others (Gough 1968). 54

PAGE 65

Persons who scored high on the CPI Dominance Scale have been differentially characterized according to gender. A male who scored high on dominance .. will be forceful, resolute, sure of his goals and their philosophical and moral worth, and inexorable in his demands upon self and others The high scoring female is equally strong, but more likely to be coercive. She wears her mantle more aggressively and is more impatient with those who delay or obstruct her progress. (Gough 1968:60) Gough's (1968) high dominant scoring male shared Gerber's (1988) definition of agentic qualities which characterized the powerful person. Agentic traits involved self-assertion, self-expansion, and the urge to master. Gough's (1968) high dominant scoring male also shared Bartol and Butterfield's (1976) definition of the person rated high in initiating structure: identifiable behavior which involved a clear definition of the leader's role, and lets followers know what is expected. Subject's scoring low in dominance were described as: Males: apathetic, indifferent, interests narrow, irresponsible, pessimistic, restless, rigid, reckless, suggestible, submissive Females: cautious, gentle, inhibited, 55

PAGE 66

peaceable, quiet, reserved, shy, submissive, trusting, unassuming. (Gough 1968:60) The low dominant females unmistakenly possessed the nonpowerful communal characteristics that Gerber -(1988) described as accommodating, concern for others, and a desire to be at one with others. The low dominant male's characteristics were also consistent with the description of the male follower who was perceived as having fewer socially desirable characteristics than his female counterpart (Gerber 1988:658). A person that was dominant was disposed to behave in a dominant fashion in a variety of settings (Jones and Davis 1965). In addition, it was the behavior of a person during some span of time rather than single discrete responses or acts that is the subject of interest in discussion of personality Organization is represented by the fact that discrete items of individual behavior appear to fit into a meaningful pattern. (Levy 1970:13) Overall, the CPI Dominance Scale seemed to be a good predictor of leadership when examining same-sex pairs and for men in mixed-sex pairs. Further, the CPI Dominance Scale has predictive validity when leadership was made salient (Megargee et al. 1966). 56

PAGE 67

The predictive validity of the CPI Dominance Scale in absence of sex-role conflict (i.e. high dominant females paired with low dominant males), was confirmed in a number of studies (cf. Bogard 1960; Butt and Fiske 1968; Carbonell 1984; Carson and Parker 1966; Dicken 1963; Fenelon and Megargee 1971; Fleischer and Chertkoff 1986; Gough 1960; Klein and Willerman 1979; Knapp 1960; Megargee 1969; Megargee, Bogart and Anderson 1966; Nyquist and Spence 1986; Rawls and Rawls 1968). Conclusions To the extent that dominance was a factor in predicting the emergence of leadership, Megargee (1969) found in his four dyads that dominance predicted leadership emergence in same-sex and in the mixed-sex dyads of high dominant males paired with low dominant females. Dominance did not predict leadership emergence for high dominant women paired with low dominant men. In this case, only 251. of the high dominant women became leaders which Megargee attributed to sex-role conflict which inhibited the manifestation of dominance. Other studies have replicated and extended Megargee's (1969) study with the dyad of major

PAGE 68

interest continuing to be the high dominant women paired with the low dominant men. This sex-role conflicted dyad found the high dominant woman assuming leadership 301. of the time in Carbonell's (1984) replication, 351. in Nyquist and Spence's (1986) replication, and 501. in Fleischer and Chertkoff's (1986) replication. While the sex-role conflict of the high dominant woman paired with a low dominant man inhibited the woman's emergence as leader so, too, will the sexual orientation of the task. Megargee (1969), Carbonell (1984), Nyquist and Spence (1986), and Fleischer and Chertkoff (1986) all presented a task that they described as sex neutral. However, three of the four studies employed tasks which involved commands or dictated instructions to be carried out by another person. Issuing of orders and carrying out commands have strong social role implications. Further,_ Fleischer and Chertkoff utilized a task that involved spatial relations, numerical skill, and memory. Studies have shown that there was a sex-linked perception of expectations of task success. Elementary school girls had lowered expectations of success than boys on tasks which required the above mentioned skills. 58

PAGE 69

While few if any tasks could be viewed as purely feminine, most tasks favored a more masculine orientation. Megargee (1969) and the replicated studies that followed examined the sex-role conflict depicted in the dyad of the high dominant woman paired with the low dominant man and the possible sexual orientation of the task. If high dominant individuals generally assumed leadership over low dominant individuals, as the California Psychological Inventory Dominance Scale claimed, then the question centered on why high dominant women paired with low dominant men had not assumed leadership as often as the high dominant individuals in other dyad compositions. The surveyed literature above examined sexual stereotyping of women as leaders through evaluations of leadership style. Evaluations of leadership style discussed prevalent perceptions that certain styles were feminine and other styles were masculine. The knowledge that there were perceptions of styles that were feminine and styles that were masculine assisted in the understanding of the social constraints which placed the high dominant woman in particular, and women in general, in a sex-role conflicted position when the woman was 59

PAGE 70

faced with leadership opportunities. The formation of perceptions and cultural expectations that leadership was evaluated in terms of styles traditionally defined as male has led to sexual stereotyping which sabotaged the emergence of women as leaders. There was some disagreement which surrounded the importance placed on personality research which centered on understanding individual action through examination of human dispositions such as dominance. Some have argued that situational variables, not personality traits, were the most important determinants of human behavior. Megargee (1969) and Aries et al. (1983) took an interactionist perspective and concluded that in order to predict behavior, an understanding of both the situational context and personal dispositions was necessary. Aries et al. (1983) in their study which examined whether dominance predicted overt verbal and nonverbal behaviors, found that the situational variables of the sex composition of the group was not sex specific. In other words, both the men and the women in mixed-sex groups invoked sex-role expectations that inhibited the manifestation of dominance. 60

PAGE 71

Aries et al. (1983) had not found that the California Psychological Inventory Dominance Scale was a tool that provided predictability in terms of the overall pattern of dominance-oriented behaviors displayed by either men or women in mixed-sex groups. The lack of predictive validity of the Dominance Scale in this study was a result of another situational variable: a leaderless, consensus building task group (i.e., Aries et al.) versus Megargee's (1969) study where there was a request that a leader be chosen who would be responsible for strategy and outcome of the task. Megargee et al. (1966) introduced the situational variable of language into their study. Megargee et al.'s study was partially replicated by Megargee (1969). In the first study the authors emphasized the task in the instructions and found that dominance had not predicted leadership. In the second study the authors emphasized leadership and found the leader role assumed by 901. of the high dominant individuals in the dyads. Thus, when leadership was made salient, dominance became a viable factor in leadership determination. The level of dominance between the dyad participants was in terms of satisfaction 61

PAGE 72

or solidarity. Smelser (1961) examined solidarity by measuring group achievement. The California Psychological Inventory Dominance Scale was administered to all male subjects who were then placed into seven categories. These categories placed high and low dominant subjects together and either allowed the "natural" personality trait dictate behavior or made assignments of their roles as leader or follower. The combination of dominant and submissive individuals who were most productive were the pairs where the dominant subject was assigned the dominant role and submissive partner assigned the submissive role. These studies indicated that while dominance was a factor in predicting leadership, it was meaningless to discuss characteristics of individual behavior without specifying the situational context in which the behavior was manifested. Further, personality theory claimed that individuals sought interactions which kept anxiety at a minimum; as a result, the level of dominance within a dyad determined the quality of solidarity for the pair. The predictive validity of the Dominance Scale of the CPI has enjoyed confirmation by numerous studies. However, most authors agreed that the 62

PAGE 73

conditions (i.e. situational variables) under which leadership was exercised was as important as the personality trait of dominance in determining whether or not dominant behavior will be manifested (Megargee et al. 1966:295). Attribution and Expectation States Megargee's (1969) study and the replications that followed (Carbonell 1984; Fleischer and Chertkoff 1986; and Nyquist and Spence 1986) all agreed that dominance was a predictor of leadership in same-sex dyads and when the act of leadership rather than the task.was made salient. High dominant individuals had a much greater likelihood of assuming leadership over low dominant individuals in same-sex dyads and mixed-sex dyads with the exception of the high dominant female paired with the low dominant male. These studies supported the idea that sex-role prescriptions overrode dominance as a predictor of leadership in the high dominant female/low dominant male dyad. These sex-role prescriptions represented cultural expectations of the male and female roles. The current study examined how cultural expectations became sex-role stereotyping which 63

PAGE 74

a of women and men into social (Eagly and Steffen 1984). But the study wished to go beyond to explain the in female male assumption of In fact, was a symptom of a significant phenomenon--status inequality. Specifically, the study examined how gender the distribution of influence, and among of such Cohen, and Zelditch 1972). Expectation states and a to analyze how status such as gender determined the distribution of influence and which led to either the assumption or denial of leadership. Simmel in 1908 that .the first condition of having to deal with somebody at all is to know with whom one has to deal. Twenty years Simmel's Soziologie, Park (1928) already assumed a conception of interaction in which one individual classified by age, sex, race, and social type, and behaved him on the basis of these (Berger et al. 1972:241). 64

PAGE 75

When status characteristics such as age, sex and race discriminated some group member from another or were used to make evaluations, then status characteristics were activated. An example would be If p is a woman and believes men know more than women about anything legal, then we say that for p sex is activated in a jury if she believes that specific male jurors know more simply because they are male. (Berger et al. 1972:244) Mixed-Sex Dyads In order to understand status characteristics, we will examine its application to the present study where there were three mixed-sex dyads. These dyads were composed of high dominant females/high dominant males, high dominant females/low dominant males, and middle dominant females/middle dominant males. The subjects were unacquainted and were assigned to dyads controlling for race. Except for gender, one could fairly assume that there were no prior status factors which determined the emergent power-prestige within the dyad (Berger, Cohen, Connor, and Zelditch 1966). Gender was the single status characteristic that had the potential for activation. 65

PAGE 76

When a male and female were asked to choose a leader between themselves, one might assume that if sex were not a status characteristic or had not been activated, then leadership would be fairly evenly distributed between the male and female. It is true that given various situational conditions, status characteristics may not be activated. It is true that activation of D [diffuse status characteristic] is not an inevitable fact. There are certain situations in which D does not in this sense become significant to p, so that it is an important theoretical problem to understand when D does become significant. (Berger et al. 1966:34) When the status characteristic of sex was activated within a mixed-sex dyad, for instance, the result was the occurrence of a diffuse status characteristic. When members of a task group are not the same age, sex or occupation or other institutionalized status factors, it is apparently these factors that determine the power and prestige order that emerges in the group. Such characteristics will be called diffuse status characteristics (Berger et al. 1966:30) There are certain conditions that need to occur before a diffuse status characteristic can be activated. Berger et al. (1966) outlined four conditions for the activation of the diffuse status 66

PAGE 77

characteristic. First, the assumption was that the individuals would be required to perform a valued, collective task (Berger et al. 1966:34). Second, one assumed that the task required some characteristic or ability that could lead to a successful outcome of the task
PAGE 78

The second condition involved the assumption that there were characteristics or abilities that would be positively evaluated for the successful outcome of the task or negatively evaluated for a failed outcome of the task. In expectation states literature, the assumptions concerning positive and negative task outcome mostly involve complex tasks, requiring a variety of different performance abilities [and] analyzing these tasks into their components is a theoretical problem in its own right (Berger et al. 1966:35) This study had not employed a complex task which required a variety of different performance abilities. On the contrary, the subjects had no knowledge of what the task entailed when they chose a leader. This brought us to the third condition, the assumption that the subjects in the mixed-sex dyads had not assigned specific states of characteristics or abilities to each other nor had such an assignment been provided as an initial condition of the situation (Berger et al. 1966). Without a description of the task, subjects could not have assessed one another's abilities to determine who might be best suited for leadership. Whether or not the task was described in detail or simply referred to as "a task" as it was in the 68

