• TABLE OF CONTENTS
HIDE
 Title Page
 Dedication
 Acknowledgement
 Table of Contents
 List of Tables
 Abstract
 Introduction
 Review of related literature
 Methodology
 Results and discussion
 Conclusions, implications, summary...
 Appendices
 Reference notes
 References
 Biographical sketch














Title: Individual differences in counselors' causal attributions for performance outcomes
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00099496/00001
 Material Information
Title: Individual differences in counselors' causal attributions for performance outcomes sex, sex role identities and levels of self-esteem
Physical Description: xii, 121 leaves : ; 28 cm.
Language: English
Creator: Hersh, Mindy S., 1955-
Publication Date: 1983
Copyright Date: 1983
 Subjects
Subject: Counseling   ( lcsh )
Attribution (Social psychology)   ( lcsh )
Sex role   ( lcsh )
Self-esteem   ( lcsh )
Counselor Education thesis Ph. D
Dissertations, Academic -- Counselor Education -- UF
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
 Notes
Thesis: Thesis (Ph. D.)--University of Florida, 1983.
Bibliography: Bibliography: leaves 112-119.
General Note: Typescript.
General Note: Vita.
Statement of Responsibility: by Mindy S. Hersh.
 Record Information
Bibliographic ID: UF00099496
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000365935
oclc - 09952423
notis - ACA4763

Downloads

This item has the following downloads:

PDF ( 4 MBs ) ( PDF )


Table of Contents
    Title Page
        Page i
        Page ii
    Dedication
        Page iii
    Acknowledgement
        Page iv
        Page v
    Table of Contents
        Page vi
        Page vii
        Page viii
    List of Tables
        Page ix
        Page x
    Abstract
        Page xi
        Page xii
    Introduction
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
    Review of related literature
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
    Methodology
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
    Results and discussion
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
    Conclusions, implications, summary and recommendations
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
    Appendices
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
    Reference notes
        Page 112
    References
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
    Biographical sketch
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
Full Text










INDIVIDUAL DIFFERENCES IN COUNSELORS' CAUSAL ATTRIBUTIONS FOR
PERFORMANCE OUTCOMES: SEX, SEX ROLE IDENTITIES AND LEVELS
OF SELF-ESTEEM














By

MINDY S. HERSH














A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE
REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA




























Copyright 1983

by

Mindy S. Hersh




























This dissertation would not
have been possible if
not for my parents

ARTIE AND HENY HERSH

to whom I dedicate this work














ACKNOWLEDGEMENTS


There were many people who gave freely of their time and love to

help make this project possible. Dr. Roderick McDavis, my chairperson

and friend who has advised and supported me through many significant

events of my life. I respected your judgments, appreciated your

patience, and valued your standards of excellence. Thank you for the

years of guidance and friendship. Dr. Paul Fitzgerald and Dr. H.

Joseph Reitz my committee members whose encouragement, suggestions, and

advice were very helpful throughout different stages of the study.

Gilda Josephson whose unconditional love and never ending belief

in my potential have seen me through many a rough spot. Thank you my

friend. Judith McBride for knowing so well what it is like to write a

dissertation. Sharing this experience with you made the difference.

Marcia, Sandi, Karan, Rochelle, Betsy, Herb, and Helen who at different

times and in different ways gave me the love, encouragement, and moral

support .I needed to complete this project. Joe McCann who never

doubted I could complete this study or anything else I put my mind to.

Thank you for your advice, your support, and for caring.

Skip Everitt for your faith in me and for allowing me the

flexibility I needed to complete this work.

Connie Carrol my friend and typist. I appreciated your expert

skills almost as much as your caring, humor, and warmth. Behrokh

Ahmadi who expertly and patiently took me through the data analyses.








I appreciated your enthusiasm and valued your commitment to complete

the project on time.

Finally, I would like to thank my mother for loving and accepting

me as I am, and my father for his love and pride in my accomplishments.

You never doubted I would do it -- afterall, I'm your daughter!















TABLE OF CONTENTS


PAGE

ACKNOWLEDGMENTS................................................

LIST OF TABLES.............................................................

ABSTRACT........................................................

CHAPTER

ONE INTRODUCTION......................................... 1

Statement of the Problem ... .......................... 1
Purpose of the Study................................. 5
Need for the Study................................... 5
Significance of the Study............................ 7
Definition of Terms.................................. 8
Organization of the Study............................ 9

TWO REVIEW OF RELATED LITERATURE......................... 11

Attribution Theory .................................. 11
Theoretical Background............................. 11
Theoretical Models................................. 14
Studies Revealing Sex Differences
in Causal Attributions .......................... 17
Studies Revealing No Sex Differences
in Causal Attributions........................... 21
Future Directions for Attribution Research.......... 23
Sex Roles.......................................... 25
Sex Role Identity ................................. 25
Sex Role Identity and
Causal Attribution Studies....................... 28
Self-Esteem....................................... .. 31
Theoretical Background.............................. 31
Self-Systems..................................... 32
Multidimensional Theories of Self-Esteem........... 34
Traditional and Nontraditional Career Women.......... 36
Summary .................. ........................... 39

THREE METHODOLOGY.......................................... 40

Null Hypotheses..................................... 40
Population and Sample................................ 41
Instruments....................................... .. 45








The Bem Sex Role Inventory.......................... 45
The Performance Self-Esteem Scale.................. 47
The Causal Attribution Scale Forms A and B......... 50
The Demographic Information Questionnaire.......... 52
Procedures........................................ .. 52
Analyses of Data .................................... 53
Limitations of the Study ............................. 54

FOUR RESULTS AND DISCUSSION .... ............................. 55

Results............................................. 55
Hypothesis One..................................... 55
Hypothesis Two..................................... 57
Hypothesis Three................................... 59
Hypothesis Four.................................... 62
Hypothesis Five.................................... 63
Hypothesis Six..................................... 66
Additional Results .................................. 68
Discussion........................................ .. 73

FIVE CONCLUSIONS, IMPLICATIONS,
SUMMARY AND RECOMMENDATIONS.......................... 79

Conclusions....................................... .. 79
Implications...................................... .. 80
Summary............................................. 82
Recommendations for Future Research.................. 83

APPENDICES

A BEM SEX ROLE INVENTORY............................... 84

B PERFORMANCE SELF-ESTEEM SCALE........................ 87

C FINAL CAUSAL ATTRIBUTION SCALE FORMS A AND B......... 90

D LETTER TO ATTRIBUTION RESEARCHERS.................... 94

E ORIGINAL CAUSAL ATTRIBUTION SCALE FORMS A AND B...... 95

F EVALUATION FORMS FOR CAUSAL
ATTRIBUTION SCALE FORMS A AND B...................... 99

G ATTRIBUTION RESEARCHERS' EVALUATIONS
OF ORIGINAL CAUSAL ATTRIBUTION SCALE
FORMS A AND B....................................... 103

H DEMOGRAPHIC INFORMATION QUESTIONNAIRE................ 106

I LETTER TO PARTICIPANTS............................... 107

J SUMMARY OF TABLES. .................................. 108








REFERENCE NOTES................................................. 112

REFERENCES....................................................... 113

BIOGRAPHICAL SKETCH............................................. 120














LIST OF TABLES


TABLE PAGE

1 Causes of Success and Failure Classified According
to Locus of Causality, Degree of Stability,
and Controllability....................................... 7

2 AMHCA Membership by Primary Employment Setting............ 42

3 AMHCA Membership by Academic Credentials.................. 43

4 AMHCA Membership by Primary Job Responsibility............ 43

5 AMHCA Membership by Race ................................. 44

6 One-Way Analyses of Variance of Causal
Attributions for Counselors' Successful
Performance Outcomes by Sex............................... 56

7 One-Way Analyses of Variance of Typical
Effort Attribution for Counselors' Successful
Performance Outcomes by Sex Role Identities............... 57

8 Least Significant Differences Test
of Counselors' Sex Role Identities
for Typical Effort Attribution............................ 58

9 Means and Standard Deviations for Typical Effort
Attribution by Counselors' Sex Role Identities............ 59

10 One-Way Analyses of Variance of Ability and
Typical Effort Attributions for Counselors'
Successful Performance Outcomes by Levels of
Agentic Self-Esteem...................................... 60

11 Least Significant Differences Tests of
Counselors' Levels of Agentic Self-Esteem
for Ability and Typical Effort Attributions............... 61

12 Means and Standard Deviations for Ability
and Typical Effort Attributions by Counselors'
Levels of Agentic Self-Esteem............................. 63

13 One-Way Analyses of Variance of Causal
Attributions for Counselors' Unsucessful
Performance Outcomes by Sex............................... 64

ix








14 One-Way Analyses of Variance of Causal
Attributions for Counselors' Unsuccessful
Performance Outcomes by Sex Role Identities............... 65

15 One-Way Analysis of Variance of Immediate
Effort Attribution for Counselors' Unsuccessful
Performance Outcomes by Levels of Agentic Self-Esteem..... 66

16 Least Significant Differences Test
of Counselors' Levels of Agentic Self-
Esteem for Immediate Effort Attribution................... 67

17 Means and Standard Deviations for
Immediate Effort Attribution by
Counselors' Levels of Agentic Self-Esteem................. 68

18 Classification of Counselors' Successful
and Unsuccessful Performance Outcomes
by Achievement Domains.................................. 69

19 Distribution of Responses to Causal
Attributions for Counselors' Successful
Performance Outcomes by Achievement Domains............... 70

20 Distribution of Responses to Causal
Attributions for Counselors Unsuccessful
Performance Outcomes by Achievement Domains............... 72

21 One-Way Analyses of Variance of Causal
Attributions for Successful Performance
Outcomes Not Significant by Sex Role Identities............ 109

22 One-Way Analyses of Variance of Causal
Attributions for Successful Performance
Outcomes Not Significant by Levels of
Agentic Self-Esteem.................................... 110

23 One-Way Analyses of Variance of Causal
Attributions for Unsuccessful Performance
Outcomes Not Significant By Levels of
Agentic Self-Esteem.................................... 111














Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



INDIVIDUAL DIFFERENCES IN COUNSELORS' CAUSAL ATTRIBUTIONS FOR
PERFORMANCE OUTCOMES: SEX, SEX ROLE IDENTITIES AND LEVELS
OF SELF-ESTEEM


By

MIND S. HERSH

April, 1983


Chairman: Roderick J. McDavis
Major Department: Counselor Education


The purpose of this study was to investigate the effects of

counselors' sex, sex role identities, and levels of agentic self-esteem

on their causal attributions for successful and unsuccessful

performance outcomes. Agentic self-esteem refers to that dimension of

self-esteem concerned with ability and performance. The causal

attributions investigated were ability, typical effort, interest in

task, immediate effort, mood, supervisor, luck, task ease, and help

from others.

The instruments were mailed to 600 members of the American Mental

Health Counselors Association. The Bem Sex Role Inventory was used to

assess counselors' sex role identities, the Performance Self-Esteem

Scale was used to assess counselors' levels of agentic self-esteem, and

the Causal Attribution Scale was used to assess counselors' causal

attributions for self-reported performance outcomes. The final sample








included 62 male counselors and 65 female counselors. One-way analyses

of variance were used to analyze the data.

The results of the data analyses indicate that differences did not

exist in counselors' causal attributions for successful outcomes when

compared by sex. The results also show that differences did exist in

counselors' causal attributions for successful outcomes when compared

by their sex role identities. Undifferentiated sex-typed counselors

were less likely than the androgynous, masculine, and feminine

sex-typed counselors to attribute their successful outcomes to their

typical effort.

The results further indicate that differences did exist in

counselors' causal attributions for successful performance outcomes

when compared by their levels of agentic self-esteem. Lower level

agentic self-esteem counselors were less likely than the higher level

agentic self-esteem counselors to attribute their successful outcomes

to their typical effort and ability. The results also show that

differences did not exist in counselors' causal attributions for

unsuccessful performance outcomes when compared by their sex.

The results of the data analyses reveal that differences did not

exist in counselors' causal attributions for unsuccessful outcomes when

compared by their sex role identities. Finally, the results indicate

that differences did exist in counselors' causal attributions for

unsuccessful outcomes when compared by their levels of agentic

self-esteem. Higher level agentic self-esteem counselors were less

likely than the lower level agentic self-esteem counselors to attribute

their unsuccessful outcomes to their immediate effort.














CHAPTER ONE

INTRODUCTION


Statement of the Problem

Differences in the achievement behaviors of men and women have

been well documented. The attributional model of achievement behavior

has been used extensively to explain these differences in terms of

corresponding differences in causal attributions assigned to

performance outcomes (Teglasi, 1977). The attributional model assumes

that following success or failure at an achievement task, outcome

causality is attributed to some combination of the following factors:

ability, effort, luck, and task difficulty. These four factors can be

conceptualized along two dimensions: locus of causality which

classifies the factor as internal or external to the individual, and

degree of stability whicn refers to the factor's stability over time

(Weiner, Frieze, Kukla, Reed, Rest, & Rosenbaum, 1971).

In spite of the substantial body of literature produced by

attribution research, a consensus concerning sex differences in

achievement attributions has not been reached. The problem as commonly

recognized by researchers is the inconsistency in findings across

investigations (Sohn, 1982). Some studies have found that women tend to

be more external than males in their attributions for both success and

failure outcomes (Feather, 1969; Simon & Feather 1973). Other studies

report that females but not males have self-derogatory attributional








patterns where they accept responsibility for failure outcomes but

attribute success to external and unstable sources (Nicholls, 1975).

Still other studies have found no sex differences at all in attribution

patterns (Sohn, 1982).

Two explanations for the inconsistencies in attribution literature

have been offered. The first explanation addresses the problem of

viewing men and women as homogeneous groups. Research separating

subjects into nonhomogeneous groupings is needed to investigate the

impact of individual differences among men and women on causal

attributions (McHugh, Frieze, & Hanusa, 1982). The attribution

literature suggests examining differences in sex role identity and

self-esteem may contribute to determining the effect of individual

differences on patterns of causality (Brewer & Blum, 1978; Levine,

Gillman, & Reis, 1982).

Sex role identities are determined by the extent to which persons

conceptualize their own degree of masculinity and femininity. These

masculine and feminine definitions exert real societal pressures upon

men and women to behave in specifically prescribed ways. The need to

keep behavior consistent with these internalized sex role definitions is

identified as a powerful motivational force (Broverman, Vogel,

Broverman, Clarkson, & Rosenkrantz, 1972; Kagan, 1964; Kohlberg, 1966).

The traditional feminine sex role embraces many personality

characteristics that generally conflict with achievement such as

nonassertiveness, avoidance of competition, and dependency. The

traditional masculine sex role includes characteristics most often

associated with achievement such as aggression, independence and a

willingness to take risks (Bem, 1974; Stein & Bailey, 1973).