PAGE 79

study, subjects still had notions of what it meant to successfully unsuccessfully complete a task. was the assumption that the subjects in the mixed-sex dyads possessed states of a single status so that a basis of discrimination existed between the subjects in to the task. et al. (1966) conceded that most studies dealt with situations in which status simultaneously. This study its focus to a situation in which only one status and that status was Not a task, that the and assigning subjects to dyads was planned in to keep the decision as as possible of competing status and ability When the status of sex became activated, sex became a diffuse status characteristic and behaved as though it symbolized status that had been activated in the situation et al. 1966:30). had available cues: 69 The status

PAGE 80

sex may function as a cue that shapes expectations about other opposite-sex relationships This does not imply that when perceiver make inferences about a person's attributes, they retrieve stereotypes associated with interdependent social roles. Instead, perceivers retrieve their schema for the typical woman and man, and judgments are made in terms of attributes already associated with these roles. (Eagly and Steffen 1984:737) Not only was there a schema for the typical woman and typical man with judgments made on attributes already associated with these roles (Eagly and Steffen 1984) but with judgments there were also beliefs about who would do well and who would do poorly at a task (Berger et al. 1966). Beliefs about task performante is called specific performance expectations, a phenomenon clearly at work when mixed-sex dyads must choose a leader who will plan the strategy and a follower who must obey the leader's commands. The question now became, which sex would be differentially valued when leadership was the issue in mixed-sex dyads? The literature was clear that men have been attributed the qualities of leader and are perceived as having higher status than women (cf. Eagly and Wood 1982; Meeker and Weitzel-O'Neill 1977; Unger 1976). Once beliefs about the relative capacities of individuals to contribute to the group 70

PAGE 81

task were formed, expectation states determined the observable power and prestige order of the group (Hembroff, Martin, and Sell 1981:422). Same-Sex Dyads This study also examined two same-sex dyads: dominant females/low dominant females and high dominant males/low dominant males. Given the circumstances of this study, expectation states literature described these same-sex dyad participants as status equals. As in the mixed-sex dyads, there was no information provided concerning the task and, as a result, there were no opportunities to assess mutual characteristics or abilities. This study examined same-sex dyads in order to replicate Megargee's (1969) study and the studies which followed (Carbonell 1984; Fleischer and Chertkoff 1986; and Nyquist and Spence 1986) to determine whether dominance was a predictor of leadership. Same-sex dyads provided a comparative base against which gender was held constant so that the variable. dominance could be analyzed. In the mixed-sex dyads, diffuse status formed expectation states about leadership/task abilities while in 71

PAGE 82

same-sex dyads was a missing status and, as a result, no expectations available about leadership/task abilities. Conclusions Megargee (1969) and the which followed (Carbonell 1984; and Chertkoff 1986; and Nyquist and Spence 1986) all stressed that the disparity in the assumption of leadership between males and females was due to sex-role sex-role prescriptions were believed to override the dispositional trait of dominance in mixed-sex dyads. This study traveled beyond sex-role explanations and looked at status inequalities outlined in attribution and expectation states theory. According to and expectation states if a actors, [and] if those actors have no other basis of discrimination, and if nothing specifically prohibits the from being used, that will organize subsequent behavior. This phenomenon will occur whether not the characteristic is relevant to the interaction of the actors. (Martin and Sell 1985:178) 72

PAGE 83

Sex was a status characteristic which, when activated in mixed-sex dyads, found that males possessed more prestige and influence than females. Males will be attributed leadership qualities and females will be attributed those qualities of the follower. Sex was the variable which determined leadership assumption. In same-sex dyads, the status characteristic was missing and the dyad participants were viewed as status equals. Controlling for gender, however, provided an examination of the variable dominance for replication purposes. General Conclusions As more women seek leadership positions in our society, it is necessary that they have an informed understanding of the barriers they will encounter. Much of the leadership literature has attributed sex-role stereotyping as a primary barrier. Eakins and Eakins (1978) stated: Sex cannot be so easily dismissed when we consider that it is one of the primary indexing factors we use when we interact with people or think about them. In our society a person's gender matters very much, for to a great extent it determines how others act and react (p.4). 73

PAGE 84

The dispositional trait of dominance, as measured by the California Psychological Inventory Dominance Scale has also been examined in leadership literature as a contributing variable of leadership determination. Previous studies (Megargee 1969; Carbonell 1984; Fleischer and Chertkoff 1986; Nyquist and Spence 1986) have shown that dominance predicted leadership in all dyads except the mixedsex dyad of the high dominant female paired with the low dominant male. This study wished to offer gender role identification as a competing or supplementary consideration with that of dominance. Leadership studies have largely ignored gender role identification as a potential explanation of emergent leadership. The Bem Sex-Role Inventory (BSRI) was designed to measure an individual's adherence to a particular sex-role identification. The BSRI provided a masculine, feminine or androgynous self-definition. By 1977 this inventory had been administered to approximately 1500 undergraduates at Stanford University. Semester after semester, they found that about 50 percent were "appropriately" sex-typed, about 35 percent were androgynous, and about 15 percent were "cross" sex-typed (Bem 1977). One of the questions raised 74

PAGE 85

in this study was it was the level of dominance or gender identification which was the more potent variable in More specifically, this study was the following reasons: (1) to examine how attribution and expectation states could inform traditional studies; (2) to interaction, if any, between the BEM Sex-Role and leadership decisions; (3) to analyze the addition of two new mixedsex, equal dominance dyads (high dominant female/high dominant male and middle dominant female/middle dominant male); (4) to subject's initial desire for leadership with dyad composition and the CPI Dominance Scale scores; (5) to examine dyad composition and CPI Dominance Scale with subjects who themselves as (6) to examine dyad composition and CPI Dominance Scale with subjects who nominated their partners as leader. To a certain extent, sex-role and dominance were appropriate to the discussion of 75

PAGE 86

leadership. This study, however, had not explained emergent leadership as exclusively a result of sex role stereotyping and/or dominance, rather, it focused on how the status characteristic of gender could result in status inequalities. Status inequality was taken to represent differentiation within groups. It was this generalization of information for decision-making [which] is the topic investigated by the expectation-states research (Martin and Sell 1985:178). This study utilized attribution/expectation states as a competing theory to explain emergent leadership in mixed-sex dyads, while the consideration of gender role identification has been offered as a competing or supplementary proposition to the dispositional trait of dominance. 76

PAGE 87

CHAPTER III METHODOLOGY The general question under study asked the extent to which gender and/or dominance was a predictor of leadership emergence. The design required to test the effects of the variables was a 3 X 2 X 2 incomplete factorial design. The factors included were: dominance distribution (high high, high low, middle middle); gender (male-female); and dyad composition (mixed gender-same gender). Of the 12 possible cells, two were repeated which left 10 unique dyads. Four of these remaining 10 dyads were omitted: (1) high dominant women/high dominant women; (2) high dominant men/high dominant men; (3) middle dominant women/middle dominant women; and (4) middle dominant men/middle dominant men. The reason these dyads were omitted was because none of the relevant factors, dominance or gender, varied. The remaining six dyads were relevant for considera-tion and included: (1) high dominant women/high dominant men; (2) high dominant women/low dominant women; (3) high dominant women/low dominant men;

PAGE 88

(4) high dominant men/low dominant men and; (5) middle dominant women/middle dominant men. Five of these treatments were relevant for consideration since they directly addressed the issue under study. The treatment condition of high dominant men paired with low dominant women was omitted for reasons stated below. The high dominant men/low dominant women treatment condition was depicted as the traditional sex-role dyad composition. This treatment was omitted because this study did not find it to be of particular interest or one that would provide additional knowledge to the state of leadership research. Conversely, the dyad of most interest in replicated studies (Carbonell 1984; Fleischer and Chertkoff 1986; Megargee 1969; and Nyquist and Spence 1986) was the high dominant women/low dominant men treatment condition. This dyad was an important focus of this study because it was depicted as the sex-role conflicted dyad which provided information on interaction effects and assisted in the comparison of relative strengths of both gender and dominance. 78

PAGE 89

The high dominant women/high dominant men treatment was an extension of replicated studies (Carbonell 1984; Fleischer and Chertkoff 1986; Megargee 1969; and Nyquist and Spence 1986). This treatment provided a baseline for gender which addressed the main effects of gender and dominance. Further, while this study and replicated studies have focused on high and low dominance as variables for leadership determination, this study further recognized that subjects who scored in the middle range of dominance might have a different perspective on leadership decisions. As a result, middle dominant women paired with middle dominant men were analyzed and this treatment also provided a baseline for gender and examined the main effects of gender and dominance. The same-sex treatments of high dominant women/low dominant women and high dominant men/low dominant men were included because they were baselines for dominance. Finally, fully crossing the two factors in this experimental design has yielded five dyad compositions: (1) high dominant women/low dominant women; (2) high dominant men/low dominant men; (3) high dominant women/high dominant men; (4) high 79

PAGE 90

dominant women/low dominant men and; (5) middle dominant women/middle dominant men. Each of the five treatment conditions consisted of ten dyads. The overall N was 50 dyads or 100 subjects. Recruitment for this study was done in various liberal arts classes at the University of Colorado at Denver. Three hundred and ninety students completed consent forms (see Appendix A) agreeing to participate in the experiment. In an effort to hold race constant, 85 non-white subjects were not contacted to participate in the experiment. Instruments Two hundred and four subjects were contacted by telephone and scheduled to complete a 27-item adaptation of the California Psychological Inventory Dominance Scale (CPI) (see Appendix B). The Dominance Scale consists of 46 items, 27 of which are pure dominance items. The current study utilized only the 27 pure dominant items. (The other 11 items tapped such things as good social intellectual capacity for sense of etc.). The CPI Dominance Scale 80

PAGE 91

framed its questions in a true or false manner. While any one item was not a good indicator of the level of dominance, an individual with a score of 27 possessed a perfect dispositional trait of high dominance. Past studies had not developed a consensus for the establishment of a range for high and low dominance scores. The current study found the mean for women to be 17 and the mean for men to be 17.606. This study established the range of dominance in the following manner: subjects who scored 19-27 were designated high dominant; subjects who scored 1-15 were designated low dominant; and subjects who scored 17 and 18 were designated as possessing middle range dominance. This study, however, was primarily interested in the relativity of high and low dominance. The assignment of an absolute distance between high and low dominance can be extremely misleading. At best, the CPI Dominance scale should be regarded as ordinal in nature. Therefore, it is assumed that the scale scores reflected some relative ranking of increasing levels of dominance. 81

PAGE 92

In addition to the 27-item Psychological Dominance Scale, the Bem Sex Role (BSRI) was completed by the The Bem Sex Role the extent to which a person's self-definition was masculine, feminine or androgynous (See Appendix C). The BSRI is an inventory scored on a ?-point scale, "never or almost true" (1) to "always or almost always true" (7). The androgyny score has a of -6 (extremely masculine) to +6 (extremely feminine). A score of zero indicated a complete androgynous self-definition of gender role identification. Because individuals score in the extremes, the BSRI placed the of the androgyny in the following manner: a feminine sex-role identification has an > +1; a near-feminine sex-role identification has an androgyny of > +.5 and < +1; an androgynous sex-role identification has an androgyny score of > -.5 and < +.5; a near-masculine identification has an androgyny score > -1 and < -.5; and a masculine sex-role identification has an androgyny score< -1 1977). 82