A-

Changes in social, cultural, and economic norms are resulting in

the acceptance of new interpretations of appropriate behaviors for men

and women. The value of the traditional sex role dichotomy is being

questioned and the concept of androgyny, the blending of both masculine

and feminine characteristics, seriously examined. Consequently,

increasing numbers of men and women are becoming less sex-typed in their

sex role identifications. Where wide gaps once existed in behaviors

exhibited by the sexes, women are now reported as possessing many of the

same characteristics as men including assertiveness and Machiavellianism

(Wertheim, Widom, & Wortzel, 1978). In view of these changes, the

validity of many global statements concerning men and women must be

questioned.

The literature also suggests level of self-esteem may affect causal

attributions (Fitch, 1970; Levine et al., 1982). Self-esteem has

traditionally been conceptualized as a unidimensional factor constant

across all aspects of the self (Stake, 1981). More recently, the view

of self-esteem as unifactorial is being challenged by researchers that

argue self-esteem cannot be defined by a single measure (Stake, 1981).

Consequently, theories examining self-esteem as a multidimensional

construct are achieving acceptance.

Stake (1979) has developed a self-esteem scale that measures the

"agentic" dimension of self-esteem. The term agentic, derived from

Bakan's (1966) model of orientations toward life, refers to that aspect

of self-esteem concerned with ability and performance. Researchers

argue that a measure of agentic self-esteem should predict achievement

behavior more accurately than a global measure (Franks & Marolla, 1976).

However, research investigating the relationship between self-esteem and








attributions for achievement outcomes relies solely on global

self-esteem measures (Fitch, 1970; Levine et al., 1982). In order to

achieve a more precise assessment of the effect of self-esteem on causal

attributions, this study considered individual differences in the

agentic dimension of self-esteem.

Failure to consider the effects of situational determinants on

attributions may also account for the inconsistencies found in

attribution research. Situational variables such as the nature of the

task itself have not been adequately investigated. Relationships

between qualitative characteristics of performance tasks and the

subsequent causal attributions assigned to task outcomes should be

examined. Travis, Burnett-Doering, and Reid (1982) suggest

investigating this relationship by classifying recalled success and

failure experiences according to the type of experiences reported, then

examining the attributional factors assigned causality for these

performance outcomes.

Other situational variables, such as the context in which the task

is performed, must also be considered. The vast majority of attribution

studies examine student populations' causal attributions for the

outcomes of controlled laboratory tasks. Few studies have been

conducted investigating attribution patterns in real life achievement

settings such as the workplace. Those studies that have been conducted

in the field concentrate on occupations that are nontraditional for

women such as the management profession (Deaux, 1979).

Just as it is necessary to extend attribution research further

into real life settings, it is important not to limit these

investigations to nontraditional occupations. The traditional-








nontraditional dichotomy may be significant as women choosing

traditional career paths have been shown to differ from women choosing

nontraditional career paths on a variety of dimensions (Karman, 1972;

Tangri, 1972). Traditional occupations are defined as those where over

50 percent of the workers are women (Tangri, 1972). Occupations

falling into the traditional category include teaching, nursing,

secretarial work, and counseling (U.S. Bureau of the Census, 1981).

This investigation focused on men and women in the counseling

profession.


Purpose of the Study

The purpose of this study was to investigate the effects of

counselors' sex, sex role identities, and levels of agentic self-esteem

on their causal attributions for successful and unsuccessful performance

outcomes.

Need for the Study

A commonality throughout attribution literature is the belief that

systematic differences in men's and women's causal attributions may

explain why women are not equaling men in the work force, politics, and

other fields traditionally associated with achievement (Frieze, Parsons,

Johnson, Ruble, & Zellman, 1978b). Major changes in society's social

and economic structures are making the identification of factors

contributing to women's low achievement status increasingly important.

Women are entering and re-entering the labor force in

unprecedented numbers accounting for 42.8 percent of all workers.

However, whereas 20 years ago women sought employment largely for

self-fulfillment purposes, today's women are seeking employment out of

economic necessity (National Commission on Working Women, 1981).








Forty-three percent of employed women are single, widowed, divorced, or

separated. Twenty-three percent are married to men earning incomes of

less than $17,000 a year. Although economic need is the primary

motivating factor for obtaining employment, the majority of working

women continue to be concentrated in low paying and often dead-end

jobs. Average full time women workers earn 59.64 to every dollar

earned by male full time workers (National Commission on Working Women,

1981).

If attribution research is to shed light upon the problem of

female underachievement it must broaden its scope. Recent research

indicates that the assignment of outcome causality is not limited to

the four factors identified in the Weiner et al. (1971) landmark

publication. However for the most part, attribution studies continue to

assume that ability, effort, luck, and task difficulty are the only

factors used to assess performance outcomes. Research investigating a

wider range of causal factors is needed (Weiner, 1979).

This study employed the three dimensional classification system

developed by Weiner (19/9) and investigated nine causal factors.

Weiner's (1979) three level taxonomy classifies causal factors according

to locus of causality and degree of stability, the original attribution

model dimensions, and according to controllability. Controllability

refers to whether or not the factor is subject to volitional control.

Table 1 presents the nine factors investigated in this study classified

according to these three dimensions.








Table 1

Causes of Success and Failure Classified According to
Locus of Causality, Degree of Stability, and Controllability



INTERNAL EXTERNAL

STABLE UNSTABLE STABLE UNSTABLE


UNCONTROLLABLE Mood at Task
Ability Time of Task Difficulty Luck

Interest
In Task

CONTROLLABLE Typical Immediate Help
Effort Effort Supervisor From Others



Significance of the Study

Results of attribution research continue to be inconclusive

indicating the necessity for moving beyond the research paradigm

generally used to investigate causal attributions. This study expanded

upon the traditional research model by employing a new methodology and

by considering variables not adequately addressed in the past. The data

obtained from this study could support researchers who believe new

directions in attribution research are needed if an understanding of sex

differences in achievement behavior is to be reached.

Results from this study could also provide information on the

relationship of sex role identity and self-esteem to causal

attributions. Emerging patterns in these relationships may have

important implications for mental health professionals. Underestimating

ability can result in underachievement (Pheterson, Kiesler, &

Golderberg, 1971). If counselors are aware of differences in the

attribution patterns of persons with differing sex role identities and








varying levels of self-esteem, they may be better prepared to

facilitate growth in clients with underachievement problems.


Definition of Terms

For the purpose of this study, the terms below will be defined as

follows:

Achievement Domain -- Classification of reported success or failure

experiences according to area or type of achievement. The three domains

employed in this study were Intrapersonal Events,

Interpersonal-Affiliative Events, and Mastery-Control Events.

Agentic Self-Esteem -- Based upon Bakan's (1966) model of agency and

communion orientations toward life, agentic self-esteem refers to that

dimension of self-esteem concerned with ability and performance.

Agentic self-esteem is concerned with the individual's achievement and

is reflected in such behaviors as assertiveness and independence.

Androgynous -- Sex role identity embracing both masculine and feminine

characteristics. Behaviors are exhibited dependent on situational

appropriateness rather than in a fashion predetermined on the basis of

sex.

Causal Attributions -- The factors to which persons assign

responsibility for the successful or unsuccessful outcomes of

achievement tasks. In this study nine causal attributions were

investigated: ability, typical effort, interest in task, immediate

effort, mood at time of task, supervisor, luck, task ease/difficulty,

and help from others.

Controllability -- Dimension that classifies causal attribution factors

according to whether or not they are subject to volitional control.








Counselors -- Persons in public or private employment settings holding a

master's degree or above, who report counseling as their primary job

responsibility.

Degree of Stability -- Dimension that classifies causal attribution

factors as fixed or variable over time.

External Attributions -- Causal explanations for success or failure at a

task reflecting the belief that outcomes are primarily due to outside

influences over which individuals have no control.

Feminine -- Sex role identity that embraces many personality attributes

that conflict with achievement such as nonassertiveness, an avoidance of

competition, and an unwillingness to take risks.

Fixed Attributions -- Causal explanations for success or failure at a

task that are stable over time.

Internal Attributions -- Causal explanations for success or failure at a

task reflecting the belief that characteristics of the self are

primarily responsible for performance outcomes.

Locus of Causality -- Dimension that classifies causal attribution

factors as internal or external to the individual.

Masculine -- Sex role identity that includes those personality

characteristics associated with achievement such as independence,

aggression, competitiveness, and a willingness to take risks.

Undifferentiated Sex-Type -- Sex role identity classification that

results from a low endorsement of both the masculine and feminine

characteristics on the Bem Sex Role Inventory.

Variable Attributions -- Causal explanations for success or failure at a

task that are subject to change over time.





-10-


Organization of the Study

The remainder of this study will be organized into four chapters.

Related literature on attribution theory, sex roles, self-esteem, and

comparing traditional career women to nontraditional career women will

be reviewed in Chapter Two. The methodology used in this study will be

described in Chapter Three. The results of the study and a discussion

of these results will be presented in Chapter Four. The conclusions,

implications, summary, and recommendations for future research will be

presented in Chapter Five.














CHAPTER TWO

REVIEW OF RELATED LITERATURE


The review of literature related to this investigation will be

divided into four sections: attribution theory, sex roles,

self-esteem, and traditional and nontraditional career women. The

section on attribution theory will include theoretical background,

theoretical models, studies revealing sex differences in causal

attributions, studies revealing no sex differences in causal

attributions, and future directions for attribution research. The

section on sex roles will include sex role identity, and sex role

identity and causal attribution studies. The section on self-esteem

will include theoretical background, self-systems, and multidimensional

theories of self-esteem. The last section of the review will include a

discussion of traditional and nontraditional career women.


Attribution Theory

Theoretical Background

The attributional model of achievement behavior was developed by

Heider (1958) and subsequently expanded by Weiner et al. (1971). This

model has been used extensively in efforts to explain sex differences

in achievement by identifying corresponding differences in causal

attributions assigned to performance outcomes (Teglasi, 1977).

Weiner et al. (1971) originally hypothesized individuals utilize

four attribution factors to both predict and explain the outcome of an





-12-


achievement related event. These four factors, ability, effort, luck,

and task difficulty, are employed for both self-evaluations and for

assessing the performances of others. Future expectations of success

and failure are also based upon the assumed level of ability in relation

to perceived task difficulty as well as on an estimate of intended

effort and anticipated luck.

The four causal factors can be classified into two dimensions:

locus of causality and degree of stability. The first dimension is

derived from Rotter's (1966) internal-external control of construct and

refers to whether a person perceives success as contingent upon internal

or external causal sources. Ability and effort describe qualities of

the person undertaking the activity and 'are therefore internal

components. Task difficulty and luck are environmental factors and are

external to the person. The degree of stability dimension indicates

whether the factor is fixed or variable over time. Ability and task

difficulty are somewhat enduring characteristics whereas effort and luck

are relatively variable.

More recently, Weiner (1979) has stated his concern that the four

factors identified in his original research were not meant to be

considered the only determinants of success and failure. Although

ability, effort, luck, and task difficulty consistently emerge among

the main perceived causes of achievement performance, research

employing open-ended response methodologies reveals these four are not

the only factors assigned responsibility for performance outcomes.

Bar-Tal and Darom (Note 1) investigated attributions made in

academic settings and found seven causes emerged as those used to

explain success or failure including ability, immediate effort, and








interest in the subject matter. Cooper and Burger (Note 2) also

investigated attributions for success or failure in a classroom setting

and found that several factors in addition to Weiner's et al. (1971)

original four were identified including typical effort, mood, and

help/bias of others.

Weiner (1979) developed a new classification system to categorize

the list of conceivable causes for success and failure. This taxonomy

expands on the original classification dimensions of locus of causality

and degree of stability to include controllability as the third

dimension. This dimension classifies factors according to whether or

not they are subject to volitional control.

Employing these three dimensions, causal factors can theoretically

be classified within one of eight cells. Among the most frequently

identified internal causes of success and failure, ability is stable and

uncontrollable, typical effort is stable and controllable, mood at time

of task and interest in task unstable and uncontrollable, and immediate

effort unstable and controllable. Among the external causes most often

employed, task difficulty is stable and uncontrollable, luck is

unstable and uncontrollable, supervisor may be perceived as stable and

controllable, and help from others as unstable and controllable.

Weiner (1979) concedes that problems exist with this

classification scheme. One unanswered question is whether or not

controllability assumes only the perspective of the actor. The factors

typical effort, immediate effort, and help from others are controllable

from the vantage point of the actor. However, supervisor cannot

necessarily be controlled except by the supervisors themselves.








Theoretical Models

A number of investigators have reported sex differences in their

attribution research. Researchers have proposed several models to

explain the occurrence of these sex differences. These models vary in

terms of their predictions and the degree of empirical support each has

received. They are similar in that the models are based upon research

limited to investigations of the original Weiner et al. (1971) two

dimensional taxonomy of causes. The three models most cited are General

Externality, Self Derogation, and Low Expectancy (Frieze, Whitley,

Hanusa, & McHugh, 1982).

One of the first models developed to explain sex differences in

attributions, General Externality, suggested that women tend to

attribute both their success and failures to external sources, i.e.,

luck and task difficulty (Feather, 1969; Simon & Feather, 1973).

Several explanations for this externality have been presented. Feather

(1969), whose work supported this theory, suggested that the greater

tendency to attribute success and failure to external sources rather

than to ability among women may be a result of how women perceive the

feminine role. Women have traditionally been taught to appear modest

and dependent. Assertiveness and self-confidence, reflected in

self-acknowledgements of ability, are masculine traits and therefore

inappropriate behaviors for women.

A second proposed explanation for externality among women is that

women are higher in both fear of success and fear of failure and

therefore withdraw from achievement situations. Horner (1968)

originally identified fear of success as the tendency for women to

internalize early in life the notion that femininity and individual








achievement reflecting intellectual competence are mutually exclusive,

antagonistic goals (Bremer & Wittig, 1980). Using task ease and luck to

explain success allows women to maintain a distance from achievement and

a lack of involvement in future tasks. Using these external

attributions also reduces personal responsibility for success and

decreases feelings of shame for failure (Frieze, Fisher, Hanusa, McHugh,

& Valle, 1978a).

A third explanation for externality provides a sociological

perspective. Wiley, Crittenden and Berg (1979) suggest that because

women and other low status groups tend to have less control over their

destinies than those of higher status groups, they tend to attribute

performance outcomes to external factors.

The Self-Derogation model predicts that women attribute successes

to external factors but attribute failures to internal sources (Heilman

& Kram, 1978; Ickes & Layden, 1978; Nicholls, 1975). This second model

assumes that people attempt to maintain a consistent set of beliefs

about themselves and process information according to these beliefs

(Aronson & Mettee, 1968). If persons have high self-esteem they will

only believe positive information about themselves and therefore

internalize success but not failure outcomes. If persons have low

self-esteem they will only believe negative information about themselves

and therefore internalize failures but not successes (Fitch, 1970).

Women have been reported as typically having low self-esteem in

achievement settings (Frieze et al., 1978a). Consequently, in order to

maintain consistency in self-beliefs, the self-derogation model predicts

that women will accept negative information about themselves and

discount positive information about themselves.








Women have been shown to have generally low expectations of their

performance in achievement situations (Deaux, 1976). According to the

Low Expectancy model, low expectations lead to unstable attributions,

i.e., luck and effort for success, and stable attributions, i.e., lack

of ability and task difficulty for failure. Even in cases of success,

the outcome is discounted and therefore does not lead to an increase in

expectancies for future performance. In cases of failure, attributions

to stable sources, i.e., poor ability, serve to further lower future

expectations (Deaux, 1976; Frieze et al., 1978a).

Low expectancy has been shown to be greatest for women performing

unfamiliar tasks. In these cases generalized expectations or

stereotypic assumptions are relied upon to replace knowledge gained

through previous experience. This has been found to be especially true

for women performing tasks defined as "masculine." As these tasks are

perceived as inappropriate for females, women have low expectations for

their own success. Even if they should subsequently succeed at these

tasks, the outcomes are attributed to unstable causes (McHugh, Fisher, &

Frieze, 1982; Deaux, 1976).

In spite of the substantial energy devoted to achievement

attribution research, as illustrated by the theoretical models

discussed, a consensus concerning sex differences has not been reached

(Sohn, 1982). The reason for this lack of agreement is the

inconsistency among research findings across investigations. Studies

have failed to replicate the findings of similar earlier ones (Feather,

1969 and Feather & Simon, 1973) or when replicated produce contradictory

results (Nicholls, 1975 and Simon & Feather, 1973).

Many published studies reveal sex differences on at least one of

the four original causal factors. However several recent studies have





-17-


concluded that main effect sex differences do not exist at all (Frieze

et al., 1982; Sohn, 1982; Sweeny, Moreland, Gruber, 1982; Travis, 1982;

Travis et al., 1982). The following section will review the empirical

findings of research investigating sex differences in causal

attributions. The investigations reviewed will be divided into two

sections: studies revealing sex differences in attribution patterns and

studies revealing no sex differences in attribution patterns.


Studies Revealing Sex Differences in Causal Attributions

One of the earliest studies to uncover sex differences in causal

attributions was conducted by Feather (1969). In this study,

undergraduate students performed laboratory tasks then indicated whether

they believed the outcomes of their performances were due to their lack

of ability. Measures were also taken of students perceived

self-competence, self-esteem, and feelings of inadequacy prior to the

task, and of students' satisfaction with their performances after the

task. Results of this study revealed female students were more likely

to attribute their success or failure to external factors, i.e., good or

bad luck, than were males. Females were also found to have higher

inadequacy scores and lower initial confidence scores. No other

significant sex differences were reported.

In a later study, Simon and Feather (1973) asked undergraduate

students to identify those factors that influence their initial

expectations of success on a college examination to determine under what

conditions these factors would be perceived as causes of the outcome.

The analysis of data revealed a significant main effect due to sex for

ratings of initial ability. Males rated their preperformance ability

higher than females. Females who failed the examination were especially








likely to rate their preperformance ability as low. Females also

attached more importance to luck and task difficulty as factors causing

their outcomes than males.

The tendency for males to credit their success to ability more so

than women has been replicated in several studies (Deaux, 1979; Deaux &

Farris, 1977) as has the greater tendency for women to use luck to

explain their performance outcomes (Bar-Tal & Frieze, 1977; Deaux &

Farris, 1977). Deaux (1979) investigated differences in male and female

managers' performance self-evaluations and attribution of causality

patterns and found that males saw themselves as performing significantly

better overall than did females.

When comparing themselves on a number of job related

characteristics to other persons holding similar positions, male

managers evaluated their ability greater and their intelligence higher

than female managers. Males also viewed their jobs as more difficult,

reported having a better relationship with their supervisors, and felt

they received relatively more approval for their work than did the

female managers. Assessments of causality patterns revealed males

claimed ability as more responsible for their success than did females.

Luck, effort, and task difficulty were not differentially used as

explanations for success or failure.

Women often take less credit for their successes than men and in

some instances have been reported to blame themselves more for failure

(Levine, Reis, Sue, & Turner, 1976; McHugh, Fisher & Frieze, 1982;

Nicholls, 1975; Rosenfield & Stephan, 1978; Stephan, Rosenfield, &

Stephan, 1976). The nature of the task in determining attributions for

success and failure has been mentioned by a number of researchers as








especially relevant to these findings (Deaux & Farris, 1977; Luginbuhl,

Crowe & Kahan, 1975; McHugh, Frieze, & Hanusa, 1982; Nicholls, 1975;

Zuckerman, 1979). For example, sex differences in causal attribution

have been found in studies where subjects were asked to perform tasks

defined as traditionally masculine.

In a review of causal attribution studies, Zuckerman (1979)

concluded that women tend to take less responsibility for success and

accept more responsibility for failure for masculine tasks. Rosenfield

and Stephan (1978) found women took less credit than men for success on

a masculine task but took more credit than men for success on a feminine

task. Deaux and Farris (1977) found in two experiments that males

evaluated their ability as higher particularly when the task was labeled

masculine. Females were more likely to use luck to explain performance

regardless of how the task was defined.

Whether a task is competitive or cooperative has also been shown to

affect causal attributions (Ames, 1978; House, 1974; Stephan et al.,

1976; Teglasi, 1977; Teglasi, 1978). Teglasi (1977) found females gave

more external and more unstable attributions for success in a

competitive condition than for success obtained cooperatively.

Cooperating females did not differ from cooperating males in their

causal attributions. Ames' (1978) study suggested that females take

less responsibility for success and more responsibility for failure in a

competitive context while males have the reverse pattern.

Stephan et al. (1976) asked female and male subjects to attribute

their performance outcomes in a competitive game to one of the four

original causal factors. The authors found males and females competing

against females took more credit for success than they gave successful








opponents and blamed themselves less for failure than they blamed

failing opponents. Females competing against females did not show the

same attributional patterns. Stephan et al. (1976) suggested that

females competing against males were less defensive because they

considered the task masculine and their opponents as males were expected

to be more skilled than they.

The Stephan, Rosenfield, and Stephan (1976) study was replicated

by Heilman and Kram (1978). Heilman and Kram predicted that when

paired with a female as opposed to a male coworker, both men and women

would take more responsibility for success and less responsibility for

failure. They further predicted that men with female coworkers and

women with male coworkers would respectively be the least and most

self-derogating in their attributions. The results of this study

partially supported the authors' hypotheses. When paired with men on

laboratory tasks, women's tendencies to derogate themselves were

evident. This did not hold true however when women worked with other

women. Men also had higher self-ratings when paired with women

indicating that situational factors as well as individual

predisposition impact on attributions for performance outcomes.

Several studies have shown that perceptions of outcomes explain

individual differences in persons having high and low achievement

motivation (Bar-Tal & Frieze, 1977; Kukla, 1972; Teglasi, 1978; Wiegers

& Frieze, 1977). Bar-Tal and Frieze (1977) explored the attributional

patterns of high and low achievement motivated women and compared these

patterns with those of men. The authors found that the only main effect

for sex to reach the significance level was the tendency of women to

employ higher luck ratings.








In contrast, many of the interaction effects involving sex and

achievement motivation were found to be significant, replicating

earlier studies and/or confirming the Bar-Tal and Frieze (1977)

hypotheses. Males with high achievement motivation had very high

estimates of their abilities. Males also tended to attribute success

primarily to internal sources and to see failure as the result of

external sources. High achievement motivated women had strong beliefs

in effort as a causal factor for both success and failure. They also

tended to be somewhat more external for success than the high

achievement males.


Studies Revealing No Sex Differences in Causal Attributions

Travis (1982) found that sex differences provide very little

information about the perception of causality. After investigating

subjective evaluations of success and the degree of correspondence

between actual and subjective success measures for women and men, no

differences were found. Travis' work further revealed similarities on

expressions of expectations for future performances and a nearly

identical relationship of specific attributions to expectations for

men and women.

House (1980) found significant differences in attributions of

causality made by subjects observed by others and by subjects not

observed by others. Observed subjects evidenced less tendency to

attribute their failure to low ability. However, there were no

significant main or interaction effects for the sex variable. Houser

and Beckman (1978) also failed to find significant sex differences in

the casual attributions of male and female college students as did

Feather and Simon (1971) in an early investigation of attribution

patterns.








Frieze, Whitley, Hanusa, and McHugh (1982) conducted an overall

assessment of the three theoretical models for sex differences in

causal attributions reported earlier in this chapter. The authors

sought to determine which if any of the three theories was best

supported by the literature. A meta-analysis of 21 published studies

addressing sex differences in ratings for causal attributions was

conducted. Results of the meta-analysis revealed that women were found

to have a very slight tendency to attribute failure to luck more than

men, and men made somewhat stronger attributions to ability for

situational success. However, overall there were no strongly supported

sex differences in attributions and not one of the three theoretical

models was supported.

Sohn (1982) conducted an effect-size analysis of previously

published studies to determine the consequences of the relationships

found in attribution research. The author investigated two categories

of findings: the overall or main effect for sex and the simple sex

difference effects for the separate levels of achievement outcome. The

investigation was an attempt to validate two popular but partially

contradictory hypotheses: females are more external than males in

their attributions (Feather, 1969) and females are less egotistic than

males by being more external for success and internal for failure

(Nicholls, 1975). Sohn concluded that the empirical evidence does not

support the contention that the sexes significantly differ in their use

of any of the four main types of achievement attributions or that the

sexes significantly differ in their use of internal and external

attributions.

The McHugh, Fisher, and Frieze (1982) study analyzed the

simultaneous effects of two situational factors on the attributions








made by females and males. These factors were whether the task was an

intellectual and therefore masculine task or a social and therefore

feminine task, and whether the task was performed in a situation of high

or low competitiveness. The study failed to reveal sex differences in

evaluations of self-performance. There were no task x competition x sex

of subject interaction effects therefore disconfirming the authors'

predictions.


Future Directions for Attribution Research

McHugh, Frieze, and Hanusa (1982) state that the literature on sex

differences in attributions is characterized by inconsistencies and has

not fulfilled its promise as the key to understanding differential

achievement in men and women. Two explanations are offered for these

inconsistencies: The failure to adequately consider various

situational determinants of sex differences in attributions and the

problems of viewing women as an homogeneous group (Frieze et al.,

1978a).

Situational variables may be viewed as falling into two

categories: the task itself and the context in which the task is

performed. What may outwardly appear to be the same task for male and

female participants may actually be subjectively two very different

tasks in terms of cultural or individual beliefs about the sex

appropriateness of the task, gender, or individual differences in

attainment value or ego involvement, and the novelty/familiarity of the

task. Similarly, the context in which the task is performed, while

objectively identical for males and females, may have different

subjective implications.








McHugh, Frieze, and Hanusa (1982) suggest that in order to discern

attributional tendencies it may be necessary to measure causal

attributions across a variety of situations so situational influences

affecting one situation do not lead to generalizing results to all

situations. For example, Travis et al. (1982) asked respondents to

recall an accomplishment and failure they actually experienced and

attribute causality for these outcomes. This approach controls for

contextual influences that may distort attributions although possible

biases against important situational factors must also be considered.

McHugh, Frieze, and Hanusa (1982) also suggest that to understand

particular groups of people it makes sense to assess their attributions

in their most typical or desirable environments. By asking respondents

to define their achievement domains, researchers are also able to

classify people into various achievement domain preference groups. The

advantages of this approach are that recalled experiences may provide

information about the most salient achievement domains of men and women.

The recalled experiences are also more likely to have been relevant and

important to the respondent, and to be interpreted as a success or

failure than are outcomes manipulated by a researcher.

Travis (Note 3) suggests classifying recalled success and failure

experiences by achievement domain -- intrapersonal, interpersonal-

affiliative, and mastery-control -- to identify corresponding patterns

in the assignment of causal attributions to performance outcomes.

These three domains are based in part upon the work of Stein and Bailey

(1973) who examined differences in the the types of achievements

claimed by men and women.

Stein and Bailey (1973) challenge the traditional definition of

achievement that narrowly defines it as a collection of events








involving manipulation, mastery, or control such as commanding power or

accumulating wealth. The authors argue that interpersonal-affiliative

events are also legitimate achievements that are simply less public or

monetarily profitable. This affiliative domain is seen as the

speciality of women, as reflected in their skills in the interpersonal,

social realm. In addition to the two classifications based upon Stein

and Bailey's work, Travis (Note 3) has identified a third achievement

domain. This intrapersonal category identifies self-growth and

expansion as a legitimate area of achievement for both sexes.

Finally, dispositional or motivational variables may be another

important source of variance in the responses of women to attributional

questions (McHugh, Frieze, & Hanusa, 1982). Generalizing findings

concerning all men or all women may fail to account for potentially

significant individual differences within each sex. Research that

separates men and women into nonhomogeneous groupings is preferable to

studies that continue to look only at sex differences.


Sex Roles

Attribution literature suggests examining differences in sex role

identities may contribute to determining the effect of individual

differences on patterns of causality (Brewer & Blum, 1978). This

section will provide an overview of theories addressing sex role

identity development and conclude with a review of studies that have

examined the relationship between causal attributions and sex role

identity.


Sex Role Identity

Kagan (1964) and Kohlberg (1966) both define sex role identity as

an acquired self-concept of being masculine or feminine. Kagan





-26-


postulates that sex role identity is the product of differences between

individuals' sex role attributes and their perceptions of sex role

stereotypes. Specifically, Kagan theorizes that individuals compare

their own attributes against sex role stereotypes to arrive at a

relativistic concept of their sex role identity. Simply put, Kagan

views sex role identity as the product of sex role stereotypes and sex

role attributes (Storms, 1979).

Kohlberg theorizes that sex role identities cause the difference

between individuals' sex role attributes and their perceptions of sex

role stereotypes. Kohlberg asserts that sex role identities are firmly

established early in life and thereafter serve to mediate the influence

of sex role stereotypes on the development of sex role attributes. Sex

role identities guide individuals attachment to, evaluation of, and

desire to emulate specific adult role models and abstract or

stereotypical role models. In short according to Kohlberg, sex role

attributes are the product of individuals' sex role identities and

perceptions of sex role stereotypes (Storms, 1979).

Psychological theory and research on sex role identities have

undergone a dramatic shift in the past decade (Major, Carnevale, &

Deaux, 1981). Traditionally, psychologists have accepted the critical

input of sex roles in personality development and psychopathologists

have considered gender identity essential to personal adjustment.

However in recent years, researchers have expressed concern over

possible detrimental effects of traditional sex role standards upon the

full development capabilities of men and women (Bem, 1974; Broverman,

Vogel, Broverman, Clarkson, & Rosenkrantz, 1972; Deutsch & Gilbert,

1976; Erdwins, Small, & Gross, 1980; Gump, 1972).