PAGE 93

Procedures Based on dominance scores, subjects were randomly assigned and then paired for the dyad treatment condition consistent with the factor levels in the experimental design. In an attempt to discourage subjects from establishing a connection between the content of the CPI and BSRI test and the intent of the experiment, four to six weeks elapsed before the experimenter contacted the subjects. Four to six weeks was considered a reasonable length of time to ensure that the subjects would have only a vague recollection of the tests. The experimenter requested that each dyad view a videotaped message (see Appendix 0) which stated that the study was interested in how people interacted to complete a task under time stress. They were told that the experimenter was in another room preparing the task. The videotaped message stated that the leader would plan the strategy for the completion of the task and the follower would follow the leader's instructions. The subjects were directed to choose a leader between themselves. The subjects were videotaped as they made their leadership decision. 83

PAGE 94

Once the decision of leadership was made, the experimenter asked each subject to complete a brief questionnaire (see E) before they began the task. The questionnaire probed each subject's satisfaction with the leadership decision, whether the leadership decision-making process was comfortable, and if they believed they would do well on the task. The questions of most interest to this study centered on whether or not there were any preconceived notions concerning the task and if the subjects initially desired leadership, offered to be the leader, or nominated their partner leader. Upon completion of the questionnaire, the subjects were questioned to ensure that they understood the videotaped instructions and to ascertain that they had never been acquainted (see Appendix F). The experimenter then told the subjects that there had been some deception. There would be no task. The subjects were debriefed concerning the study's focus and were provided the scores for their CPI and BSRI tests. Each subject was provided a key to the BSRI and a complete of what the CPI and BSRI test measured. 84

PAGE 95

Hypotheses This design allowed a direct examination of the effects of gender on the determination of a leader; the direct effects of dominance on leader determination; and the interaction of gender and dominance. Specifically, this study tested the following hypotheses: I. The proportion of males that assumed leadership would be greater than the proportion of females who assumed leadership in mixed-sex, equal dominance dyads (high women/high dominant men; middle dominant women/middle dominant men). This hypothesis followed from the general discussion in Chapter II regarding appropriate gender roles. Leadership was assumed to be more appropriate for males than females and therefore, this study anticipated that men would follow traditional expectations and assume the leadership role while women would follow traditional expectations and assume the role of follower. II. The proportion of high dominant individuals that assumed leadership would be greater than the proportion of low dominant individuals in 85

PAGE 96

same-sex dyads (high dominant women/low dominant women and high dominant men/low dominant men). In same-sex treatment conditions, individuals with relatively greater dominance (as determined by the CPI Dominance Scale) became leaders significantly more often than individuals with low dominance scores. As a result, the following hypothesis originated from replications and leadership and dominance literature from Chapter II. I I I When dominance was unequal in a mixed-sex dyad where the female was high in dominance, there will be an equal proportion of leadership distributed between males and females (high dominant women/low dominant men). In the mixed-sex, unequal dominance treatment condition where the female was high dominant and the male was low dominant, the variable gender and dominance competed for the determination of leadership. Because this condition held cultural perceptions of sex-role conflict, it enjoyed the most attention in the replicated studies. The literature in this particular condition found that the best a high dominant female had done in terms of leadership assumption was to become leader SO'l. of 86

PAGE 97

the time. The status characteristic of gender worked against women for determination of leadership and the relatively high dispositional trait of dominance provided the woman an opportunity no better than chance for leadership assumption. While the above three hypotheses tested the effects of gender, dominance, and the interaction of gender and dominance on leader determination, this study also introduced the Bem Sex Role Inventory and the subject's scores were correlated with the CPI Dominance scores. This study tested the hypothesis that: IV. CPI Dominance scores and BSRI scores will be negatively correlated. Specifically, this implied that high dominance scores will be associated with masculinity scores on the BSRI. This hypothesis followed from the notion that dominance was a traditionally male quality while docility was the corresponding female trait. Therefore, persons should evaluate their role in a more masculine perceptual scheme if they possessed high dominance. The CPI Dominance scores were examined in terms of how subjects in particular treatment conditions 87

PAGE 98

approached the leadership decision-making process. In other words, based on the level of dominance in the treatment conditions, this study tested who initially desired leadership, who offered to be the leader, and who nominated themselves leader. study predicted that: This Males will initially desire leadership in greater proportions than females in dyads. Va. In same-sex dyads, high dominant individuals initially desire leadership in proportionally larger numbers than their low dominant partners. This hypothesis followed from the general discussion in Chapter II where the role of leader was traditionally attributed to males. As a result, it would not be surprising that because males were attributed the role of leader, they would also initially desire the role which was of the male gender. Females, on the other hand, would not necessarily initially desire the leadership role because they have been attributed the role of follower and to desire leadership would, to some extent, be out of role for them. 88

PAGE 99

VIa.. Males will offer to be leader proportionally more often than females in dyads. VIc. In same-sex dyads, high dominant individuals will offer to be leader in proportionally larger numbers than their low dominant partners. This hypothesis followed closely hypothesis V because the assumption was that initially desiring leadership and offering to be leader were closely related. Again, this hypothesis drew from attribution/expectation states theory which discussed the cultural expectations that males are viewed as leaders and females are viewed as followers. VI I4 Regardless of dominance levels in mixed-sex dyads, females nominate their male partners leader proportionally more often than males nominate their female partners leader. In same-sex dyads, low dominant individuals nominate the high dominant partner leader proportionally more often than high dominant individuals nominate their low dominant partner leader. 89

PAGE 100

This hypothesis, as in hypotheses V and VI, predicted that relatively high dominance will determine leadership only in same-sex treatment conditions. Regardless of the level of dominance in mixed-sex treatment conditions, women nominate men leaders more often than men nominate women leaders. This is a fulfillment of the culturally defined attribution of male as leader. Except for hypothesis I, these hypotheses followed directly from the preceding discussion of replications in Chapter II. Hypothesis I provided an extension of the replicated studies. (The mixed-sex, equal dominance dyads of high dominant women/high dominant men and middle dominant women/middle dominant men had not been examined in replicated studies). The Bem Sex Role Inventory was also an extension of the replicated studies. A subject's gender role identification provided another variable, in addition to dominance and gender, for the analysis of leadership determination. 90

PAGE 101

CHAPTER IV RESULTS The present study replicated three of Megargee's (1969) Study II dyads which examined the variables of gender and dominance in connection with the assumption of leadership. This replication looked at the number of high dominant subjects who assumed the leader and follower role in the following three treatments: high dominant women/low dominant women (Group I); high dominant men/low dominant men (Group II); and high dominant women/low dominant men (Group III). The present study employed a chi-square test to determine whether or not there was a significant difference between the number of high dominant subjects who assumed the leader and follower role. A chi-square test was chosen because the level of measurement utilized two nominal scales and the samples were random and independent. A correction for continuity was computed because some expected frequencies were five or less. this study was set at .05. The alpha level for

PAGE 102

The present study found no significant difference in the number of high dominant subjects who assumed leader and follower role (X22=4.8; p >.05). Megargee's study, however, found a significant difference between the number of high dominant subjects who assumed leader and follower role (X22=6.56; p <.05, phi-square .14). The present study would have supported Megargee's results for Group I and Group II because the percentage of high dominant subjects who assumed leadership was fairly similar (69'l. in Megargee's study and 85'l. in the present study). However, the present study's Group III realized an equal distribution of leadership between high dominant females and their low dominant male partners. In Megargee's study, only 25'l. of the high dominant female subjects in Groups III assumed leadership over their low dominant male partners. A two sample difference of proportions test was computed to provide a comparison of these two independent, randomly assigned samples for Group III. There was a significant difference between this study's Group III and Megargee's Group III (t24=2.252; p < .01). This comparison revealed that there was a 92

PAGE 103

significant difference with respect to the high dominant women's assumption of leadership over her low dominant male partner (Group III) between Megargee's (1969) study and the present study. Table 4.1. Number of High Dominant Subjects Assuming Leader and Follower Roles Present Group Megargee's Group Study I I I I I I Study I I I I I I Leader 9 8 5 Leader 12 10 Follower 1 2 5 Follower 4 6 N 10 10 10 N 16 16 Group !=high dominant woman/low dominant woman; Group II=high dominant man/low dominant man; Group III=high dominant woman/low dominant man 4 12 16 While the present study was a replication of three of Megargee's (1969) treatments. it was also an extension of Megargee's research because this study examined two new treatments: high dominant women/high dominant men and middle dominant women/middle dominant men. Along with two new treatments. the present study also introduced the Bem Sex Role Inventory (BSRI) as an additional variable for consideration in leadership assumption. As with previous replications (Carbonell 1984; Fleischer and Chertkoff 1986; Nyquist and Spence 93

PAGE 104

1986), the present study used Megargee's (1969) research as a springboard for a somewhat different focus which included a correlation of CPI Dominance scores and Bem Role Inventory scores; individuals who initially desired leadership; individuals who offered to be leader; and those who nominated their partner leader. This focus was then correlated with gender and/or dominance effects. Leader Gender and Gender Hypothesis I predicted that the proportion of that assumed would be greater than the proportion of females who assumed leadership in mixed-sex, equal dominance Specifically, this hypothesi& was tested using two mixed-sex, equal dominance treatments: high dominant females/high dominant males; and middle dominant females/middle dominant males. A single sample test involving proportions was chosen because there was a dichotomized nominal level of measurement. The sample was randomly assigned. The null hypothesis was Pu=.S. Normally a single sample test which involved proportions cannot measure comparisons as in this case where there was a 94

PAGE 105

comparison between males and females. However, designating is a method by which a single sample test can present a test hypothesis comparing two variables. provided the attendant implication that qu=.5. There was a significant difference 2.236; p < .025). The hypothesis was confirmed. Males assumed leadership in greater proportions than females in equal dominance dyads. Leader Selection and Dominance Hypothesis II predicted that the proportion of high dominant individuals that assumed leadership would be greater than the proportion of low dominant individuals in same-sex dyads. This hypothesis was tested for two same-sex unequal dominance treatments: high dominant females/low dominant females and high dominant males/low dominant males. A single sample test involving proportions was performed with a null hypothesis of Pu=.5. Again, single sample tests normally are not engaged in comparisons as in the case of comparing high dominant individuals with low dominant individuals for leadership determination. However, 95

PAGE 106

designating that provided the implication that A significant difference was found 3.1306; p < .005). The hypothesis was confirmed. High dominant individuals assumed leadership in greater proportions than their low dominant partners in same-sex dyads. Leader Gender: Gender Composition of Dyad and Dominance Distribution in the Dyad Hypothesis III predicted that when dominance is unequal in a mixed-sex dyad where the female is high in dominance, there will be an equal proportion of leadership distributed between males and females. This hypothesis specifically tested the high dominant female/low dominant male treatment. A single sample test involving proportions was performed and there was no significant difference (When the hypothesis of interest is in fact the null hypothesis, convention dictates that alpha levels be set at least at .2) The hypothesis was confirmed that the distribution of leadership between high dominant females and low dominant males was equal. 96