Traditional sex roles define the masculine identity as possessing

such traits as "aggressive," "dominant," "independent," "willing to

take risks." The traditional feminine sex role identity is one of

"passivity," "nurturance," "submissiveness," "dependence" (Bem, 1974;

Osofsky & Osofsky, 1971; Stake, 1981). Researchers have asserted that

adherence to these traditional sex role standards has outlived its

usefulness and serves only to prevent men and women from developing

their full potential as complete human beings (Bem, 1974; Broverman,

Vogel, Broverman, Clarkson, & Rosenkrantz, 1972).

Rosenkrantz, Vogel, Broverman, and Broverman (1968) assessed the

relationship of self-concept to differentially valued sex role

stereotypes. College students were asked to indicate using 122 bipolar

items what typical adult males, females and they themselves were like.

Results revealed a higher valuation of stereotypically masculine than

feminine characteristics. Broverman, Broverman, Clarkson, Rosenkrantz,

and Vogel (1970) found that mental health clinicians' concepts of a

healthy male did not differ significantly from their concepts of a

healthy adult sex unspecified. However, their concepts of a mature

healthy woman did differ significantly from their adult healthy

concepts.

Deutsch and Gilbert (1976) found women's sex role concepts

regarding their real self, perceptions of their ideal self, and their

belief of what men desire in women were highly dissimilar. This

dissimilarity was not true for the males in this study. Men's

evaluations of their real self, ideal self, and their belief of what

women desire in men were highly similar.

Recent research on sex role identity has proposed that androgynous

individuals exhibit greater behavioral flexibility and interpersonal








adjustment than either sex-typed or sex-reversed individuals do. An

androgynous individual is one who has both masculine and feminine

characteristics and chooses to behave as is situationally appropriate

rather than in a fashion predetermined on the basis of sex (Bem, 1974).

Bem (1975) reported that androgynous individuals of both genders were

able to engage in either instrumental (masculine) or expressive

(feminine) behavior depending on situational contexts and demands.

However, sex-typed and sex-reversed persons engaged in relatively rigid

and stereotyped behaviors regardless of situational characteristics.

Flaherty and Dusek (1980) reported male and female college

students differed on measures of self-esteem factors depending upon

their sex role identity. Subjects classified as androgynous or

masculine scored significantly higher than those classified as

undifferentiated or as feminine on an achievement/leadership factor of

self-esteem. Androgynous students scored higher than undifferentiated

students on an adjustment factor. Androgynous and feminine students

scored higher than masculine and undifferentiated students on a

congeniality/socializability factor. Flaherty and Dusek concluded the

results of their study supported the theory of androgynous flexibility.

Major, Carnevale, and Deaux (1981) found that male and female

androgynous persons were liked by male and female others better than,

and perceived by others as more adjusted than, undifferentiated

persons. Androgynous persons were also generally liked better than,

and perceived as more adjusted than, feminine or masculine individuals.


Sex Role Identity and Causal Attribution Studies

A limited number of studies have examined the relationship between

sex role identification and patterns of causal attributions for








performance outcomes. Brewer and Blum (1979) investigated the

relationship between sex role identity and causal attributions for

performance in math/physical Science courses in an academic setting.

Freshman students completed the Bem Sex Role Inventory (Bem, 1974) and

a questionnaire assessing their attributions of causality for course

performance. Responses to the causal questions were pooled resulting

in a Total Control score for each subject.

Results revealed Total Control scores were significantly more

external among female than male respondents. Females tended to make

more internal attributions for failures in math/physical Science

courses and more external attributions for successes than males. The

authors further found androgynous persons were more likely to perceive

math/physical Science achievements as due to internal sources.

Feminine sex typed persons were more likely to perceive math/physical

Science achievements as due to external sources.

Teglasi (1978) extended the Bar-Tal and Frieze (1977) study

concerning causal attributions and achievement motivation to examine

intercorrelations between these two variables plus subject sex role

orientation. Teglasi found that female undergraduate students with a

traditional role orientation had lower achievement motivation than their

less traditional counterparts. Traditionally oriented women also were

more likely than nontraditional women to attribute failure to lack of

ability but only when working with a male partner. It was further

found that traditional women attributed both success and failure

outcomes to ability or inability when paired with a woman in a

cooperative situation significantly more so than did nontraditional

women. Traditional women also were more likely to attribute success to








ease of task when in the presence of a man than were nontraditional

women.

Wiegers and Frieze (1977) investigated the effects of achievement

level, gender, and female traditionality of career aspirations on

success and failure. Subjects were high school seniors who were

classified according to career choice and grade point average. Results

indicated that males and females high and low in achievement level, and

traditional and nontraditional females, experienced success and failure

differently.

Levine et al. (1982) conducted a study to identify and assess

individual differences between men and women to determine their

mediating role in producing sex-differentiated styles of attribution.

The variables examined were self-esteem, achievement motivation, fear

of success, attitudes toward women, and androgyny. Results revealed

males were more likely than females to attribute their outcomes to

ability and less likely to assign their performance outcomes to effort

and luck.

However, sex differences in attribution patterns were eliminated

when the effects of individual differences were controlled.

Achievement motivation was the best predictor successfully explaining

each of the sex differences for ability, effort, and luck attributions.

Masculinity scores also successfully accounted for sex differences in

luck attributions. The authors suggested that sex of subject was not a

strong predictor of attributions and, in fact, predicted more poorly

than the sex role measure. Levine et al. (1982) suggest levels of

self-esteem may affect causal attributions.





-31-


Self-Esteem

The concept of self-esteem has found an eminent place in

sociological and psychological studies. A formidable body of

literature exists reflecting diverse conceptualizations of this

important construct. This section will examine the theoretical

background of self-esteem. This will be followed by a discussion of

two major self-systems, self-enhancement theory, and self-consistency

theory and their relationships to causal attributions. Finally, the

section will conclude with a discussion of multidimensional theories of

self-esteem and look at the concept of agentic self-esteem.


Theoretical Background

William James (1890, 1892) is generally identified as the earliest

"self" psychologist. James' major theoretical contribution to the

study of the self is the development of the I-Me dichotomy. In this

delineation, the total self or person is differentiated into two

discriminantt aspects": the self as the knower and the self as that

which is known. James perceived identity as divided into three parts:

the material Me which refers to all material things with which a person

experiences a sense of unity, the social Me which refers to recognition

received from others and the spiritual Me which is the states of

consciousness perceived by persons themselves.

Cooley (1902) was the next major theorist to deal with the idea of

the self. He confined his work to that aspect James labeled the Social

Me. Cooley developed the concept of the looking-glass self which

postulates that individuals' conceptions of themselves are determined

by perceptions of other people's reactions to them. According to

Cooley, a sense of self always involves a sense of other people. Mead








(1934, 1956) adopted principles from James and Cooley to expound a

theory that views the self as an I-Me distinction that is also the

product of interactions where persons experience themselves as

reflected in the behaviors of others.

Mead viewed language as an essential part of the development and

operation of the self, i.e., the self as a symbol-using process.

Mead's concept of the "generalized other" was an important contribution

to self theory as it was the first concept to account for a global

cross-situational sense of self. The concept of "generalized other"

implies persons develop the ability to take the role of a group of

others -- real or inferred -- which correspond to society's

representation within the individual.


Self-Systems

Recent research on self-esteem reflects the development of self

theories based upon empirical studies of self-esteem and its correlates

(Wells & Marwell, 1976). Two major theoretical approaches or

self-systems have emerged from self theory: self-enhancement theory

and self-consistency theory. These theories will be discussed as they

apply to performance outcomes of achievement related tasks.

According to self-enhancement theory which is exemplified in the

writings of Cartwright and Zander (1960), Dittes (1959), and Homans

(1961), individuals seek to maximize their self-esteem. That is,

individuals have a need to view themselves as favorably as possible. If

this need is not satisfied it becomes stronger. Consequently,

individuals with any level of chronic self-esteem (high, medium, or

low) will react more favorably to success even if unexpected and less

favorably to failure even if expected.








Self-consistency theory reflected in the work of Heider (1958) and

Lecky (1945) assumes that individuals strive to maintain consistent

attitudes toward themselves. Thus, performance outcomes that are

inconsistent with those attitudes will produce a negative reaction.

Consequently, individuals with high self-esteem will respond more

favorably to success and less favorably to failure than individuals with

low self-esteem. Individuals with low self-esteem will respond more

favorably to failure than success, actually avoiding success to remain

consistent with their own negative expectations.

Self-consistency and self-enhancement theories both contend that

high self-esteem persons will ascribe successful performance outcomes

to internal causal sources and failure outcomes to external sources.

However, the two theories suggest contradictory hypotheses concerning

how low self-esteem persons ascribe causality when they experience

success or failure outcomes. Self-enhancement theory suggests low

self-esteem persons ascribe less internal causality for failure than

success. Self-consistency theory predicts the opposite finding.

Researchers have attempted to produce empirical evidence to

support the legitimacy of these conflicting theories. A limited number

of these studies have examined self-enhancement and self-consistency

theories as applied to attribution research. Fitch (1970) investigated

the causal attributions for performance outcomes of undergraduate

students. Fitch found students attributed significantly more causality

to internal sources for success outcomes than for failure outcomes

supporting the self-enhancement theory. Fitch also found that students

with low self-esteem receiving failure feedback attributed

significantly more causality to internal sources than did high








self-esteem students receiving failure feedback, thereby supporting the

self-consistency theory for low self-esteem subjects.

Burke (1978) investigated the relationship between self-esteem and

causal attributions for performance outcomes. Undergraduate students

with low, medium, and high self-esteem were asked to attribute success

and failure outcomes to ability, effort, luck, or task difficulty.

Results of this study revealed that success was attributed to internal

sources more than failure. Performances consistent with self-esteem

were attributed more than inconsistent outcomes to ability and task

difficulty, and performance outcomes inconsistent with self-esteem were

attributed more than consistent outcomes to luck. Only the attribution

pattern for effort was not found to be largely a function of maintaining

self-consistency.

Burke suggests the results of this study lend general support to

the self-consistency theory. However, Burke also suggests that because

of their unidimensional focus, neither self-system theory was able to

adequately anticipate the relationships that emerged between self-esteem

and performance outcome variables as they related to causal

attributions.


Multidimensional Theories of Self-Esteem

Burke's (1978) belief in the inadequacy of unidimensional theories

of self-esteem is shared by several theorists who in the last decade

have devoted much attention to examining self-esteem from a

multidimensional perspective. Multidimensional theories tend toward the

conceptualization of self-esteem as the function of two processes. The

first process refers to the selected appraisals of significant others

in persons' social environments in the form of social approval. The








second process addresses persons' feelings of efficacy and competence

derived from their own perceptions of their effect on their

environments. The first process leads to feelings of self-worth. The

second process is more closely associated with feelings of achievement,

power, and confidence (Franks & Marolla, 1976).

These two qualitatively different processes or types of

self-esteem are reflected in Coopersmith's (1967) self-esteem theory.

Coopersmith stressed the generality of cross-situational stability of

self-esteem but also recognized its multi-faceted nature and identified

four sources of self-validation. These sources are power -- the

ability to influence and control others, significance -- the acceptance

of attention and affection from others, virtue -- adherence to moral

and ethical standards, and competence -- successful performance in

meeting demands for achievement.

Brissett (1972) suggested a distinction between self-worth, a

concept that pertains to one's effectiveness or mastery in the world,

and self-evaluation, a more socially oriented dimension. Gordon (1969)

proposed four senses of the self: competence, self-determination, unity

and moral worth, and suggested that each of these dimensions is linked

to a universally shared human concern.

The majority of self-esteem measures employed in self-esteem

research are based on unidimensional theories. The assumption

underlying these global measures is that self-evaluations are constant

across dimensions of the self and across situations. The reliance of

studies on purely global measures has been criticized (Franks &

Marolla, 1976). Simpson and Boyle (1975) argue that global self-esteem

measures are too vague and general to allow for specific behavioral

predictions.








In response to this criticism, Stake (1979) developed an

instrument specifically designed to tap one dimension of self-esteem.

The Performance Self-Esteem Scale (PSES) is based upon Bakan's (1966)

work in which he explored his perceptions of the duality of humankind's

nature. Bakan adopted the terms "agency" and "communion" to

characterize two fundamental modalities of human existence.

Agency addresses existence as the individual and manifests in

self-protection, self-assertion, self-expansion, isolation, and

alienation. Agency is further manifested in a repression of thought,

feeling and impulse, and the urge to master. Communion refers to the

participation of the individual in some larger organism of which the

individual is a part. This collective experience is manifested in

contact, openness, union, noncontractual cooperation, and in a lack of

repression.

The Performance Self-Esteem Scale assesses an individual's agentic

qualities. This "agentic" self-esteem is concerned with achievement

and performance behaviors and is reflected in assertiveness,

independence, and an ability to be influential in a task oriented

group. The PSES assesses a broad range of self-evaluations of ability

and performance while excluding other self-evaluations such as the

ability to develop positive interpersonal relationships, likeability,

and moral goodness which are communal characteristics.


Traditional and Nontraditional Career Women

Research investigating causal attributions in the field has

focused on nontraditional occupations. Studies have shown that women

choosing nontraditional careers differ from women choosing traditional

careers across a wide variety of dimensions. Results of studies








examining one population cannot be generalized to the other creating

the need for studies examining traditional career groups. This section

will examine studies comparing women choosing traditional careers to

women choosing nontraditional careers.

The phrase nontraditional career choice refers to the selection of

an occupation which has been traditionally stereotyped as the exclusive

domain of the opposite sex (Auster & Auster, 1981). Women choosing

traditional career paths have been shown to differ from women choosing

nontraditional career paths across various dimensions.

Karman (1972) compared female college students who chose

traditional (stereotypically feminine) or nontraditional

(stereotypically masculine) careers on the basis of family background,

socioeconomic, and personality characteristics. The students with

nontraditional aspirations came from families with higher incomes, had

more educated mothers, were more theoretically oriented, held more

liberal attitudes toward the role of women in society, and did better

academically than their traditional counterparts.

Nagley (1971) found similar differences in a study of "pioneers"

and traditional career women. Women in nontraditional fields were more

committed to their careers, had well educated and supportive fathers,

had husbands employed at higher occupational levels, and had

cooperative home decision making responsibilities. Tangri (1972)

studied women students in their senior year of college that chose

traditionally masculine professions. The author found that these

"role-innovators" were less likely to displace their achievement

concerns onto their present or future husbands than women choosing





-38-


traditional careers. Role-innovators were also found to have higher

success goals and to express greater commitment to their careers.

Peng and Jaffee (1979) examined multiple factors that may

influence the entry of women into nontraditional fields as they move

from high school into college. Specifically, the study investigated

the direct relationships of family background characteristics, high

school experience, academic ability, and plans and attitudes toward the

choice of nontraditional fields among college women students. The

results of this study revealed that high school course work, academic

ability, a success orientation, and educational plans were important

predictors for women's entry into nontraditional fields.

Heilman (1979) suggested that the sexual composition of an

occupation influences preferences for and perceptions of personal

success in that occupation. In an investigation of high school

students' attitudes toward nontraditional occupations, Heilman found

that females perceived greater potential personal success in

occupations with high percentages of women than did males. Harren,

Ross, Tinsley, and Moreland (1979) found that sex role orientation and

gender were significant predictors of college students choice of

occupation.

Collins, Reardon, and Waters (1980) examined the influence of both

sex and sex role orientation on expressed interest and perceived

personal success of female and male college students for male-dominated

occupations. The authors concluded that perceptions of success for

females but not for males were influenced by the percentage of females

in the occupation. High masculine males reported greater interest and

perceived success in male-dominated occupations than low masculine








males. Low feminine oriented women perceived greater success than high

feminine women. Wertheim, Widom, and Wortzel (1978) found men and

women in traditionally female occupations scored higher on femininity

scales and reported significantly lower educational status for their

mothers than those subjects in traditionally masculine occupations.


Summary

A review of the literature indicates that future research must

focus on situational and individual differences in order to determine

the significance of sex differences in patterns of causal attributions

to achievement behavior. Sex role identity and self-esteem were

suggested as two possible variables affecting causal attributions to be

investigated. The number of studies examining individual differences in

causal attributions was limited, particularly those studies conducted in

real life settings. Those studies that were conducted in the field

examined nontraditional or stereotypically masculine professions.

Differences between women in traditional and nontraditional fields were

discussed supporting the legitimacy of extending investigations of

causal attributions to examining persons in traditionally female

fields.














CHAPTER THREE

METHODOLOGY


The purpose of this study was to investigate the effects of

counselors' sex, sex role identities, and levels of agentic self-esteem

on their causal attributions for successful and unsuccessful

performance outcomes. The hypotheses, population, and sample,

instruments, research procedures, analyses of data, and limitations of

this study are discussed in this chapter.


Null Hypotheses


Hypothesis One. There are no significant differences in counselors'

causal attributions for successful performance outcomes when compared

by their sex.

Hypothesis Two. There are no significant differences in counselors'

causal attributions for successful performance outcomes when compared

by their sex role identities.

Hypothesis Three. There are no significant differences in counselors'

causal attributions for successful performance outcomes when compared

by their levels of agentic self-esteem.

Hypothesis Four. There are no significant differences in counselors'

causal attributions for unsuccessful performance outcomes when compared

by their sex.








Hypothesis Five. There are no significant differences in counselors'

causal attributions for unsuccessful performance outcomes when compared

by their sex role identities.

Hypothesis Six. There are no significant differences in counselors'

causal attributions for unsuccessful performance outcomes when compared

by their levels of agentic self-esteem.


Population and Sample

The population for this study consisted of the 5,648 members of

the American Mental. Health Counselors Association (AMHCA). AMHCA is

one of 13 divisions of the American Personnel and Guidance Association

(APGA). Any person in public or private practice whose primary

responsibility is in the area of mental health counseling or

consultation is eligible for AMHCA membership. Students enrolled in an

academic program for mental health counseling may also join. There are

no additional requirements to obtain membership in AMHCA.

Membership statistics are gathered through an AMHCA survey that is

completed by all new members. Due to AMHCA's rapidly growing

membership roles, information is not available on all current members.

However, statistics are available on the approximately 4,700 members

included in the December, 1981, American Personnel and Guidance

Association statistics report.

In December, 1981, there were 2,080 male AMHCA members and 2,636

female members. The majority of these counselors were employed in

either a college/university or private practice setting. Table 2

presents the AMHCA membership according to primary employment setting.





-42-


Table 2

AMHCA Membership by Primary Employment Setting


Employment Setting Number of Members


College or University 671

Private Practice 650

Private Counseling Center 577

Community Mental Health Center 524

Community Agency 433

Rehabilitation Program or Agency 210

State or Local Government Agency 183

Secondary or Senior High School 158

Parochial or Private Institution 109

Elementary School 107

Junior or Community College 96

Federal Government 76

Middle School 74

Association or Foundation 51

Corrections 28

Probation and Parole 24

Family Services 7

Other Settings 391


The typical AMHCA member holds a master's degree. Table 3 presents the

AMHCA membership according to academic credentials.








Table 3

AMHCA Membership by Academic Credentials


Academic Credential


Number of Members


Doctoral Degree 933

Educational Specialist Degree 146

Master's Degree 3,050

Bachelor's Degree 478

Associated Degrees or Certificates 26

Other 133


The statistics on AMHCA members' primary job responsibilities reveal

over one half of the membership have positions as counselors. Table 4

presents the AMHCA membership according to primary job responsibility.


Table 4

AMHCA Membership by Primary Job Responsibility


Primary Job Responsibility


Counselor

Counselor Educator

Administrators

Supervisors/Consultants

Student Personnel Work

Paraprofessional

Other

Student


Number of Members


2,467

369

298

280

37

28

670

589








AMHCA's membership by race is presented in Table 5.


Table 5

AMHCA Membership by Race



Race Number of Members


White 4,172

Black 310

Native American 65

Hispanic 54

Asian American 32

Other 67


The American Personnel and Guidance Association offers access to

the membership rolls of its' 13 divisions for a fee. Mailing lists in

lots of 1,000 can be purchased according to division, state, zip code

area, members' primary job responsibility, members' work setting, or any

combination of these categories. For this study, 1,000 names were

randomly selected by APGA computer from the approximately 2,500 AMHCA

members who specified their primary job responsibility as counselors.

An additional criterion for inclusion in this study was that

subjects have a minimum of a master's degree. As mailing lists

indicating members' academic credentials could not be purchased, a

Demographic Information Questionnaire (DIQ) to obtain this and other

demographic information was completed by all respondents. Subjects

indicating on the OIQ they had less than a master's degree or had a

primary job responsibility other than counselor were not included in

the data analyses.








Every second counselor from the 1,000 name mailing list was chosen

until 600 names were selected for inclusion in the study. The final

sample included 127 counselors. Of this 127, 62 were males and 65 were

females. The total return sample was 158 counselors or a 26 percent

return rate. Of the respondents not included in the final sample, 12 did

not adequately complete the instruments, 12 returned their instruments

after the data analyses were completed, four did not meet the criteria

for inclusion in the study and three stated they did not wish to

participate.


Instruments


The Bem Sex Role Inventory

The Bem Sex Role Inventory (BSRI) was developed in 1974 by Sandra

Bem to assess respondents' sex role identities (Appendix A). The BSRI

consists of a Masculinity and Femininity Scale each containing 20

personality characteristics. Bem selected these characteristics on the

basis of sex-typed social desirability rather than on a differential

endorsement by men and women as is true of other sex role inventories.

In developing the instrument, characteristics qualified as feminine if

they were judged to be more desirable for a woman than a man in American

society, and qualified as masculine if they were judged to be more

desirable for a man than a woman in American society.

The BSRI also contains characteristics that are completely neutral

with respect to sex. These items comprise the Social Desirability

Scale which provides a neutral context for the Masculinity and

Femininity Scales. The Social Desirability Scale was originally








included to ensure the BSRI was not merely tapping a general tendency

to endorse socially desirable traits.

The BSRI is a self-administering paper and pencil instrument. It

requires respondents indicate on a seven point Likert scale how

accurately each of 60 personality characteristics describes them.

Possible responses range from one indicating the characteristic is

"never or almost never true" for the respondent to seven, indicating

the characteristic is "true or almost always true" for the respondent.

On the basis of their responses, respondents receive a Masculinity

Score and a Femininity Score. Masculinity Scores are determined by the

mean self-rating for all endorsed feminine characteristics. Group

medians for the Masculinity and Femininity Scores for the subject

population are then calculated to determine the respondents' sex role

classification.

Those respondents whose scores fall above the masculinity median

and below the femininity median are classified as masculine. Those

respondents whose scores fall above the femininity median but below the

masculinity median are classified as feminine. Those respondents whose

scores fall above both the masculinity and femininity medians are

classified as androgynous. Those respondents whose scores fall below

both the masculinity and femininity medians are classified as

undifferentiated.

Bem (1974) collected normative data on 444 male and 279

undergraduate students from Stanford University and 117 male and 77

female paid volunteers from Foothill Junior College. Coefficient alpha

was computed separately for the Masculinity, Femininity, and Social

Desirability Scales for each of the two normative samples to estimate








the internal consistency of the BSRI. The Stanford students' and

Foothill Junior College volunteers' scores were respectively as

follows: Masculinity .86 and .86, Femininity .80 and .82, and Social

Desirability 75 and .70.

Using a formula for linear combinations, coefficient alpha was

computed for the androgynous difference scores. The reliability of the

androgyny difference score was .85 for the Stanford University sample

and .86 for the Foothill Junior College sample. Test-retest reliability

for the BSRI was also computed using students from the Stanford sample.

Approximately four weeks after the initial test administration, 28 males

and 28 females were re-administered the BSRI.

Product moment correlations were computed between the first and

second administrations for the Masculinity, Femininity, Androgynous,

and Social Desirability Scales. All four correlations proved to be

highly reliable over this time interval. Respectively correlations

were .90, .90, .93, .89. All correlations were statistically

significant at the p < .05 or better level. Finally, mean scores were

computed by sex for the two normative samples. Males scored

significantly higher than females on the Masculinity Scale (p < .001)

while females scored significantly higher than males on the Femininity

Scale (p < .001).


The Performance Self-Esteem Scale

The Performance Self-Esteem Scale (PSES) was developed by Jayne E.

Stake in 1979 to measure self-evaluations of ability and performance

(Appendix B). The PSES was based upon the argument that global

self-esteem measures were too general to allow for specific behavioral

predictions. Therefore, specific self-esteem measures that identify





-48-


separate aspects or types of self-esteem were needed (Simpson & Boyle,

1975).

The PSES is a self-rating instrument on which respondents are

asked to indicate, using a seven point Likert scale, the extent to

which each of 47 items are self-descriptive. Possible responses range

from one to seven. A response of one indicates the item is "never or

almost never true" for the respondent. A response of seven indicates

the item is "always or almost always true" for the respondent.

The PSES contains a 40 item performance self-esteem scale that

relates to ability and performance. Scoring procedures indicate

whether each of these 40 items is scored in a positive or negative

direction. The PSES also contains a seven item social self-esteem

scale. These seven items relate to the respondents' perceptions of how

they are liked by others. A total performance self-esteem score is

derived for each respondent by summing the responses to the positive

performance self-esteem items, summing the responses to the negative

performance self-esteem items, and subtracting the sum of the negative

items from the sum of the positive items.

For the purposes of this study the performance self-esteem scores

of all respondents were divided into quartiles to determine the

respondents self-esteem classifications. Those respondents whose scores

fell above the first quartile were classified as high-high self-esteem.

Those respondents whose scores fell below the first quartile but above

the second quartile were classified high-low self-esteem. Those

respondents whose scores fell below the second quartile but above the

third quartile were classified low-high self-esteem. Those respondents








whose scores fell below the third quartile were classified low-low

self-esteem.

A check of the internal consistency of the PSES using 35 male and

54 female undergraduate students revealed a coefficient alpha of .90.

The discriminative validity of the PSES was tested in three ways.

First, the relationship between the performance self-esteem score and

the sum of the social self-esteem items was examined. Since both

represent aspects of overall self-evaluation, the relationship was

expected to be low but positive. The Pearson correlation between the

two measures was +.27 (p < .01) for the same sample of 89 undergraduates

that supplied the reliability data. The author concluded that this

finding offered evidence that the PSES measures a separate and distinct

factor of self-esteem.

Stake next examined the relationship between the PSES and a

defensive response set. Forty-four male and 44 female undergraduates

were administered the PSES and the K subscale of the Minnesota

Multiphasic Personality Inventory. The K scale taps defensiveness in

responding and is associated with high global self-esteem and adjustment

(Dahlstrom, Welsh, & Dahlstrom, 1972). The correlation between the PSES

and the K scale was expected to be low but positive as both measure

aspects of self-esteem. The Pearson correlation was found to be +.25

(p < .05) indicating the PSES was not closely related to this measure of

defensiveness.

A third concern was that the PSES measure self-evaluations

independent of general sex role attitudes. This concern was due to the

PSES including many items that represent traits and behaviors considered

nontraditional for women. A sample of 80 female and 94 male








undergraduates completed the PSES and Attitudes Toward Women Scale

(Spence & Helmrich, 1972). This instrument measures traditionality of

sex role attitudes. The correlation was not significant for females

(r = +.04) or males (r = -.10) indicating the PSES measures

self-evaluations that are independent of general sex role attitudes.


The Causal Attribution Scale

The Causal Attribution Scale (CAS) Forms A and B is a structured

rating measure developed by this researcher to assess causal

attributions for self-reported performance outcomes (Appendix C). The

original form of the CAS was developed to evaluate the extent to which

respondents perceive the causal factors ability, effort, luck, and task

difficulty as responsible for their performance outcomes. The final

form of the CAS used in this investigation was expanded to examine a

broader range of causal factors than the four factors typically

investigated in attribution research.

The CAS is a self-administering paper and pencil instrument that

has three sections. The first section of form A requires respondents

describe a time when they were very successful in their present

positions. Section two asks respondents to characterize this

experience according to one of the following three achievement areas:

interpersonal event, intrapersonal-affiliative event, mastery-control

event.

Finally in section three, respondents are asked to rate on a five

point Likert scale the extent to which each of the following nine

factors was responsible for their successful performance outcomes:

ability, typical effort, mood at time of task, immediate effort, task

ease/difficulty, luck, interest in task, supervisor, help from others.








A response of one indicates the factor was not at all a cause of the

experience reported. A response of five indicates the factor was very

much a cause of the experience reported. Form B asks respondents to

provide answers to the same three sections for a time they were very

unsuccessful in their present positions.

To obtain an external evaluation of the content validity of the

CAS, 10 experts were asked to examine the original CAS and evaluate its

relevancy to this research. These 10 experts were chosen based upon

their publications and research in the area of sex differences and

causal attributions. Each expert was sent a letter introducing this

researcher, the purpose of the research, and requesting assistance

(Appendix D), the original CAS Forms A and B (Appendix E), and two

evaluation forms one each for Forms A and B (Appendix F).

Each evaluation form corresponded to the CAS by section. The

experts were asked to indicate whether each section was adequate as is,

adequate with the following changes, or inadequate for the following

reasons. Five of the 10 experts contacted responded to the request for

assistance in establishing the content validity of the CAS. Of the five

experts, for both Forms A and B four said section one was adequate as

is, four said section two was adequate as is, and five said section

three was adequate as is. One expert said section one and section two

were inadequate.