PAGE 107

It should be noted that in particular instances the decision for leadership was reached by a coin toss. The coin tosses were distributed in the following treatments: one coin toss in the high dominant woman/high dominant man dyad which resulted in the male winning the toss; two coin tosses in the high woman/low man dyad with the woman winning one toss and the male winning the second toss; one coin toss in the high man/low man dyad with the high man winning the coin toss; two coin tosses in the middle dominant woman/middle dominant man dyad with the woman winning one toss and the male winning the second toss. In the dyads, the idea for a coin toss to determine leadership was always initiated by the male. In Nyquist and Spences (1986) study, eight of the dyads (lOX) elected to toss a or draw lots. Half of the dyads which elected to use a random procedure were the male/male pairs. Nyquist and Spence attributed this to the reconciliation of social prescriptions that men should be leaders with the demand that one man be leader by random (and thus "fair") process (p.89). This study chose to treat the coin tosses as a valid leadership 97

PAGE 108

determination although it was recognized that this was not a negotiated decision between dyad participants. The decision not to omit coin tosses as a leadership decision was based on the fact that four of the six coin tosses were in two mixed-sex dyads where the toss went to both a male and a female. In other words, neither gender benefitted from the toss when the whole treatment condition was examined. Table 4.2. Leader Outcomes Dyad Composition Leader Follower Hi-Do Man/ *1 8 2 Low-Do Man 2 8 Hi-Do Woman/ 9 1 Low-Do Woman 1 9 Hi-Do Woman/ 5 5 Lo-Do Man $2 5 5 Mid-Do Man/ $2 7 3 Mid-Do Woman 3 7 98

PAGE 109

Table 4.2. (contd.) Dyad Composition Leader Follower Hi-Do Man/ *1 8 2 Hi-Do Woman 2 8 N 50 50 Number of coin tosses for leadership determination CPI Dominance Scores and Bem Sex Role Inventory Scores Hypothesis IV predicted that the CPI Dominance scores and the BSRI scores will be negatively correlated. Specifically, this implies that high dominant scores will be associated with masculinity scores on the BSRI. The Spearman rank-order correlation coefficient r4 was utilized in order to measure the association between CPI Dominance scores and BSRI scores. Spearman's r1 was chosen as the most appropriate test because this hypothesis asked for a comparison of rankings on two sets of scores. Spearman's rs also required that the original data were ordinal. Once the two scores were ordered then a rank order 99

PAGE 110

correlation determined the relation between CPI Dominance scores and BSRI scores. There was a moderate relationship between CPI Dominance scores and BSRI scores = -.4034, one tailed test, (N=lOO). The hypothesis was confirmed that the CPI Dominance scores and the BSRI scores were negatively correlated. Specifically, CPI high dominant scores were moderately associated with masculinity scores on the BSRI in the population of which the 100 subjects were sampled. Initial Desires for Leadership and Gender Hypothesis predicted males will initially desire leadership in greater proportions than females in mixed sex dyads. This hypothesis was tested on the following treatments: high dominant women/low dominant men; high dominant woman/high dominant men; and middle dominant women/midd le dominant men. This hypothesis and subsequent hypotheses utilized a two sample difference of proportions test. A two sample difference of proportions test 100

PAGE 111

BSRI Scores 1.!11 A 0.!1 0 I -D. !I p -1.!1 _, lo I ''I I ., I ., I 17 I I l11l :ro I 21 In I n In In l:r I 10 II IJ 14 1!1 17 17 II II ::Ill ::111 21 77 2'2 2J 24 21 BSRI Ranks Figure 4.1. CPI Scores CPI Scores CPI and BSRI Scores and Rankings. Lowest BSRI rank is 1 and progresses in ascending order. 101

PAGE 112

was chosen because a comparison was being made between the dichotomized variable of gender and relating it to a second variable, initial desire for leadership. The two samples were independent because whether a subject initially desired leadership did not depend on the other gender's initial desire for leadership. No significant difference was found (tsg=1.096; p > .05). The hypothesis was not confirmed. There was no difference in the proportion of males and females in terms of initially desiring leadership. Initial Desires for Leadership and Dominance Hypothesis Vb predicted that in same-sex dyads, high dominant individuals will initially desire leadership in proportionally larger numbers than their low dominant partners. This hypothesis specifically tested the following two treatments: high dominant women/low dominant women and high dominant men/low dominant men. A two sample difference of proportions test found that there was a significant difference ( t3g=2.9411 p < .005). The hypothesis was confirmed. High dominant individuals initially desired leadership in 102

PAGE 113

proportionally larger numbers than low dominant individuals. The Offer of Leadership and Gender Hypothesis predicted that males will offer' to be leader proportionally more often than females in mixed-sex dyads. This hypothesis tested the following treatments: high dominant women/low dominant men; high dominant women/high dominant men; and middle dominant women/middle dominant men. A two sample difference of proportions test found that there was no significance difference (t58=1.274; p > .05). The hypothesis was not confirmed. There was no difference in the proportion of males and females who offered to be leader. The Offer of Leadership and Dominance Hypothesis Vlb predicted that in same-sex dyads, high dominant individuals will offer to be leader in proportionally larger numbers than their low dominant partners. This hypothesis specifically tested the following two treatments: high dominant women/low dominant 103

PAGE 114

women and high dominant men/low dominant men. A two sample difference of proportions test found that there was no significant difference (t38:1.253; p > .05). The hypothesis was not confirmed. High dominant individuals in same-sex dyads did not offer to be leaders in proportionally larger numbers than their low dominant partners. Nomination of Partner for Leadership and Gender Hypothesis predicted that regardless of dominance levels in mixedsex dyads, females nominate their male partner leader proportionally more often than males nominate their female partners leader. Three treatments were tested: high dominant women/low dominant men; high dominant women/high dominant men; and middle dominant women/middle dominant men. A two sample difference of proportions test found that there was no significant difference (t2g=.825; p > .05). The hypothesis was not confirmed. Females had not nominated their male partners leader proportionally more often than males nominated their female partners leader. 104

PAGE 115

Nomination of Partner for Leadership and Dominance Hypothesis VIIb predicted that in same-sex dyads, low dominant individuals nominate the high dominant partner leader proportionally more often than the high dominant individuals nominate their low dominant partner leader. This hypothesis tested the following treatments: high dominant women/low dominant women and high dominant men/low dominant men. A two sample difference of proportions test found that there was a significant difference p < .025). The hypothesis was confirmed. Low dominant individuals nominated their high dominant partners leader proportionally more often in same-sex dyads. In summary, the present study could not replicate Megargee's (1969) study results in terms of the number of high dominant subjects who assumed the leader and follower role. The present study found no significant difference in the number of high dominant subjects who assumed the leader and follower role in the three mixed dominance groups (high dominant women/low dominant men; high dominant women/low dominant women; high dominant men/low dominant men). The inability to replicate these 105

PAGE 116

three groups was due primarily to the differences in the number of high dominant subjects who assumed the leader and follower role in the three dominance groups. The inability to replicate these three groups was due primarily to the differences in the high dominant/women/low dominant men treatment condition. Only 25/. of the high dominant women assumed leadership over their low dominant male partners in MegarQees study, while in the present study, there was an equal distribution of leadership within that treatment condition. Hypothesis I, which provided a baseline for gender, was confirmed. Males assumed leadership in greater proportions than females in mixed-sex, equal dominance dyads. Hypothesis II, which provided a baseline for dominance, was confirmed. High dominant individuals assumed leadership in greater proportions than low dominant subjects in same-sex dyads. Hypothesis III, which provided information on interaction effects and assisted in the comparison of relative strengths of both gender and dominance was confirmed. There was an equal distribution of leadership between high dominant women and low dominant men. Hypothesis IV confirmed 106

PAGE 117

that the CPI Dominance scores and the BSRI scores were negatively related. More specifically, negative correlation meant that high dominant scores were associated with the BSRI masculinity scores. The last six hypotheses made predictions based on those who offered to be leader (Table 4.3a.), which subjects initially desired leadership (Table 4.3b.), and who nominated whom for leadership (Table 4.3c.). The independent variables of gender and dominance these three questions. Hypothesis predicted that males would initially desire leadership in greater proportions than females. This hypothesis was not confirmed as there was no difference between males and females in their initial desire for leadership. However, Hypothesis Vo, which predicted that high dominant individuals would initially desire leadership in proportionally larger numbers than low dominant individuals in same-sex treatments was confirmed. Hypothesis predicted that males would offer to be leaders proportionally more often than females. This hypothesis was not confirmed as there were no differences in the proportion of males and females who offered to be the leader. Hypothesis VIo was 107

PAGE 118

Table 4.3a. Dominance of Subjects in Five Dyad Treatments Offering to Be Leader. Dyad Composition Same-Sex Mixed-Sex Mid Mid Hi Hi # Ss M F r.a. M F r.a.. M F M F Offer Leadership Hi-Do 1 4 251. X 0 Of. X X X X Lo-Do 1 0 51. 1 X 101. X X X X Mid-Do X X X X 1 1 X X Hi-Hi X X X X X X 5 0 f.b 101. 201. 101. Of. 101. 101. 50% Of. the percentage of subjects in each dominant category that offered to be leader bRepresents the total percentage of males or females who offered to be leader The N's do not represent dyad totals because of the arbitrariness in the leadership decision exchanges 108

PAGE 119

Table 4.3b. Dominance of Subjects in Five Dyad Treatments Initially Desiring to Be a Leader Dyad Composition Same-Sex Mixed-Sex Mid Mid Hi # Ss M F %a.M F %a. M F M Desire Leadership Hi-Do 6 7 65% X 6 60{. X X X Lo-Do 2 0 10% 3 X 30% X X X Mid-Do X X X X 6 3 X Hi-Hi X X X X X X 9 %1:1 40% 35% 30% 60% 60% 30% 90Y. the percentage of subjects in each dominant category that initially desired to be leader Hi F X X X 3 30% the total percentage of males or females that initially desired to be leader The N's do not represent dyad totals because of the arbitrariness in the leadership decision exchanges 109

PAGE 120

Table 4.3c. Dominance of Subjects in Five Dyad Treatments Who Nominated Their Partner to Be Leader Dyad Composition Same-Sex Mixed-Sex Mid Mid. Hi Hi # Ss M F f.O.,. M F {.Q.. M F M F Nominate Partner Hi-Do 1 1 101. X 3 301. X X Lo-Do 6 6 601. 3 X 30/. X X Mid-Do X X X X 2 5 Hi-Hi X X X X X X f. I:> 35/. 351. 30/. 30/. 20% 50% the percentage of subjects in each dominant category that nominated their partner leader X X X X X X 1 0 101. Of. the total percentage of males or females that nominated their partner leader The N's do not represent dyad totals because of the arbitrariness in the leadership decision exchanges 110

PAGE 121

also not confirmed as high dominant individuals in same-sex dyads did not offer to be leader in proportionally larger numbers than low dominant individuals. Hypothesis predicted that females would nominate their male partners leader proportionally more often than males would nominate their female partners leader. This hypothesis was not confirmed, there was no difference in the proportion of males and females who nominated their partner leader. Hypothesis VIIb confirmed that in same-sex dyads, low dominant individuals nominated the high dominant partner leader proportionally more often than the high dominant individual nominated the low dominant partner leader. The last three hypotheses and were split in an interesting manner. Except for Hypothesis the independent variable dominance was the predictor for those individuals who initially desired leadership and those who nominated their partner as leader. The independent variable of gender did not provide the predictive element for those two Hypotheses. Neither gender nor dominance made a difference in the ability to 111