Although the majority of experts stated all three sections were

adequate, several suggestions were offered that led to the final form

of the CAS employed in this investigation. The experts' evaluations of

the CAS are reported in full in Appendix G. Overall, the experts

evaluating the CAS believed it to be adequate in assessing causal








attributions for successful and unsuccessful performance outcomes in

naturalistic contexts thereby establishing the scales' content

validity.


The Demographic Information Questionnaire

The Demographic Information Questionnaire (DIQ) was developed by

this researcher to obtain statistical data on the participants

(Appendix H). The DIQ consists of four fill-in questions requesting

the following information: respondents' age, sex, educational level

and primary job responsibility. Information concerning respondents'

sex, educational level, and job responsibility was needed for accurate

classification of subjects and for a final determination concerning

respondents inclusion in the subject pool. The categories under

educational level and primary job responsibility were those used by the

American Mental Health Counselors Association to obtain their

demographic membership information.

The Demographic Information Questionnaire is a self-administering

paper and pencil instrument. Respondents are asked to either fill in a

blank or select among several choices the appropriate answer to four

questions.


Procedures

A mailing list consisting of 1,000 American Mental Health

Counselors Association (AMHCA) members' names and addresses was

purchased through the American Personnel and Guidance Association

(APGA). These 1,000 names were randomly selected by APGA computer from

the 2,500 AMHCA members who indicated on the AMHCA membership form their

primary job responsibility as counselor.








Every other counselor from the 1,000 name mailing list was

included in this study. This sampling procedure was repeated until 600

names were selected. These 600 counselors were mailed a packet

containing a letter introducing this researcher, the purpose of the

research, a request for participation in the study, and general

instructions for completing the instruments (Appendix I), the Bem Sex

Role Inventory, the Performance Self-Esteem Scale, the final form of

the Causal Attribution Scale Forms A and B, the Demographic Information

Questionnaire, and a stamped self-addressed envelope.

The general letter of instructions requested participants

carefully read the individual instructions included with each

instrument. Participants were asked not to omit any items answering

each to the best of their capabilities. Instructions were also

included to return all forms in the enclosed self-addressed envelope as

soon as possible. Only those participants who adequately completed

each instrument were included in the data analyses. Respondents

indicating a primary job responsibility other than counselor and

respondents with less than a master's degree were also excluded from

the sample. To ensure anonymity, code numbers were assigned to each

respondent.


Analyses of Data

One way analyses of variance (ANOVA) were used to determine if

differences existed in the causal attributions counselors used to

explain their successful performance outcomes when compared by their

sex. One way analyses of variance were also used to determine if

differences existed in the causal attributions counselors used to

explain their successful performance outcomes when compared by their








sex role identities and when compared by their levels of agentic

self-esteem. Where significant main effects were found, Least

Significant Differences (LSD) tests were conducted for follow-up

analyses.

One way analyses of variance were used to determine if differences

existed in the causal attributions counselors used to explain their

unsuccessful performance outcomes when compared by their sex. One way

analyses of variance were also used to determine if differences existed

in the causal attributions counselors used to explain their unsuccessful

performance outcomes when compared by their sex role identities and when

compared by their levels of agentic self-esteem. Where significant main

effects were found, Least Significant Differences (LSD) tests were

conducted for follow-up analyses.


Limitations of the Study

One of the limitations of this study was the self-selection of

counselors as participants. Counselors who chose to participate in

this study may not be representative of counselors in general. A

second limitation was the possibility that counselors' responses to the

instruments were inaccurate or biased self-reports.

Although the sample for this survey was drawn from a nationwide

population and attempts were made to establish the diversity of AMHCA's

membership, there were no comparison groups in this study. Therefore, a

third limitation of this study was that any conclusions drawn were

limited to counselors who meet AMHCA's membership requirements and the

requirements for inclusion in this study.














CHAPTER FOUR

RESULTS AND DISCUSSION


Results

The purpose of this study was to investigate the effects of

counselors' sex, sex role identities, and levels of agentic self-esteem

on their causal attributions for successful performance outcomes. One

hundred and twenty-seven counselors participated in the study. Of that

number, 62 were males and 65 were females. Participants completed the

Bem Sex Role Inventory (BSRI), the Performance Self-Esteem Scale (PSES),

the Causal Attribution Scale Forms A and B (CAS), and the Demographic

Information Questionnaire (DIQ). Data analyses were conducted as

outlined in Chapter Three.


Hypothesis One

There are no significant differences in counselors' causal

attributions for successful performance outcomes when compared by their

sex. Nine one-way analyses of variance were conducted to test this

hypothesis. The results in Table 6 indicate there were no significant

differences in the nine causal attributions counselors used to explain

their successful performance outcomes when compared by their sex

(p < .01). These results mean male and female counselors did not

attribute the causes of their successful performance outcomes to

different factors. Therefore, the first null hypothesis was not

rejected.








Table 6

One-Way Analyses of Variance of Causal Attributions for Counselors'
Successful Performance Outcomes by Sex


Causal Source of Degrees of Sums of F-Ratio
Attribution Variation Freedom Squares


Sex 1 0.76322268 1.51
Ability Explained 1 0.76322268
Residual 125 63.33126551
TOTAL 126 64.09448819

Sex 1 4.34356695 5.13
Typical Explained 1 4.34356695
Effort Residual 125 105.84540943
TOTAL 126 110.18897638

Sex 1 1.31228399 3.42
Interest Explained 1 1.31228399
in Task Residual 124 47.54485887
TOTAL 125 48.85714286

Sex 1 0.26030167 0.34
Immediate Explained 1 0.26030167
Effort Residual 125 94.62158809
TOTAL 126 94.88188976

Sex 1 2.50371036 1.29
Mood at Explained 125 2.50371036
Time of Task Residual 125 242.85062035
TOTAL 126 245.35433071

Sex 1 0.08138230 0.08
Supervisor Explained 1 0.08138230
Residual 117 115.90181098
TOTAL 118 115.98319328

Sex 1 2.22365540 2.04
Good Luck Explained 1 2.22365540
Residual 124 135.07793190
TOTAL 125 137.30158730

Sex 1 1.94994432 6.50
Task Ease Explained 1 1.94994432
Residual 125 37.48312655
TOTAL 126 39.43307087

Sex 1 3.48765988 1.89
Help From Explained 1 3.48765988
Others Residual 124 229.36948298
TOTAL 125 232.85714286








Hypothesis Two

There are no significant differences in counselors' causal

attributions for successful performance outcomes when compared by their

sex role identities. Nine one-way analyses of variance were conducted

to test this hypothesis. The results in Table 7 reveal there were

significant differences in counselors use of the attribution typical

effort to explain their successful performance outcomes when compared by

their sex role identities (p < .01). These results mean counselors did

not assign the same causal responsibility to their typical effort when

compared by their sex role identities.


Table 7

One-Way Analysis of Variance of Typical Effort Attribution for
Counselors' Successful Performance Outcomes by Sex Role Identities



Causal Source of Degrees of Sums of F-Ratio
Attribution Variation Freedom Squares


SRI 3 11.92296653 4.97*

Typical Explained 3 11.92296653
Effort
Residual 123 98.26600985

TOTAL 126 110.18897638


Note: SRI is the abbreviation for sex role identities

*p < .01

A Least Significant Differences test was conducted to further

investigate these differences. The results of this analysis are

reported in Table 8. Undifferentiated sex-typed counselors differed

significantly from androgynous sex-typed counselors (p < .001) and from

masculine sex-typed counselors (p < .01) in their use of the attribution








typical effort to explain their successful performance outcomes. These

results mean undifferentiated sex-typed counselors did not assign the

same causal responsibility to their typical effort as androgynous and

masculine sex-typed counselors.

Table 9 shows the mean score of undifferentiated sex-typed

counselors was significantly lower than the mean score of androgynous

and masculine sex-typed counselors. While the mean score of

undifferentiated sex-typed counselors was also lower than the mean score

of feminine sex-typed counselors, this difference did not reach

significance. These results mean undifferentiated sex-typed counselors

were significantly less likely than the androgynous and masculine

sex-typed counselors to attribute their successful outcomes to their

typical effort.


Table 8

Least Significant Differences Test of Counselors'
Sex-Role Identities for Typical Effort Attribution



Causal Source of Estimate T-Value
Attribution Variation


Androgynous vs Masculine 0.17142857 0.80

Androgynous vs Feminine 0.34285714 1.51

Typical Androgynous vs Undifferentiated 0.83054187 3.70**
Effort
Masculine vs Feminine 0.17142857 0.76

Masculine vs Undifferentiated 0.65911330 2.94*

Feminine vs Undifferentiated 0.48768473 2.06


*p < .01

**p < .001








Table 9

Means and Standard Deviations for Typical Effort Attribution
by Counselors' Sex Role Identities


Causal Group N Mean SD
Attribution


Androgynous 35 4.4857 0.781

Typical Masculine 35 4.3142 0.832
Effort
Feminine 28 4.1428 0.890

Undifferentiated 29 3.6551 1.078


The results in Table 21 (Appendix J) reveal that counselors did not

differ significantly in their use of the remaining eight causal

attributions when compared by their sex role identities. Therefore, for

the causal attribution typical effort, hypothesis two was rejected. For

the remaining eight causal attributions investigated in this study,

hypothesis two was not rejected.


Hypothesis Three

There are no significant differences in counselors' causal

attributions for successful performance outcomes when compared by their

levels of agentic self-esteem. Nine one-way analyses of variance were

conducted to test this hypothesis. The results in Table 10 reveal there

were significant differences in counselors use of the attributions

ability (p < .001) and typical effort (p < .0001) to explain their

successful performance outcomes when compared by their levels of agentic

self-esteem. These results mean counselors did not assign the same

causal responsibility to their ability and typical effort when compared

by their levels of agentic self-esteem.








Table 10

One-Way Analyses of Variance of Ability and
Typical Effort Attributions for Counselors' Successful
Performance Outcomes by Levels of Agentic Self-Esteem


Causal Source of Degrees of Sums of F-Ratio
Attribution Variation Freedom Squares


LASE 3 8.53327639 6.30*

Ability Explained 3 8.53327639

Residual 123 55.56121180

TOTAL 126 64.09448819

LASE 3 22.37411871 10.45**

Typical Explained 3 22.37411871
Effort
Residual 123 87.81485767

TOTAL 126 110.18897638


Note: LASE is the abbreviation for levels of agentic self-esteem.

*p < .001

**p < .0001

Least Significant Differences tests were conducted to further

investigate these differences. The results of this analysis are

reported in Table 11. Low-low agentic self-esteem counselors differed

significantly from high-low agentic self-esteem counselors (p < .0005)

and from high-high agentic self-esteem counselors (p < .0005) in their

use of the attribution ability to explain their successful performance

outcomes. These results mean low-low agentic self-esteem counselors did

not assign the same causal responsibility to their ability as high-low

and high-high agentic self-esteem counselors.





-61-


Table 11 further reveals low-low agentic self-esteem counselors

differed significantly from low-high, high-low, and high-high agentic

self-esteem counselors in their use of the causal attribution typical

effort to explain their successful performance outcomes (p < .0001).

These results mean low-low agentic self-esteem counselors did not assign

the same causal responsibility to their typical effort as counselors

with higher levels of agentic self-esteem.


Table 11

Least Significant Differences Tests of Counselors' Levels of
Agentic Self-Esteem for Ability and Typical Effort Attributions



Causal Source of Estimate T-Value
Attribution Variation


Low-low vs Low-high -0.39338493 -1.92

Low-low vs High-low -0.58823529 -3.58*

Ability Low-low vs High-high -0.65517241 -3.83*

Low-high vs High-low -0.26565465 -1.59

Low-high vs High-high -0.33259177 -1.92

High-low vs High-high -0.06693712 -0.39


Low-low vs Low-high -0.86999022 -4.12**

Low-low vs High-low -0.83868093 -4.06**

Typical Low-low vs High-high -1.10135841 -5.12**
Effort
Low-high vs. High-high 0.03130930 0.15

Low-high vs High-high -0.23136819 -1.06

High-low vs. High-high -0.26267748 -1.23


*p < .0005

**p < .0001








Table 12 shows the mean score of low-low agentic self-esteem

counselors was lower than the mean scores of high-low and high-high

agentic self-esteem counselors for the attribution ability. While the

mean score of the low-low agentic self-esteem counselors was also lower

than the mean score of the low-high agentic self-esteem counselors, this

difference did not reach significance. These results mean low-low

agentic self-esteem counselors were significantly less likely than the

high-low and high-high agentic self-esteem counselors to attribute

their successful outcomes to their ability.

Table 12 further reveals the mean score of the low-low agentic

self-esteem counselors was significantly lower than the mean scores of

the low-high, high-low, and high-high agentic self-esteem counselors

for the causal attribution typical effort. These results mean low-low

agentic self-esteem counselors were significantly less likely than the

low-high, high-low and high-high agentic self-esteem counselors to

attribute their successful outcomes to their typical effort.

The results in Table 22 (Appendix J) reveal that counselors did

not differ significantly in their use of the remaining seven causal

attributions when compared by their levels of agentic self-esteem.

Therefore, for the causal attributions ability and typical effort,

hypothesis three was rejected. For the remaining seven causal

attributions investigated in this study, hypothesis three was not

rejected.


Hypothesis Four

There are no significant differences in counselors' causal

attributions for unsuccessful performance outcomes when compared by

their sex. Nine one-way analyses of variance were conducted to test





-63-


Table 12



Means and Standard Deviations for Ability and Typical Effort
Attributions by Counselors' Levels of Agentic Self-Esteem



Causal Group N Mean SD
Attribution


Low-low 33 4.0000 0.790

Low-high 31 4.3225 0.701

Ability High-low 34 4.5882 0.608

High-low 29 4.6551 0.552

Low-low 33 3.4848 1.064

Typical Low-high 31 4.3548 0.709
Effort
High-low 34 4.3235 0.806

High-high 29 4.5862 0.732


this hypothesis. The results in Table 13 indicate there were no

significant differences in the nine causal attributions counselors used

to explain their unsuccessful performance outcomes when compared by

their sex (p < .01). These results mean male and female counselors did

not attribute the causes of their unsuccessful performance outcomes to

different factors. Therefore, hypothesis four was not rejected.


Hypothesis Five

There are no significant differences in counselors' causal

attributions for unsuccessful performance outcomes when compared by

their sex role identities. Nine one-way analyses of variance were

conducted to test this hypothesis. The results in Table 14 indicate

there were no significant differences in the nine causal attributions








Table 13

One-Way Analyses of Variance of Causal Attributions for
Counselors' Unsuccessful Performance Outcomes by Sex


Causal Source of Degrees of Sums of F-Ratio
Attribution Variation Freedom Squares


Inability




Typical
Effort



Interest
in Task



Immediate
Effort



Mood at
Time of Task



Supervisor




Bad Luck




Task Difficulty


Lack of
Help From
Others


Sex
Explained
Residual
TOTAL

Sex
Explained
Residual
TOTAL

Sex
Explained
Residual
TOTAL

Sex
Explained
Residual
TOTAL

Sex
Explained
Residual
TOTAL

Sex
Explained
Residual
TOTAL

Sex
Explained
Residual
TOTAL

Sex
Explained
Residual
TOTAL

Sex
Explained
Residual
TOTAL


4.49169119
4.49169119
178.61941992
183.11111111

3.79801107
3.79801107
163.69405242
167.49206349

2.76684207
2.76684707
205.28039702
208.04724409

2.84174987
2.84174987
223.26848635
226.11023622

1.34925265
1.34925265
267.98933002
269.33858268

0.07340148
0.07340148
82.80263158
82.87603306

0.02813544
0.02813544
87.57816377
87.60629921

7.89374963
7.89374963
241.52357320
249.41732283


1 3.02599366
1 3.02599366
125 277.33114919
126 280.35714286


3.12




2.88




1.68




1.59




0.63




0.11




0.04




4.09


1.35








Table 14

One-Way Analyses of Variance of Causal Attributions for
Counselors' Unsuccessful Performance Outcomes by Sex Role Identities



Causal Source of Degrees of Sums of F-Ratio
Attribution Variation Freedom Squares


SRI 3 9.53606040 2.23
Inability Explained 3 9.53606040
Residual 122 173.57505071
TOTAL 125 183.11111111

SRI 3 5.64226633 1.42
Typical Explained 3 5.64226633
Effort Residual 122 161.84979716
TOTAL 125 167.49206349

SRI 3 2.53886971 0.51
Interest Explained 3 2.53886971
in Task Residual 123 205.50837438
TOTAL 126 208.04724409

SRI 3 7.29742834 1.37
Immediate Explained 3 7.29742834
Effort Residual 123 218.81280788
TOTAL 126 226.11023622

SRI 3 6.44400140 1.00
Mood at Explained 3 6.44400140
Time of Task Residual 123 262.89458128
TOTAL 126 269.33858268

SRI 3 3.91428571 0.58
Supervisor Explained 3 3.91428571
Residual 117 276.44285714
TOTAL 120 280.35714286

SRI 3 2.75038788 1.33
Bad Luck Explained 3 2.75038788
Residual 123 84.85591133
TOTAL 126 87.60629921

SRI 3 12.87446569 2.23
Task Difficulty Explained 3 12.87446569
Residual 123 236.54285714
TOTAL 126 249.41732283

Lack of SRI 3 3.91428571 0.58
Help From Explained 3 3.91428571
Others Residual 122 276.44285714
TOTAL 125 280.35714286








counselors used to explain unsuccessful performance outcomes when

compared by their sex role identities. These results mean androgynous,

masculine, feminine, and undifferentiated sex-typed counselors did not

attribute the causes of their unsuccessful performance outcomes to

different factors. Therefore, hypothesis five was not rejected.


Hypothesis Six

There are no significant differences in counselors' causal

attributions for unsuccessful performance outcomes when compared by

their levels of agentic self-esteem. Nine one-way analyses of variance

were conducted to test this hypothesis. The results in Table 15 reveal

there were significant differences in counselors use of the attribution

immediate effort to explain their unsuccessful performance outcomes when

compared by their levels of agentic self-esteem (p < .01). These

results mean counselors did not assign the same causal responsibility to

their immediate effort when compared by their levels of agentic

self-esteem.


Table 15

One-Way Analyses of Variance of Immediate
Effort Attribution for Counselors' Unsuccessful
Performance Outcomes by Levels of Agentic Self-Esteem



Causal Source of Degrees of Sums of F-Ratio
Attribution Variation Freedom Squares


LASE 3 23.40979564 4.74*

Immediate Explained 3 23.40979564
Effort
Residual 123 202.70044058

TOTAL 126 226.11023622

*p < .01








A Least Significant Differences test was conducted to further

investigate these differences. The results of this analysis are

reported in Table 16. High-high agentic self-esteem counselors differed

significantly from low-low and low-high agentic self-esteem counselors

in their use of the causal attribution immediate effort to explain their

unsuccessful performance outcomes (p < .01). These results mean

high-high agentic self-esteem counselors did not assign the same causal

responsibility to their immediate effort as low-low and low-high agentic

self-esteem counselors.


Table 16

Least Significant Differences Test of Counselors' Levels of
Agentic Self-Esteem for Immediate Effort Attribution



Causal Source of Estimate T-Value
Attribution Variation


Immediate Low-low vs Low-high -0.13978495 -0.44
Effort

Low-low vs High-low 0.57843137 1.84

Low-low vs High-high 0.94252874 2.88*

Low-high vs High-low 0.71821632 2.25

Low-high vs High-high 1.08231368 3.26*

High-low vs High-high 0.36409736 1.12


*p < .01

Table 17 shows the mean score of high-high agentic self-esteem

counselors was significantly lower than the mean scores of low-low and

low-high agentic self-esteem counselors for the causal attribution

immediate effort. While the mean score of the high-high agentic





-68-


self-esteem counselors was also lower than the mean score of the

high-low agentic self-esteem counselors, this difference did not reach

significance. These results mean high-high agentic self-esteem

counselors were significantly less likely than low-low and low-high

agentic self-esteem counselors to attribute their unsuccessful outcomes

to their immediate effort.


Table 17

Means and Standard Deviations for Immediate Effort
Attribution by Counselors' Levels of Agentic Self-Esteem



Causal Group N Mean SD
Attribution


Low-low 33 2.6666 1.290

Immediate Low-high 31 2.8064 1.352
Effort
High-low 34 2.0882 1.311

High-High 29 1.7241 1.161


The results in Table 23 (Appendix J) reveal that counselors did not

differ significantly in their use of the remaining eight causal

attributions when compared by their levels of agentic self-esteem.

Therefore, for the causal attribution immediate effort, hypothesis six

was rejected. For the remaining eight causal attributions investigated

in this study, hypothesis six was not rejected.


Additional Results

To investigate the relationship between type of experience reported

and subsequent causal attributions assigned to performance outcomes,

counselors were asked to classify their successful and unsuccessful








performance outcomes according to achievement domain. The results in

Table 18 reveal the largest percentage of counselors classified their

successful performance outcomes as interpersonal-affiliative events.

While this was also true for the unsuccessful performance outcomes, more

than twice as many counselors also classified their unsuccessful

outcomes as intrapersonal or mastery-control events.


Table 18

Classification of Counselors' Successful and Unsuccessful
Performance Outcomes by Achievement Domains



ACHIEVEMENT DOMAIN

Intrapersonal Interpersonal- Mastery-Control
PERFORMANCE OUTCOME Event Affiliative Event Event


Successful 9 108 10
Performance Outcome 7.90% 85.04% 7.87%

Unsuccessful 24 81 22
Performance Outcome 18.90% 63.78% 17.32%


The relationship of causal attributions to achievement domains was

investigated by examining the frequencies of responses to the nine

causal factors for successful and unsuccessful performance outcomes.

Results in Table 19 indicate for counselors' successful performance

outcomes, patterns emerged in the causal responsibility assigned to

their typical effort and mood at time of task when compared by

achievement domains.

Fifty percent of counselors reporting mastery-control events

indicated their typical effort was either very much or quite a cause of

their successful outcomes as compared to 88.8 percent of counselors

reporting intrapersonal events and 82.4 percent of counselors reporting








Table 19

Distribution of Responses to Causal Attributions for
Counselors' Successful Performance Outcomes by Achievement Domains

RESPONSES
Not Very
ACHIEVEMENT At All Somewhat A Quite Much a
DOMAINS a Cause a Cause Cause a Cause Cause

Ability
Intrapersonal 0% 0% 11.11% 22.22% 66.67%
Interpersonal 0% 0.93% 12.04% 37.04% 50.00%
Mastery/Control 0% 0% 0% 50.00% 50.00%

Typical Effort
Intrapersonal 0% 0% 11.11% 22.22% 66.67%
Interpersonal 1.85% 2.78% 12.96% 37.96% 44.44%
Mastery/Control 10.00% 0% 40.00% 30.00% 20.00%

Interest in Task
Intrapersonal 0% 0% 0% 44.00% 55.56%
Interpersonal 0% 0.93% 4.67% 31.78% 62.62%
Mastery/Control 0% 0% 10.00% 10.00% 80.00%

Immediate Effort
Intrapersonal 0% 11.11% 11.11% 55.56% 33.33%
Interpersonal 2.78% 11.11% 11.11% 24.07% 62.04%
Mastery/Control 0% 10.00% 10.00% 30.00% 60.00%

Mood at Time of Task
Intrapersonal 33.33% 11.11% 11.11% 33.33% 11.11%
Interpersonal 15.74% 12.04% 25.00% 23.15% 24.07%
Mastery/Control 10.00% 20.00% 0% 30.00% 40.00%

Supervisor
Intrapersonal 75.00% 0% 25.00% 0% 0%
Interpersonal 63.73% 17.65% 10.78% 5.88% 1.96%
Mastery/Control 66.67% 22.22% 11.11% 0% 0%

Good Luck
Intrapersonal 55.56% 33.33% 11.11% 0% 0%
Interpersonal 61.68% 21.50% 9.35% 2.80% 4.67%
Mastery/Control 50.00% 20.00% 20.00% 10.00% 0%

Task Ease
Intrapersonal 88.89% 0% 11.11% 0% 0%
Interpersonal 82.41% 12.96% 4.63% 0% 0%
Mastery/Control 70.00% 10.00% 20.00% 0% 0%

Help From Others
Intrapersonal 66.67% 11.11% 11.11% 0% 11.11%
Interpersonal 42.06% 18.69% 15.89% 14.02% 9.35%
Mastery/Control 40.00% 40.00% 0% 20.00% 0%





-71-


interpersonal events. Seventy percent of counselors reporting

mastery-control events indicated their mood at time of task was either

very much or quite a cause of their successful outcomes as compared to

44.4 percent and 47.2 percent of counselors reporting intrapersonal and

interpersonal events respectively.

Results in Table 20 indicate for Counselors' unsuccessful

performance outcomes patterns emerged in the causal responsibility

assigned to their inability, typical effort, interest in task, immediate

effort, and mood at time of task when compared by achievement domains.

Over thirty-six percent of counselors reporting mastery-control events

indicated their inability was either very much or quite a cause of their

unsuccessful outcomes as compared to 12.5 percent 17.5 percent of

counselors reporting intrapersonal and interpersonal events

respectively.

Twenty-five percent of counselors reporting intrapersonal events

indicated their typical effort was either very much or quite a cause of

their unsuccessful performance outcomes as compared to 11.2 percent and

9.0 percent of counselors reporting interpersonal and mastery-control

events respectively. Over four percent of counselors reporting

mastery-control events indicated their immediate effort was either very

much or quite a cause of their unsuccessful outcomes as compared to 29.1

percent and 22.2 percent of counselors reporting intrapersonal and

interpersonal events respectively. Over four percent of counselors

reporting mastery-control events indicated their mood at time of task

was either very much or quite a cause of their successful outcomes as

compared to 54.1 percent and 19.7 percent of counselors reporting

intrapersonal and interpersonal events respectively.








Table 20

Distribution of Responses to Causal Attributions for
Counselors' Unsuccessful Performance Outcomes by Achievement Domains

RESPONSES
Not Very
ACHIEVEMENT At All Somewhat A Quite Much a
DOMAINS a Cause a Cause Cause a Cause Cause

Inability
Intrapersonal 25.00% 37.50% 25.00% 8.33% 4.17%
Interpersonal 26.25% 35.00% 21.25% 11.25% 6.25%
Mastery/Control 18.18% 27.27% 18.18% 18.18% 18.18%

Typical Effort
Intrapersonal 33.33% 25.00% 16.67% 20.83% 4.17%
Interpersonal 53.75% 22.50% 12.50% 8.75% 2.50%
Mastery/Control 59.09% 9.09% 27.27% 0% 4.55%

Interest in Task
Intrapersonal 33.33% 33.33% 8.33% 0% 25.00
Interpersonal 61.73% 16.05% 11.11% 7.41% 3.70%
Mastery/Control 59.09% 13.64% 18.18% 0% 9.09%

Immediate Effort
Intrapersonal 29.17% 16.67% 25.00% 12.50% 16.67%
Interpersonal 39.51% 25.93% 12.35% 11.11% 11.11%
Mastery/Control 36.36% 22.73% 36.36% 4.55% 0%

Mood at Time of Task
Intrapersonal 12.50% 25.00% 8.33% 16.67% 37.50%
Interpersonal 43.21% 19.75% 17.28% 8.64% 11.11%
Mastery/Control 63.64% 18.18% 13.64% 0% 4.55%

Supervisor
Intrapersonal 81.82% 9.09% 0% 9.09% 0%
Interpersonal 89.61% 3.90% 3.90% 0% 2.60%
Mastery/Control 77.27% 9.09% 9.09% 0% 4.55%

Bad Luck
Intrapersonal 87.50% 8.33% 4.17% 0% 0%
Interpersonal 75.31% 16.05% 3.70% 2.47% 2.47%
Mastery/Control 77.27% 9.09% 4.55% 9.09% 0%

Task Difficulty
Intrapersonal 16.67% 16.67% 12.50% 33.33% 20.83%
Interpersonal 12.35% 8.64% 13.58% 27.16% 38.27%
Mastery/Control 18.18% 13.64% 9.09% 36.36% 22.73%

Lack of Help From Others
Intrapersonal 58.33% 8.33% 25.00% 4.17% 4.17%
Interpersonal 48.75% 12.50% 10.00% 11.25% 17.50%
Mastery/Control 40.91% 22.73% 13.64% 9.09% 13.64%





-73-


Discussion


The results of this study indicate male and female counselors do

not attribute the causes of their successful performance outcomes to

different factors. The findings support the conclusions of McHugh,

Frieze, and Hanusa (1982) that widely held beliefs concerning sex

differences in attributions are unwarranted. This finding also

supports the contention that attribution research should expand beyond

investigating sex differences to examining individual differences

within male and female groups.

Significant differences in counselors' causal attributions for

successful performance outcomes were found when compared by their sex

role identities. Undifferentiated sex-typed counselors attributed their

successful outcomes to their typical effort significantly less than

androgynous or masculine sex-typed counselors. Weiner's (1979) three

dimensional taxonomy of causes classifies typical effort as an internal,

stable, and controllable factor. Attribution theory contends internal

factors are associated with high self-worth, stable factors with future

expectancies for success, and controllable factors with appropriate

interpersonal judgements.

It is not surprising that undifferentiated sex-typed counselors

attributed their successful performance outcomes to their typical effort

less than counselors in the other sex-typed groups. An undifferentiated

sex-type classification results from a low endorsement of both the

masculine and feminine characteristics contained in the Bem Sex Role

Inventory. Counselors unwilling to rate themselves high on either

masculine or feminine attributes would not be expected to assign








successful outcomes to an attribution that indicates acceptance of

personal responsibility for achievement.