PAGE 122

predict who would offer to be leader in Hypothesis VIa.,= 112

PAGE 123

CHAPTER V SUMMARY ANDCONCLUSIONS The first question examined whether the present study replicated Megargee's (1969) study with regard to whether there was a difference in the number of high dominant subjects who assumed the leader and follower roles. The three treatment conditions were high dominant men/low dominant men; high dominant women/low dominant women; and high dominant women/low dominant men. Unlike Megargee's study, the present study found no significant difference in the number of high dominant subjects who assumed the leader and follower role. These results were mitigated by the fact that while both studies found that dominance was a predictor of leadership in same-sex dyad treatment conditions, the inability to replicate Megargee's study was due to the results of the high dominant woman/low dominant man dyad in the present study. Only 251. of Megargee's high dominant women assumed leadership while in the present study, 501. of the high dominant women assumed leadership in this particular treatment condition. As a result, the substantial increase in high dominant women who

PAGE 124

assumed when with low dominant men in the study as with study was the cause the inability to the findings of the conditions. discussion on this dyad follows). Same-Sex, Unequal Dominance Conditions Both the present study and (1969) study found quite in the same-sex, unequal dominance treatment conditions. The most finding was that high dominant women paired with low dominant women assumed the leader role in somewhat greater numbers than did high dominant men paired with low dominant men. (For an exception see Fleischer and Chertkoff, 1986.) While the were not a possible explanation for these results may be that although the range of CPI Dominance scores for low dominant men and low dominant women were identical, there may be greater on the of the low dominant woman to assume the leader role as compared to the low dominant man. The basis for this reluctance may be because males in have been attributed the leader role and women in general have 114

PAGE 125

been attributed the follower role. Low dominant males have viewed their gender as the leader and perhaps accepted that attribution and were more motivated than the low dominant woman to assume leadership. The dilemma became twofold for the low dominant woman because her gender has not only been attributed the role of follower but, according to her dominance scores, she also possessed the dispositional trait of low dominance. This combination may have resulted in a more advantageous position for high dominant women to assume leadership in greater numbers than the high dominant male counterpart in same-sex treatment conditions. The present study had the strongest proportional representation of high dominant individuals in same-sex treatment conditions who assumed the leader role (901. high dominant women and 801. high dominant men). In addition to the present study, the trait dominance predicted leadership in same-sex dyads in all replicated studies (cf. Carbonell 1984; Fleischer and Chertkoff 1986; Megargee 1969; Nyquist and Spence 1986). The present study demonstrated that even with the passage of 20 years, the CPI Dominance Scale 115

PAGE 126

continued to be a strong predictor of leadership determination in same-sex treatment conditions. While this study found that dominance was a predictor of leadership in same-sex treatment conditions, this observation was further strengthened by the fact that high dominant individuals displayed a very strong initial desire for leadership as compared to their low dominant partners. In addition, results indicated that they almost never nominated their low dominant partner leader (see table 4.3cJ. Hypothesis VIo predicted that high dominant individuals in same-sex dyad treatments would offer to be leader in proportionally larger numbers than their low dominant partners. This hypothesis was not confirmed in same-sex treatment conditions and as we shall also see in the upcoming section on mixed-sex treatment conditions, neither gender nor dominance had an effect on whether an offer for leadership was made. In summary, high dominant individuals in samesex dyads displayed a strong initial desire for leadership and almost never nominated their low dominant partner leader yet, these same high 116

PAGE 127

dominant individuals did not offer to be leader in proportionally larger numbers than their low dominant partners. There were two possible explanations for this result: (1) when subjects did not know what the task entailed, there was little incentive to volunteer or risk oneself as the leader and (2) some subjects commented that the experimental environment wasn't important enough to assert themselves by offering to be the leader. The fact that high dominant individuals had not offered to be the leader in proportionally larger numbers than their low dominant partners was certainly not a barrier in the final leadership determination. Subjects negotiated a leadership decision within a very few minutes. The subjects also knew that their discussion was being videotaped and perhaps they believed that negotiation through consensual agreement was the most appropriate path to leadership determination. High Dominant Female/Low Pominant Male Condition As discussed above, the inability of the present study to replicate Megargee's (1969) study was due primarily to the results of the high 117

PAGE 128

dominant female/low dominant male treatment condition. This replication was also addressed in Hypothesis III which predicted that in mixed-sex dyad conditions where females were high dominant and males were low dominant, there would be an equal distribution of leadership. There was an equal distribution (50%) of leadership between the high dominant females and low dominant males. The present study replicated Fleischer and Chertkoff1s (1986) study of the treatment condition of high dominant females paired with low dominant males. The present study, however, may have actually found a greater percentage of high dominant females assuming leadership because there were procedural differences between this study and Fleischer and Chertkoff's study. Fleischer and Chertkoff's subjects indicated their leadership preference position privately before being placed in the dyad decision-making laboratory setting. Having knowledge prior to the leader decision-making interaction that a leader was to be chosen and indicating on a questionnaire whether one wanted to be the leader may have accounted for more high dominant women assuming leadership. The present 118

PAGE 129

study asked whether the subjects initially desired leadership but this was done subsequent to the laboratory leadership decision-making interaction. The present study may have captured a more spontaneous and less determined response by the high dominant woman in the experimental setting. Hypothesis examined whether in mixed-sex dyads, males initially desired leadership in greater proportions than females. While no significant differences were found, when the high dominant female/low dominant male treatment condition was analyzed, 60/. of the high dominant women and 30/. of the low dominant men initially desired leadership. Hypothesis predicted that males would offer to be leader proportionally more often than females in mixed-sex dyads. This hypothesis was not confirmed but when the high dominant female paired with the low dominant male condition was examined, no females offered to be the leader while one low dominant male made that offer. These findings indicated that while twice as many high dominant women (60/.) as compared to low dominant men (30'l.) initially desired leadership, no high dominant women actually acted upon that desire by offering to be the leader. 119

PAGE 130

The fact that high dominant females initially desired leadership twice as much as low dominant males was probably not surprising as it was a manifestation of the dominance levels of that particular condition. The initial desire for leadership, however, was known by the subjects alone as they privately completed their questionnaire. Offering to be the leader, on the other hand, was an act that required some level of assertiveness. High dominant females offered to be leader 20/. of the time when paired with low dominant females and yet these same females never offered to be leader when paired with low dominant males. This sex-role conflicted dyad condition underlined more clearly than the other mixed-sex dyad conditions the role constraints placed on women when there was an opportunity for leadership. Regardless of the female's own individual characteristic of higher dominance than males, to some extent they did abide by the socially defined attributes and expectations associated with leadership determination. These socially defined attributes and expectations dictate that males are leaders and females are followers. 120

PAGE 131

Hypothesis predicted that regardless of dominance levels in mixed-sex dyads, females would nominate their male partners leader proportionally more often than males would nominate their female partners leader. This hypothesis was not confirmed and when the high dominant female paired with the low dominant male condition was independently examined, there was an equal distribution of males and females who nominated their partners leader (30% for both sexes). The present study holds out some hope that if dominance is a factor in leadership determination as we know it to be in same-sex dyads, the mixed-sex dyad of high dominant females paired with low dominant males may be the first mixed-sex treatment condition approaching some predictability based on dominance. Dominance may have produced a mitigating, modifying effect in predicting leadership only insofar as the high dominant woman had an equal chance to assume leadership when paired with a low dominant man. 121

PAGE 132

High Dominant Female/High Dominant Male and Middle Dominant Female/Middle Dominant Male Conditions Hypothesis I presented an extension of Megargees (1969) study which examined mixed-sex, equal dominance treatment conditions; specifically, high dominant females paired with high dominant males and middle dominant females paired with middle dominant males. Males assumed leadership in proportionally greater numbers in these mixed-sex, equal dominance conditions. Findings indicated, however, that in hypotheses and where the condition of the high dominant woman paired with the low dominant man was included, there were no gender effects regarding the desire, offer and nomination of the partner for leadership. That particular dyad treatment condition skewed the results for the understanding of the total findings. If one looked independently at the mixed-sex, equal dominance conditions, there were some dramatic results which may have explained how the leadership decision process produced more male leaders. Middle dominant men initially desired leadership proportionally more than their middle dominant women partners (601. versus 301.). In terms 122

PAGE 133

of acting on that desire, only one middle dominant man and one middle dominant woman offered to be leader. Five middle dominant females (507.) nomindted their middle dominant male partner leader and two (20/.) middle dominant males nomindted their middle dominant female partner leader. High dominant men initially desired leadership proportionally more than their high dominant women partners (90/. versus 30/.) and five of the high dominant males and no high dominant females offered to be the leader. Interestingly, no high dominant females nominated their high dominant male partner leader and one high dominant male nominated his high dominant female partner leader. These two dyads were the only conditions that the subjects were of equal dominance. The mixedsex, equal dominance conditions controlled for dominance which posed gender as the only predictive variable in terms of leadership determination. If dominance had predicted leadership in mixed-sex dyads, then one would expect that leadership determination would be evenly split in equal dominance conditions. However, in the present study, the two equal dominance, mixed-sex conditions 123

PAGE 134

found that males became in much than females (701. middle dominant males and 801. high dominant males became Males hold status than females and as a result, all things except gender being equal, males assumed leadership in much by High Dominant Females in Dyadic Conditions The dispositional trait of dominance became an elusive variable when predicting for high dominant women. For example, high dominant women were placed in three different decision-making dyads and they responded quite differently in each of those three situations (high dominant female/low dominant male; high dominant female/low dominant female; and high dominant female/high dominant male). In the present study high dominant women paired with low dominant women were overwhelmingly the leader in 901. of the decision-making occasions. And yet, these same high dominantly scored females dropped to equal leadership assumption when paired with low dominant men and assumed leadership in only 201. of the occasions when paired with high dominant 124

PAGE 135

men. Questions were raised surrounding the phenomenon of the high dominant woman's decline in the assumption of leadership as she moved from same sex conditions to the conditions of being paired with low dominant men and high dominant men. Perhaps high dominant women became less assertive and perhaps less competitive as they move from same sex conditions to the mixed-sex conditions and/or perhaps men become more assertive in their attempt to become leaders when faced with the possibility of being the follower of women leaders? When high dominant women were paired with low dominant women they initially desired leadership 70% of the time and 60% of the time when paired with the low dominant men. However, high dominant women, when paired with high dominant men, initially desired leadership 30% of the time as compared to their high dominant male partners who initially desired leadership 90% of the time. The desire for leadership followed the same pattern of decline as the leadership outcome described above. The high dominant woman's desire for leadership declined as her partners changed from same-sex to mixed-sex. The fluctuation in stated desire for leadership 125

PAGE 136

clearly indicated that desire was a gender based rather than a dominance based phenomenon. Perhaps even more revealing, 40/. of the high dominant women paired with the low dominant women offered to be the leader and yet in the presence of the low dominant man and high dominant man, she never offered to be the leader. The high dominant woman nominated her low dominant female partner 10/. of the time and yet in the presence of her low dominant male partner, she nominated him to be leader 30/. of the time. The high dominant woman's steady progression of deference to men was halted, however, when no high dominant women nominated their high dominant male partner leader. A partial explanation of why high dominant women nominated their low dominant male partners leader 30/. of the time and yet never nominated their high dominant partners leader may lie in the fact that the opportunity for nomination was taken away from the women 50/. of the time because the high dominant men offered to be the leader. In conclusion, high dominant men and high dominant women scored in the same high dominance range of the CPI Dominance Scale and yet these 126

PAGE 137

findings strongly suggested that when high dominant women were in the presence of high dominant men, subtle messages from the high dominant men, which were not clearly evident in the videotaping, must have occurred. Do high dominant men exhibit more assertiveness than high dominant women in their quest for leadership? Do high dominant women receive messages, however subtle, from the high dominant men which projected their strong desire to be the leader? Was the high dominant woman's selfpresentation during this interaction one which displayed a docility not witnessed in the other dyad treatments? These questions were not the scope of this investigation and yet it is an area that could be pursued for future research. Leadership Determination by High Dominant Males in various Dyadic Conditions High dominant men who were cast in two dyad treatment conditions displayed similar inconsistent responses which the high dominant women exhibited in terms of initially desiring leadership, offering to be the leader, and nominating their partner leader. It will be remembered that high dominant men were 127

PAGE 138

placed in two conditions they with low dominant men and high dominant women. (The discussion immediately above the of high dominant females discussed fully the high dominant male' s with the high dominant female in High dominant men initially desired 60/. of the time when with low dominant men but when with high dominant women, they initially desired to be the 90/. of the time. High dominant men offered to be only 10/. of the time when with low dominant men and yet they offered to be the 50% of the time when with high dominant women. High dominant men nominated low dominant male and high dominant female 101. of the time in each condition. The most seemed to be between the high dominant male's desire and offer when with low dominant One might conclude that was a males. The most explanation would be that expectations have dictated that males 128

PAGE 139

leaders and males have internalized that expectation. As a result, males perceived their gender to be leaders and might negotiate leadership with another male more readily than they might negotiate leadership with a female whose cultural expectations would be that of the follower. In conclusion, leadership assumption by high dominant women tended to decline as they moved from the same-sex treatment condition to low dominant male partners and finally to high dominant male partners. Her desire to be leader and her offer to be the leader also declined as she entered the mixed-sex treatment conditions. High dominant males, on the other hand, were unaffected by changes in dyad treatment conditions and enjoyed the same large proportion of leadership assumption in both treatment conditions. Leadership Determination by Low Dominant Males in Various DyadiL Condjtjons Low dominant males had not displayed the rather large fluctuations of responses which were seen with the high dominant males and females. Low dominant men didn't desire leadership much more when 129

PAGE 140

paired with high dominant men than when paired with high dominant women (20/. and 30/. respectively). They also offered to be the leader only one (10/.) time in each treatment condition. In the same-sex treatment condition, low dominant males nominated their partner leader 60/. of the time which was identical to the percentage of low dominant women who nominated their high dominant same-sex partner leader. Low dominant males, however, nominated their high dominant female partners leader 30/. of the time. In other words, low dominant males nominated their high dominant same-sex partners twice as much as they nominated their high dominant female partners leader. These results posed the question of whether the low dominant male was more reluctant to be the follower of a female leader then a follower of a male leader. Dr, as discussed above, the cultural expectations of males as leaders may have been the reason why the low dominant males failed to nominate the high dominant female leader with greater frequency. Further, if cultural expectations perceive women as followers, there was always the possibility that the males believed that women do 130

PAGE 141

not necessarily want to be leaders and, thus, fewer nominations of women for leadership. CPI Dominance Scores and Bem Sex Role Inventory Scores The CPI Dominance scores and Bem Sex Role Inventory scores were negatively correlated as evidenced in Figure 4.1. Although the association was moderate, this negative correlation implied that high CPI Dominance scores are associated with masculinity scores on the Bem Sex Role Inventory. It was determined that in same-sex dyads, dominance predicted leadership and high dominant individuals became leaders in substantially larger proportions than the low dominant same-sex partners. The implication from the CPI Dominance scores and the Bem Sex Role Inventory scores was that high dominant men and high dominant women were more likely to have masculine BSRI scores. This was not surprising when one recognized that the Bem Sex Role Inventory required that subjects describe themselves on a scale from 1 -7 (1 being the lowest) from a large number of personality characteristics. clear that the characteristics conducive to 131 It was

PAGE 142

leadership were those categorized as masculine. Some examples of masculine scored personality characteristics on the Bem Sex Role Inventory which are consistent with leaderships attributes are: self-reliant, willing to take a stand, leadership abilities, willing to take risks, makes decisions easily, dominant, and competitive. The likelihood that there would be a moderate relationship for high dominant individuals (men and women) to possess masculine BSRI scores was equally true in mixed-sex dyads. However, in those instances gender, not dominance, became the variable which determined leadership assumption. In the case of the high dominant female paired with the low dominant male, the implications were that even if the high dominant female had a masculine BSRI score equivalent to or more masculine than the low dominant male, the status characteristic of gender overrode the relationship of masculine BSRI scores. In this situation status was activated rather than the behavioral property of dominance and the status of female was not consistent with leadership. 132

PAGE 143

Suggestions for Future Research Because it was found that CPI Dominance scores and BSRI scores were negatively correlated, future research which would center on the variables of gender and BSRI scores would be a valuable partial replication to contrast gender role identification. and dominance for leadership determination. For example, in same-sex treatment conditions the dyads could consist of BSRI "cross-sex" typed females paired with "appropriate-sex" typed females and "cross-sex" typed males paired with "appropriatesex" typed males. The mixed-sex dyad could be comprised of one treatment condition of BSRI "cross sex typed males paired with "cross-sex" typed females which may be analogous to the high dominant females paired with the low dominant males. Another dyad composition could be "appropriate-sex" typed males paired with "cross-sex" typed females as a possible counterpart for the present study's high dominant females paired with high dominant males. Further, if one wanted to make comparisons with androgynous BSRI scores and its potential relationship to middle dominant males and females,

PAGE 144

androgynous females could be paired with androgynous males. The suggestion that dyads be based on BSRI scores rather than CPI Dominance scores would present the question of whether BSRI scores would predict leadership determination as strongly as high CPI Dominance scores predicted leadership in samesex dyads. Further, would masculine BSRI scores for women in mixed-sex dyads be a more potent indicator of leadership assumption than high CPI Dominance scores for women? One of the limitations of this study was the utilization of college students which in most instances represents a white collar culture. There would be great value in future research which tested whether the results of this study were consistent with other social, ethnic, and racial groups. The present study was also the first replication of Megargee's (1969) work which videotaped all the dyad participant's leadership decision-making discussions. In terms of future research, this videotape could supplement and develop research directed at interaction process analysis of leadership determination. The 134

PAGE 145

videotaping was done primarily as a tool to substantiate the accuracy of the questionnaire results when subjects stated who offered to be the leader and who nominated whom for leadership. The videotape would also provide a partial replication of Nyquist and Spence's (1986) study which audiotaped their dyad's discussion to ascertain the percentage speaking (speakingtime/available-time); rate of sequence suggestions (number-ofsuggestions/unfilled-time); percentage of suggestions made directly (number-ofdirect-suggestions/number-of-suggestions); percentage of summary/strategy statements statements/total-number-of-statements); rate of agreeing (number-ofagreements/number-of-partner-statements); rate of disagreeing (number-ofdisagreements/number-of-partnerstatements); and rate of interrupting (number-of-interruptions/partner's speaking-time) (p.90). The videotaped material may provide subtle manifestations of dominance level behaviors. Conclusions Whether one initially desired leadership, offered to be the leader, or nominated their partner leader did shed some light on how the leadership decision-making process resulted in more males who 135

PAGE 146

assumed leadership. But what was the underlying explanation for the substantially greater proportion of males that assumed leadership in the mixed-sex, equal dominance treatment conditions? The present study employed expectation states and attribution theory in an effort to account for the greater proportions of males that assumed the leadership role. In mixed-sex treatment conditions, gender was the single status characteristic that had potential for activation. When the status characteristic of gender was activated, general assumptions about the individual surfaced. Because females have lower status than males, they brought to the group (dyad leadership determination) characteristics which were of low status outside the group. According to expectation states and attribution theory, the diffuse status characteristic of gender operated as though it symbolized prior status orders which were activated in the situation. As a result, females were considered to have low competence until they proved otherwise (Berger et al. 1972). In this particular experimental setting, there was no opportunity for females to prove otherwise because 136

PAGE 147

not only were the dyad participants unacquainted and thus unable to assess one another's status, but the nature of the task was not revealed and this prevented assessment of mutual abilities. As a result, gender operated as though it symbolized prior status orders which in turn dictated that females possessed lower status. The prior status order resulted in females being followers in greater numbers than males. The utilization of expectation states and attribution theory to analyze the preponderance of males who became leaders in mixed-sex treatment conditions was particularly useful when there was an extremely limited range of cues. In other words, when a situation holds a greater range of cues, the prior status orders will be diminished. In the present study, the dominance levels were manipulated by the researcher in forming dyadic treatment conditions. Gender was the only visible cue upon which leadership was determined. This was the intent of the study. The results of this study strongly suggested that in situations where the only cue to status hierarchy was gender, sex-based 137

PAGE 148

judgments concerning leadership determination may well occur. Some caution should be used in making generalizations concerning the results of the current investigation. The subject pool consisted of a student population whose liberal arts background may express a more liberal view of appropriate sexrole behavior. If this is true, then the rather dismal findings that the present study found in terms of women assuming leadership would be even more discouraging in the real-world work place where there is more diversity in educational background. On the other hand, the present study's experimental setting consisted of groups that had no past and probably no future in terms of interaction. It is a quite different matter when men and women work together and have the same training for their careers and equal preparation for leadership roles. Actual on-the-job experience in mixed-sex groups may mediate the stereotyping effects. And all things being equal (e.g. intelligence, energy level, etc.) men and women should not only perform similarly, but access to leadership positions should not be as 138

PAGE 149

strongly gender-based as the results of this study have shown. Traditional sex-role perceptions are undergoing some changes and the composition of the work force has changed dramatically over the years, however, women are making little headway breaking into the ranks of corporate America. In 60% of the large corporations, fewer than 5% of the senior managers are women and in an August, 1990 Business Week/Harris Poll, 60% of management women in large corporations indicated that "a male dominated corporate culture" is an obstacle to women's success. The value of the present study is that women, whether by deference to males or males asserting themselves to become leaders, do begin the leadership quest at a definite disadvantage. Clearly, emerging as a leader is determined by one's own expectations of gender competence as well as the behavior of the individual concerned. It cannot be denied that the 1990's will bring great pressure from unprecedented numbers of women in the work force to obtain leadership roles and not just support roles. However, the results of this 139

PAGE 150

study did not hold out great hopes for more equality in terms of leadership assumption for women. Megargee (1969) responded to his results of leadership determination in the treatment condition of the high dominant woman paired with the low dominant man by stating that further situations should be undertaken to determine ways in which reluctance of high dominant females to assume leadership can be overcome (p.382). This study would suggest that until the status characteristic of gender, which has placed in the position of status unequals is abolished, men will continue to be the leaders in this society. 140

PAGE 151

APPENDIXES

PAGE 152

APPENDIX A CONSENT FORM A research study will soon begin which will examine the question of how people interact to complete a task. Your involvement in this research will not place you at any risk or danger. You will not be asked to subject yourself to any painful or distasteful experience. I would like to ask you for your assistance in this research. I am seeking volunteers for this study and if you would be interested in helping by participating, please sign the form below. You will be contacted later regarding a date and time for your participation. If, at that time, you decide not to participate you may withdraw your name and no further attempts to get you to participate will follow. If you schedule yourself for a study, you may withdraw from it at any time. Your participation in this study is completely voluntary. You will not receive any fee or other compensation for your participation. All results will be kept confidential. If you have any questions about the purpose of this study in which

PAGE 153

you participate, they will be answered at the completion of your involvement. A summary of study results will be available upon request. If you would like to participate, please complete the form below. Thank you. NAME (please print) ____________________________ SEX ______ __ AGE ______ __ RACE ________ __ PHONE NUMBER ______________ TIME(S) AVAILABLE TO CALL TIME(S) YOU ARE AVAILABLE TO PARTICIPATE IN EXPERIMENT ________________ HAVE YOU EVER PARTICIPATED IN AN EXPERIMENTAL STUDY BEFORE? IF YES, COULD YOU BRIEFLY DESCRIBE IT? SIGNATURE: ______________________________________ ___ Any questions concerning your rights as a subject may be directed to the Office of Research Administration, CU-Denver, Box 123, Denver, Co. 80204, telephone 556-2770. 142

PAGE 154

APPENDIX B CALIFORNIA PSYCHOLOGICAL INVENTORY DOMINANCE SCALE Read each question and decide how you feel about it. If you agree with a statement, or feel that it is true about you circle TRUE. If you disagree with a statement, or feel that it is not true about you, circle FALSE. T F Every citizen should take the time to find out about national affairs, even if it means giving up some personal pleasure. T F I am a better talker than a listener. T F I would be willing to give money myself in order to right a wrong, even though I was not mixed up in it in the first place. T F We should cut down on our use of oil, if necessary, so that there will be plenty left for people fifty or a hundred years from now. T F When a community makes a decision, it is up to a person to help carry it out even if he had been against it. T F I would rather have people dislike me than look down on me. T F I must admit I try to see what others think before I take a stand. T F People should not have to pay taxes for schools if they don't have children.

PAGE 155

T F I would be willing to describe myself as a pretty "strong" personality. T F There are times when I act like a coward. T F I must admit I am a pretty fair talker. T F I have strong political opinions. T F I think I am usually a leader in my group. T F Disobedience to any government is never justified. T F I enjoy planning things, and deciding what each person should do. T F I would rather not have very much responsibility for other people. T F I usually have to stop and think before I act even in trifling matters. T F I have not lived the right kind of life. T F I have a natural talent for influencing people. T F I like to give orders and get things moving. T F I am embarrassed with people I do not know well. T F The one to whom I was most attached and whom I most admired as a child was a woman (mother, sister, aunt, or other woman). T F I'm not the type to be a political leader. T F People seem naturally to turn to me when decisions have to be made. 144

PAGE 156

T F I dislike to have to talk in front of a group of people. T F I have more trouble concentrating than others seem to have. 145

PAGE 157

APPENDIX C BEM SEX ROLE INVENTORY Below you will find a large number of personality characteristics. We would like you to use those characteristics in order to describe yourself. That is, we would like you to indicate, on a scale from 1 to 7, how true of you these various characteristics are. Please do not leave any characteristics unmarked. Example: sly Mark 1 if it is NEVER OR ALMOST NEVER TRUE that you are sly. Mark 2 if it is USUALLY NOT TRUE that you are sly. Mark 3 if it is SOMETIMES BUT INFREQUENTLY TRUE that you are sly. Mark 4 if it is OCCASIONALLY TRUE that you are sly. Mark 5 if it is OFTEN TRUE that you are sly. Mark 6 if it is USUALLY TRUE that you are sly. Mark 7 if it is ALWAYS OR ALMOST ALWAYS TRUE that you are sly. Self-reliant Reliable Warm __ Yielding __ Analytical __ Solemn Helpful __ Sympathetic __ Tender __

PAGE 158

Willing to take a stand __ Cheerful __ Aggressive __ Independent __ Inefficient __ Conscientious __ Childlike __ Affectionate __ Theatrical __ Assertive __ Flatterable __ Competitive __ Strong personality __ Loyal __ Unpredictable __ Forceful __ Feminine __ Defends own Jealous __ beliefs __ Leadership Friendly __ abilities __ Sensitive to Moody __ the needs of others __ Truthful__ Gullible __ Willing to Shy __ __ take risks __ Understanding Acts as a Leader __ Secretive__ Athletic __ Makes Adaptable __ decisions easily __ Compassionate___ Individualistic __ Conventional__ Does not use harsh language ___ Unsystematic__ Self-suffi-cient __ Eager to Happy __ soothe hurt feelings __ Conceited__ Loves children __ Dominant___ Tactful __ Soft-spoken__ Ambitious __ Likable___ Gentle __ Masculine__ Sincere __ 147

PAGE 159

APPENDIX D SCRIPT OF VIDEOTAPED MESSAGE Hello. My name is Jane Hegstrom. I am a member of the research team in Social Sciences at CU-Denver. Thank you for coming today. This study concerns how people interact to complete a task. Of particular interest in this study is how time stress affects the quality of performance in task completion. The task you will perform is one which requires one of you to be the leader. The leader will have the responsibility for making final decisions about strategy and the follower must carry out those decisions. You will determine between yourselves who will be the leader. This discussion will be videotaped. I am in the other room preparing the task. I will return to explain the task shortly. When I return, please report to me who will be the leader. Thank you for your participation. 148

PAGE 160

APPENDIX E SOCIAL SCIENCE RESEARCH LABORATORY OUESTIONNAIRE NAME (please print) ______________________________ __ DATE __________________ __ 1. Were you told during the video-taped instructions that you and your partner would complete a task? _________ N.O ________ YES 2. Do you have any preconceived notions of what the task might entail? _______ N.O ________ YES 3. Were you told by the experimenter who would be the leader? _____ __,NO __________ YES 4. Before the leader decision was made, did you want to be the leader? -------'NO ________ YES

PAGE 161

5. Did you initially offer to be the leader? NO YES 6. Did you initially suggest that your partner be the leader? NO _____ YES 7. Are you satisfied with the leadership decision? _____ _,NO _____ YES 8. Are you the leader? _____ _,NO ______ YES 9. Did you find the leadership decision-making process a comfortable exercise? _____ _,NO ______ YES 10. Do you anticipate doing well on the task with your partner? _____ _,NO _____ YES 150

PAGE 162

APPENDIX F SCOPE CONDITION CHECKLIST NAME ______________________________ DATE ______________ __ SUSPICION: ________ N.O ________ YES IF NOT "NO" EXPLAIN WERE THE FOLLOWING CONDITIONS MET? Understood instructions Are you acquainted with partner Is it a fairly long-standing friendship Aware that a leader must be chosen between themselves Knew that the leader would plan the strategy and be accountable for success of the outcome Knew that the follower would follow leader's instructions WERE ANY OTHER CONDITIONS VIOLATED? YES NO YES NO YES NO YES NO YES NO YES NO IF SO, WHICH? TAKING ALL THINGS INTO CONSIDERATION, SHOULD THIS DYAD BE INCLUDED? ____ _;NO ________ YES 151

PAGE 163

BIBLIOGRAPHY

PAGE 164

BIBLIOGRAPHY Adorno, T.W. 1950. "Democratic Leadership and Mass Mani pu 1 a tion." Pp. 418-35 in StudiE?s in LE?adE?rship: Leadership and DE?mocratic Action, edited by A.W. Gouldner. New York:Russell & Russell, Inc. Aires, E.J., C. Gold, and R.H. Weigel. 1983. "Dispositional and Situational Influence on Dominance Behavior and Small Groups." Journal of Personality and Social Psychology 44:779-86. Bakan, D. 1966. The Duality of Human Existence. Chicago:Rand McNally. Bales, R.F. 1950. Interaction Process Analysis: A Method for the Study of Small Groups. Reading, MA: Addison-Wesley. Bales, R.F. 1970. Personality and IntE?rpersonal Behavior. New York:Holt, Rinehart and Winston. Bales, R.F., and P.E. Slater. 1955. "Role Differentiation in Small Decision-Making Groups." Pp. 259-306 in Family_, Socialization_, and Interaction ProcE?ss, edited by T. Parsons, R.F. Bales, et al. Glencoe, IL.:The Free Press. Bartol, K.M. 1973. "Males and Female Leaders of Small Groups." Bureau of Economic and Business Research. East Lansing:Michigan State University. Bartol, K.M. 1974. "Male Versus Female Leaders: The Effect of Leader Need for Dominance on Follower Satisfaction." Academy of Management Journal 17 (2) :225-33. Bartol, K.M., and D.A. Butterfield. 1976. "Effects in Evaluating Leaders." Journal of Applied Psychology 61:446-54.

PAGE 165

Bell, D. 1950. "Notes on Authoritarian and Democratic Leadership." Pp. 395-408 in Studies in Leadership: Leadership and Democratic edited by A.W. Gouldner. New York:Russell & Russell, Inc. Bem, S.L. 1974. "The Measurement of Psychological Androgyny." Journal of Consulting and Clinical Psychology 42:155-62. Bem, S.L. 1977. "Psychological Androgyny." Pp. 319-24 in Beyond Sex Roles, edited by A. Sargent. St. Paul:West Publishing. Bennis, W., and B. Nanus. 1985. Leaders: The Strategies Far Taking Charge. New York:Harper & Row Publishers. Berger, J. Jr. 1966. States." Progress_, Zelditch, Mifflin. B.P. Cohen, T. Connor, and M. Zelditch, "Status Characteristics and Expectation Pp. 47-73 in Sociological Theories in vol. 1, edited by J. Berger, M. Jr., and B. Anderson. New York:Houghton Berger, J., B.P. Cohen, and M. Zelditch, Jr. 1972. "Status Characteristics and Social Interactions." Sociological Review 37 (3):241-55. Bogard, H.M. 1960. "Union and Management TraineesComparative Study of Personality and Occupational Choice." Journal of Psychology 44:56-63. Bormann, E.G. 1969. Discussion and Group Methods: Theory and Practice. (2nd ed.). New York:Harper and Row. Bowman, G.W., N.B. Worthy, and S.A. Greyser. 1965. "Are Women Executive People?" Harvard Business Review (July-Aug) 43:14-17. Braverman, J.K., S.R. Vogel, D.M. Braverman, F.E. Clarkson, and P.S. Rosenkrantz. 1972. "Sex-Role Stereotypes: A Current Appraisal." Journal of Social Issues 28:50-78. Butt, D.S., and D.W. Fiske. 1968. "Comparison of Strategies in Developing Scales for Dominance." Psychological Bulletin 70:505-19. 153

PAGE 166

Calder, B.J. 1977. "An Attribution Theory of Leadership." Pp. 179-204 in New Directions in Organizational Behavior, edited by B.F. Staw and G.R. Salancik. Chicago:St. Clair Press.Carbonell, J.L. 1984. "Sex Roles and Leadership Revisited." Journal of Plpplied Psychology 69 (1):44-49. Carson, G.L., and C.A. Parker. 1966. "Leadership and Profiles on the MMPI and CPl." Journal of College Student Personnel (January), 14-18. Chemers, M. M. 1984. "The Socia 1, Organ i za tiona 1, and Cultural Context of Effective Leadership." Pp. 91-112 in Leadership: Multidisciplinary Perspective, edited by B. Kellerman. Englewood Cliffs, NJ: Prentice-Hill, Inc. Chemers, M.M., and G.J. Skrzypek. 1972. "Experimental Test of the Contingency Model of Leadership Effectiveness." Journal of Personality and Social Psychology 24:172-77. Crandall, V.C. 1969. "Sex Differences in Expectancy of Intellectual and Academic Reinforcement." Pp. 11-45 in Plchievement-Related Motives in Children, edited by S.P. Smith. New York:Russell Sage Foundation. Dicken, C.F. 1963. "Convergent and Discriminant Validity of the California Psychological Inventory. EdLICational and Psychological Measurement 23:449-59. Doll, P.A. 1965. "A Comparative Study of Top Level Male and Female Executives in Harris County." Unpublished Masters. University of Houston. Eagly, A.H., and V.J. Steffen. 1984. "Gender Stereotypes Stem from the Distribution of Women and Men into Social Roles." Journal of Personality and Social Psychology 46:735-54. Eagly, A.H., and w. Wood. 1982. "Inferred Sex Differences in Status as a Determinant of Gender Stereotypes about Social Influence." Journal of Personality and Social Psychology 43:915-28. Eakins, B.W., and R.G. Eakins. 1978. Sex Differences in Human Communication. Boston:Houghton Mifflin. 154

PAGE 167

Feather, N.T. 1968. "Change in Confidence Following Success or Failure as a Predictor of Subsequent Performance." Journal of Personality and Social Psychology 9:38-46. Feather, N.T. 1969. "Attribution of Responsibility and Valence of Success and Failure in Relation to Initial Confidence and Task Performance." Journal of Personality and Social Psychology 13:129-44. Fenelon, J.R. 1966. "Influence of Race on Leadership Prediction." Unpublished dissertation. University of Texas, Austin. Fenelon, J .R., and E. I. Megargee. 1971. "Influence of Race on the Manifestation of Leadership. u Journal of Applied Psychology 55:353-58. Fiedler, F.E. 1964. "A Contingency Model of Leadership Effectiveness." Pp. 150-90 in Advances in Experimental Social Psychology, val. 1, edited by L. Berkowitz. New York:Academic Press. Fiedler, F.E. 1967. A Theory of Leadership Effectiveness. New York:McGraw-Hill. Fiedler, F.E. 1971a. "Validation and Extension of the Contingency Model of Leadership Effectiveness: A Review of Empirical Findings." Psychological Bulletin 76:128-48. Fiedler, F.E. 1971b. Leadership. New York:General Learning Press. Fiske, D. W. 1974. "The Limits for the Conven tiona 1 Science of Personality." Journal of Personality 42:1-11. Fleischer, R.A., and J .M. Chertkoff. 1986. "Effects of Dominance and Sex on Leader Selection in Dyadic Work Groups." Journal of Personality and Social Psychology 50 (1):94-99. French, J.R.P., Jr., and B. Raven. 1968. "The Basis of Social Power." Pp. 259-69 in Group Dynamics: Research and Theory (3rd ed.), edited by D. Cartwright and A. Zander. New York:Harper & Row. Galton, F. 1879. English Hen of Science: Their Nature and Nurture. New York:Appleton Century. 155

PAGE 168

Gardner, John W. 1965. "The Anti-Leadership Vaccine." Annual Report of the Carnegie Corporation. New York:Carnegie Corporation. Gerber, G.L. 1988. "Leadership Roles and Gender Stereotype Traits." Sex Roles 18 (11/12):649-68. Gibb, C.A. 1954. "Leadership." Pp. 877-920 in Handbook of Social Psychology, vol. 2, edited by G. Linzey. Reading, PA:Addison-Wesley. Gilmer, B. 1961. Industrial Psychology. New York: McGraw-Hill. Gough, H.G. 1960. The California Psychological Inventory Nanual. Palo Alto, CA:Consulting Psychologist Press. Gough, H.G. 1968. "An Interpreter's Syllabus for the CPI." Pp. 55-79 in Advances in Psychological Assessment, val. 1, edited by P. McReynolds. Palo Alto, CA:Science and Behavior Books, Inc. Graen, G., K. Alvares, J.B. Orris, and J.A. Martella. 1970. "A Contingency Model of Leadership Effectiveness: Antecedent and Evidential Results." Psychological Bulletin 74:285-96. Harris Poll. Business Week. August 6, 1990. Heinicke, C., and R.F. Bales. 1953. "Developmental Trends in the Structure of Small Groups." Sociometry 16:7-38. Hembroff, L.A., M. Martin, and J. Sell. 1981. "Total Performance Inconsistency and Status Generalization: An Expectation Status Formulation." The Sociological Quarterly 22:42130. Hemphi 11, J. K. 1949. "The Leader and His Group." Educational Research Bulletin 28:225-229. Hollander, E.P., and J.W. Jullian. 1968. "Leadership." Pp. 890-97 in Handbook of Persona 1 it y Theory and Research' edited by E. F. Borgatta and W.W. Lambert. Chicago:Rand McNally. Hollander, E.P., and J.W. Jullian. 1969. "Contemporary Trends in the Analysis of Leadership Processes." Psychological Bulletin 71:387-97. 156

PAGE 169

Hollander, E.P., and J.W. Jullian. 1970. "Studies in Leader Legitimacy, Influence, and Innovation." Pp. 34-68 in Advances in Experimental Social Psychology, vol.5, edited by L. Berkowitz. New York:Academic Press. Hunt J.G. 1967. "Fiedler's Leadership Contingency Model: An Empirical Test in Three Organizations." Organizational Behavior and Human Performance 2:290-308. Hunt, J.G. 1984. "Organizational Leadership: The Contingency Paradigm and Its Challenges." Pp. 113-38 in Leadership: f>fultidisciplinary Perspectives, edited by B. Kellerman. Englewood Cliffs, NJ:Prentice-Hill, Inc. Jacobson, M.B., and J. Effertz. 1974. "Sex Roles and Leadership: Perception of the Leader and the Led." Organizational Behavior and Human Performance 12:383-96. Jones, E.E., and K.E. Davis. 1965. "From Acts to Dispositions: The Attribution Process in Person Perception." Pp. 220-66 in Advances in Experimental Social Psychology_, vol. 2, edited by L. Berkowitz. New York:Academic Press. Kahn, R. and D. Katz. 1960. "Leadership Practices in Relation to Productivity and Morale." Pp. 612-28 in Group Dynamics, edited by D. Cartwright and A. Zander. Evanston, IL:Row, Peterson, and Co. Kirk, R.E. 1982. Experimental Design: Procedures for the Behavioral Sciences (2nd ed.). Belmont, CA:Brooks Cole Publishing Co. Klein, H.M., and L. Willerman. 1979. "Psychological Masculinity and Femininity and Typical and Maximal Do Expression in Women." Journal of Personality and Social Psychology 37 (11):2059-70. Knapp, R.R. 1960. "A Reevaluation of the Validity of MMPI Scale of Dominance and Social Responsibility." Educational .:md Psychological Measurement 20:381-86. 157

PAGE 170

Kutner, B. 1950. "Elements and Problems of Democratic Leadership." Pp. 459-67 in Studies in Leadership: Leadership and Democratic edited by A.W. Gouldner. New York:Russell & Russe 11 Inc. Levy, L.H. 1970. Conceptions of Personality: Theories and.Research. New York:Random House. Lewin, K. 1948. "The Consequences of an Authoritarian and Democratic Leadership." Pp. 40917 in Studies in Leadership: Leadership and Democratic Action, edited by A.W. Gouldner. New York:Russell & Russell, Inc. Lippitt, R., N. Polansky, F. Redl, and S. Rosen. 1952. "The Dynamics of Power." Human Relations 5:37-64. Lombardo, M.M., and M.W. McCall, Jr. 1978. Leadership: Where Else Can We Go? Durham, NC: Duke University Press. Mann, R.D. 1959. "A Review of the Relationships Between Personality and Performance in Small Groups." Psychological Bulletin 56:241-70. Martin, C.R., Jr. 1972. "Support for Women's Lib: Management Performance." The Southern Journal of Business 7:19-28. Martin, M.W., and J. Sell. 1985. "The Effect of Equating Status Characteristics on the Generalization Process." Social Psychology Quarterly 48:178-82. McKee, J.P., and A.C. Sherriffs. 1957. "The Differential Evaluation of Males and Females." Journal of Personality 25:356-71. Meeker, B.F., and P.A. Weitzel-O'Neil!. 1977. "Sex Roles and Interpersonal Behavior in Task Oriented Groups." American Sociological Review 42:91-105. Megargee, E.I. 1969. "Influence of Sex Roles on the Manifestation of Leadership." Journal of Psychology 53, (5) 377-82. Megargee, E.I. 1972. The California Psychological Inventory Handbook. San Francisco:Jossey-Bass, Inc. Publishing. 158

PAGE 171

Megargee, E.I., P. Bogart, and B.J. Anderson. 1966. "Predic:tion of Leadership in a Simulated Industrial Task." Journal of Psychology 50 (4) 292-95. Milner, J. 1965. Studies in Hanagement Education. New York:Springer. Misc:hel, W. 1968. Personality and New York:Wiley. Nyquist, L.V., and J.T. Spenc:e. 1986. "Effec:ts of Dispositional Dominanc:e and Sex Role Expectations on Leadership Behaviors." Journal of Personality and Social Psychology 50 (1):87-93. Offerman, L.R. 1986. "Visibility and Evaluation of Female and Male Leaders." Sex Roles 14:553-43. Park, R.E. 1928. "Bases of Rac:e Prejudic:e." The 140:11-20. Peterson, D.R. 1968. The Clinical Study of Social Behavior. New York:Appleton-Century-Crofts. Pfeffer, J. 1978. "The Ambiguity of Leadership." Pp. 13-35 in Leadership: Where Else Can We Go?, edited by M.W. Mc:Call, Jr. and M.M. Lombardo. Durham, NC:Duke University Press. Pondy, L.R. 1978. "Leadership is a Language Game." Pp. 87-102 in Leadership: Where Else Can We Go?, edited by M.W. Mc:Call, Jr. and M.M. Lombardo. Durham, NC:Duke University Press. Powell, G.N., and D.A. Butterfield. 1979. "The 'Good Manager': Masc:uline or Androgenous?" of Hanagement Journal 22:395-403. Rawls, D. J., and J.R. Rawls. 1968. "Personality Charac:teristic:s and Personal History Data of Suc:c:essful and Less Suc:c:essful Exec:utives." Psychological Reports 23 (3, Part 2), 1032-34. Rosen, B., and T.H. Jerdee. 1973. "Influenc:e of Sex Role Stereotypes on Evaluations of Male and Female Supervisory Behavior." Journal of Psychology 57 (1):44-48. 159

PAGE 172

Rosenkrantz, P., S. Vogel, H. Bee, I. Braverman, and D.M. Braverman. 1968. "Sex-Role Stereotypes and Self-Concepts in College Students." JoL1rnal of Consulting and Clinical Psychology 9:287-95. Sargent, A. 1977. Beyond St. Paul:West Publishing. Schein, V.E. 1973. "The Relation Between Sex-Role Stereotypes and Requisite Management Characteristics." Journal of Applied Psychology 57:95-100. Schein, V.E. 1975. "Relationships Between Sex-Role Stereotypes and Requisite Management Characteristics Among Female Managers." Journal of Applied Psychology 60:340-44. Schweder, R.A. 1975. "How Relevant Is an Individual Difference Theory of Personality?" Journal of Personality 43:455-84. Shartle, C. 1952. Executive and Columbus, OH:Ohio State University Research Foundation. Sheridan, J.E., J.L. Kerr, and M.A. Abelson. 1982. "Leadership Activation Theory: An Opponent Process Model of Subordinate Responses to Leadership Behavior." Pp. 122-41 in Beyond Views edited by J.G. Hunt, U. Sekaran, and C.A. Schrieshein. Carbondale, IL.:Southern Illinois University Press. Simmel, G. 1950. Sociology of Simmel, edited by K.H. Wolff. Glencoe, IL.:The Free Press. Smelser, W.T. 1961. "Dominance as a Factor in Achievement and Perceptions in Cooperative Problem So 1 v ing Interactions." JoLirnal of Abnormal and Social Psychology 62 (3):535-42. Spence, J.T., and and 1 a tes and Texas Press. R. Helmreich. 1978. Nasculinity Their Psychological Austin: University of 160