Significant differences in counselors' causal attributions for

successful performance outcomes were found when compared by their

levels of agentic self-esteem. Self-consistency theory purports

individuals are motivated to interpret achievement events in ways that

allow them to maintain consistent self-beliefs. High self-esteem

persons are expected to attribute success to causal factors that

reinforce their positive self-image. Low self-esteem persons are

expected to exhibit reverse patterns of causality.

Results of this study indicate counselors with lower agentic

self-esteem ascribed their successful outcomes less to both ability and

typical effort than counselors with higher agentic self-esteem. Since

the assignment of causality to these factors indicates an acceptance of

personal responsibility for success, these findings are congruent with

self-consistency theory. However, self-consistency theory also

predicts low self-esteem persons will assign causality for their

successful outcomes to factors indicating low acceptance of personal

responsibility such as task ease or good luck. Results of this study

did not support this second prediction. Counselors did not differ in

their use of the remaining seven factors investigated in this study

when compared by their levels of agentic self-esteem.

Traditionally, attribution research has focused on investigating

the four causal factors originally identified by Weiner et al (1971):

ability, effort, luck, and task difficulty. In recent investigations

less conventional researchers employing open-ended response

methodologies have concluded the assignment of causal responsibility is








not limited to these four factors alone. However, the findings of this

study indicate the original intuitions of Weiner and his colleagues may

be correct. While a potentially unlimited number of factors may exist

that can explain success, this study revealed typical effort and

ability are the most salient and generally recognized of causes.

Male and female counselors did not attribute the causes of their

unsuccessful performance outcomes to different factors. This finding

is congruent with the absence of sex differences found in attributions

for successful performance outcomes and further substantiates the

McHugh, Frieze, and Hanusa (1982) conclusion that sex differences in

achievement behaviors will not be explained through identifying sex

differences in causal attributions.

The expectation that attributional patterns would be distinguished

by sex role identities for unsuccessful performance outcomes was not

proven. Androgynous, masculine, feminine, and undifferentiated

sex-typed counselors did not attribute the causes of their unsuccessful

performance outcomes to different factors. This finding is puzzling in

view of the significant differences found in counselors' attributions

for successful performance outcomes when compared by their sex role

identities.

Several authors have suggested that problems exit with the

assessment of causality for unsuccessful outcomes. Previous

investigations have revealed differences in causal ascriptions for

successful outcomes that did not find significant differences in

attributions for unsuccessful outcomes (Teglasi, 1978; Travis et al.,

1982). One explanation for this could be that the social stigma

attached to unsuccessful performances may inhibit the self-evaluations








that occur for successful outcomes. Therefore, no patterns of causality

emerge.

Significant differences in counselors causal attributions for

unsuccessful performance outcomes were found when compared by their

levels of agentic self-esteem. High agentic self-esteem counselors

ascribed their unsuccessful outcomes significantly less to their

immediate effort than counselors with lower levels of agentic

self-esteem. Once again, these findings lend partial support to

self-consistency theory.

Self-consistency theory contends to maintain congruent

self-beliefs, high self-esteem persons will ascribe unsuccessful

performance outcomes to external sources df causality and low

self-esteem persons will ascribe unsuccessful outcomes to internal

sources of causality. The results of this study revealed high agentic

self-esteem counselors distinguished themselves from counselors in the

lower half of the agentic self-esteem distribution by their

unwillingness to identify their immediate effort, an internal

attribution, as responsible for their lack of success. While these

high self-esteem counselors did not assign responsibility to external

sources any more than lower self-esteem counselors, they did indicate

their beliefs that an internal factor was not responsible for their

unsuccessful outcomes.

From these results it is difficult to determine just how counselors

explained their unsuccessful performance outcomes. The only pattern

that clearly emerges is that higher agentic self-esteem counselors were

more certain than lower agentic self-esteem counselors their immediate

effort was not responsible for their unsuccessful outcomes. However,








this finding provides little information on how counselors with any

level of agentic self-esteem explain their unsuccessful performance

outcomes. From these findings it is reasonable to conject that

methodologies revealing significant differences in causal attributions

for successful performance outcomes may be invalid for asssessing

causality for unsuccessful performance outcomes.

Assessment of achievement domain classifications revealed the

largest percentage of counselors classified their successful

performance outcomes as interpersonal-affiliative events. In fact, the

distribution was so heavily weighted in this category that it is

difficult to reach meaningful conclusions concerning the relationship

between achievement domains and causal attributions. One pattern that

does emerge is that counselors reporting mastery-control events

evidenced a stronger tendency toward externality than counselors

reporting intrapersonal or interpersonal-affiliative events.

Counselors reporting mastery-control events were less likely to

attribute their success to their typical effort and more likely to

attribute causal responsibility to their mood at the time of the task

than counselors in the other two domains. This may be due to the same

reasons the majority of counselors classified their successful

performance outcomes as interpersonal-affiliative events.

The counseling profession is dedicated to the enhancement of the

potential of each individual and thus to the service of society (APGA,

1981). This dedication is reflected in counselors overwhelming

classification of their successful performance outcomes as

interpersonal-affiliative events and subsequent attributions to factors

indicating an acceptance of personal responsibility. Perhaps





-78-


counselors reporting successful outcomes in the mastery-control

classification indicated less acceptance of personal responsibility

because these experiences were unusual for them.

As was true for successful performance outcomes, the largest

percentage of counselors classified their unsuccessful performance

outcomes as interpersonal-affiliative events. Counselors reporting

mastery-control events distinguished themselves by a strong tendency to

attribute causal responsibility for their lack of success to their

inability. Once again it seems counselors reporting mastery-control

events appear less secure than counselors reporting experiences in the

other two achievement domains.














CHAPTER FIVE

CONCLUSIONS, IMPLICATIONS, SUMMARY, AND RECOMMENDATIONS


Conclusions

Based on the results of this study, the following conclusions were

reached:

1. Differences do not exist in counselors' causal attributions for

successful performance outcomes when compared by their sex. Male

and female counselors do not attribute the causes of their

successful outcomes to different factors.

2. Differences do exist in counselors' causal attributions for

successful performance outcomes when compared by their sex role

identities. Undifferentiated sex-typed counselors were less

likely than the androgynous and masculine sex-typed counselors to

attribute their successful outcomes to their typical effort.

3. Differences do exist in counselors' causal attributions for

successful performance outcomes when compared by their levels of

agentic self-esteem. Counselors with lower levels of agentic

self-esteem were less likely than the higher level agentic

self-esteem counselors to attribute their successful outcomes to

their typical effort and ability.

4. Differences do not exist in counselors' causal attributions for

unsuccessful performance outcomes when compared by their sex.

Male and female counselors do not attribute the causes of their

unsuccessful outcomes to different factors.

-79-








5. Differences do not exist in counselors' causal attributions for

unsuccessful performance outcomes when compared by their sex role

identities. Androgynous, masculine, feminine, and

undifferentiated sex-typed counselors do not attribute the causes

of their unsuccessful outcomes to different factors.

6. Differences do exist in counselors' causal attributions for

unsuccessful performance outcomes when compared by their levels of

agentic self-esteem. Counselors with higher levels of agentic

self-esteem were less likely than the lower level agentic

self-esteem counselors to attribute their unsuccessful outcomes to

their immediate effort.


Implications

The results of this study have important implications for future

directions in attribution research. Researchers should begin to

examine individual differences and the effects of various personality

variables on causal attributions. In this study, differences in levels

of agentic self-esteem proved to be significant in identifying

attributional patterns. Researchers should conduct investigations that

assess the relationship between causal attributions and personality

factors not yet considered.

A second implication of this study is that researchers should

consider the effects of situational variables on causal attributions to

distinguish between situations that elicit differential attributional

responses. Attributions made in one situation may not be

representative of an enduring disposition to make the same type of

attributions in other situations. The indication of differences in

associations between achievement domains and causal attributions








implies that characteristics of the performance outcome reported may be

more influential in determining attributions than characteristics of

the individual. Attribution research has not yet adequately addressed

this possibility.

A third implication of this study is that methodologies employed

to assess causal attributions for successful performance outcomes may

be invalid for assessing causal attributions for unsuccessful

performance outcomes. Researchers should devise studies that

investigate whether or not the same process used to assign causality

for successful outcomes is applied to unsuccessful outcomes. In this

study clearer attributional patterns emerged for successful

performances than for unsuccessful ones. This could be due to

differences in perceived causality of successful and unsuccessful

experiences or because the research procedure employed did not

adequately assess both types of outcomes.

A final implication of this study is that mental health

professionals should be aware of the relationship between causal

attributions and sex role identities and levels of agentic self-esteem.

Assignment of causality to a particular attribution indicates

individuals' perceptions of responsibility for their performance

outcomes. Mental health professionals who recognize the tendency of

undifferentiated sex-typed and low self-esteem clients to discount

personal characteristics as the cause of their successes, may be better

equipped to facilitate growth in these clients. Encouraging acceptance

of personal control over life experiences would be particularly helpful

with clients with histories of underachievement.








Summary

The purpose of this study was to expand attribution research from

examining sex differences to investigating individual differences in

causal attributions. This study investigated the effects of counselors'

sex, sex role identities, and levels of agentic self-esteem on their

causal attributions for successful and unsuccessful performance

outcomes. The statement of the problem, purpose of the study, need for

the study, significance of the study, definition of terms, and

organization of the study were presented in Chapter One.

Chapter Two reviewed the literature related to attribution theory,

sex roles, self-esteem, and traditional and nontraditional career

women. The sections included under attribution theory were theoretical

background, theoretical models, studies revealing sex differences in

causal attributions, studies revealing no sex differences in causal

attributions, and future directions for attribution research. The

sections included under self-esteem were theoretical background,

self-systems, and multidimensional theories of self-esteem.

Chapter Three described the hypotheses, population and sample,

instruments, procedures, analyses of data, and the limitations of the

study. Chapter Four presented the results and a discussion of these

results. The findings of this study indicate differences do not exist

in counselors' causal attributions for successful and unsuccessful

performance outcomes when compared by their sex. Differences in

counselors' causal attributions emerged for their successful

performance outcomes but not for their unsuccessful performance

outcomes when compared by their sex role identities. Differences in

counselors' causal attributions emerged for both their successful and








unsuccessful performance outcomes when compared by their levels of

agentic self-esteem.


Recommendations for Further Research

Based on the results of this study, the following research studies

are suggested :

1. A replication study of this research should be conducted to

examine the reliability of these findings. Replication should be

conducted under conditions approximately equivalent to this study.

Replication studies that vary the occupational group could provide

useful information as to whether the differences in causal

attributions found in this study are unique to counselors or

whether they are generalizable in scope.

2. A correlational study should be conducted to investigate the

relationship between attributions for performance outcomes and

differential achievement-related behaviors. The relationship between

achievement and causal attributions has not been adequately

investigated.

3. A study should be conducted to investigate the differential

consequence of attributions for men and women. It is possible

that differences exist in the consequences men and women

experience when assigning a particular attribution to explain

performance. If differences do exist they may provide insights

into why sex differences in achievement behaviors continue to

occur.

4. A study should be conducted to investigate whether or not persons

assign causal responsibility for their successful performance

outcomes in the same manner that they assign causal responsibility

for their unsuccessful performance outcomes.














APPENDIX A

BEM SEX-ROLE INVENTORY


INSTRUCTIONS: On the opposite side of this sheet, you will find listed

a number of personality characteristics. We would like you to use those

characteristics to describe yourself, that is, we would like you to


indicate,, on a scal

characteristics is.

EXAMPLE:

Write a 1 if it

Write a 2 if it

Write a 3 if it

Write a 4 if it

Write a 5 if it

Write a 6 if it

Write a 7 if it


e from 1 to 7, how true of you each of these

Please do not leave any characteristic unmarked.


NEVER or ALMOST NEVER TRUE.

USUALLY NOT TRUE.

SOMETIMES BUT INFREQUENTLY TRUE.

OCCASIONALLY TRUE.

OFTEN TRUE.

USUALLY TRUE.

ALWAYS OR ALMOST ALWAYS TRUE.


MARK ALL ITEMS!


-84-








1 2 3 4 5 6 7


Never or Usually Sometimes but Occasionally Often Usually Always or
almost not infrequently true true true almost
never true true true always true


Defend my
own beliefs


Affectionate


Conscientious


Independent


Sympathetic


Moody


Assertive


Sensitive to
needs of others


Reliable


Strong



Understanding


Jealous


Forceful


Compassionate


Adaptable


Dominant


Tender


Conceited


Willing to


Love children


Tactful


Aggressive



Gentle


Conventional



Self-reliant


Yielding


Helpful


Athletic


Flatterable


Theatrical


Self-sufficient


Loyal


Happy


Individualistic


Soft-spoken


Unpredictable



Masculine


Gullible



Solemn


Competitive


Childlike


Likable








4 5 6 7
I I I I


Never or Usually Sometimes but Occasionally Often Usually Always or
almost not infrequently true true true almost


never true true



Truthful


Have leadership
abilities


Eager to soothe
hurt feelings


Secretive


Willing to
take risks



Warm


Cheerful


Unsystematic


Analytical


Shy



Inefficient


Make
decisions easily
I


always true


Ambitious


Do not use
harsh language



Sincere


Act as a leader



Feminine



Friendly














APPENDIX B

THE PERFORMANCE SELF-ESTEEM SCALE


On the opposite side of this sheet is a list of descriptions about

people. For each, please indicate how true the description is of you in

this way:

Mark a 1 if it is NEVER OR ALMOST NEVER TRUE of you.

Mark a 2 if it is USUALLY NOT TRUE of you.

Mark a 3 if it is SOMETIMES BUT INFREQUENTLY TRUE of you.

Mark a 4 if it is OCCASIONALLY TRUE of you.

Mark a 5 if it is OFTEN TRUE of you.

Mark a 6 if it is USUALLY TRUE of you.

Mark a 7 if it is ALWAYS OR ALMOST ALWAYS TRUE of you.



MARK ALL ITEMS!








4 5 6


Never or Usually Sometimes but Occasionally Often Usually Always or
almost not infrequently true true true almost
never true true true always true


Productive


Assertive


Friendly



Tough



Clever


Creative


Self critical


Able to
give orders



Nervous


Businesslike



Self-sufficient


Logical


Self reliant


Easily hurt



Good sense
of humor



Inefficient


Enjoys a
challenge


Pleasant


Persuasive



Has initiative


Willing to
take risks


Powerful



Acts as a leader



Intelligent


Pessimistic


Good business
sense



Individualistic


Willing to
take a stand


Makes a mistake
when flustered


Gullible


Sociable



Ambitious



Yielding


Fun to be with


Headed for
success


Avoids
competition




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs