' ”'3'. ‘:‘;“"‘?“-"_‘,-‘"*‘r. xx .-:»;v:;»."E',;r;5-;zfg,“ 4,, "M‘Jléfz‘u'i 5"}.3‘; -.o 1??”51’ _ 2 gafffii‘fi #04 ‘. ‘rq‘. ‘U'T . 3:: . :1 .‘g 4.13:53 A" ‘3 I}? 4“- , ' " "3'75“: $125.5: J?- M}. 317-”: :3"; ,. 1". 31‘ )U"‘U)J ‘ “Li; lighgr‘r' . “(V1 NW! JR“; :thqu ‘ M J. “'—.~.‘.‘~:‘ J},_‘ ‘ .521“. v M; ,. .4 a 'II“. ' ; 1. ». W». ”"9? 5,; 55:??? "o -v ”Jr?! :2» Li! as; " find. 7:; , ‘ 3&5!” P“ '3, 5." r. ...u. 2 . ‘ . a u Far}; , r» ”m a 7 Ifi%’§;,fi$§$ fing'J3 H’J ' 3&i1'JJ‘. ‘4’“4‘27’), "fi9 J. .1 ’«f ,1“ ' 9;”? ("“fii‘ if? {—3 v . :- . u'. uni»— Sac“).- '3 “-V‘ 0 ‘fi-r‘ q.“ . J A. ¢ .25., 1 $7419 J-J’}: If r'r‘? I I . ‘ "3"“ v 5.: J.;:,,13.;;;v,r$"g§.xm u} (J’LZ'Q 1'2"" 2”}; “:4; r_‘,’,‘r| ("of xi. fl; 5’" ' -n.,-_—:, 0 -'v-.4 . . I.‘ ~ whp. :; .‘ . .' ‘ «1:43-2:- . A _ . - 'I'G‘ § ".. 1' ”only“. ""4r5 7:17 J 4 2 ‘1 |L~ v 'f'J-“J a,- $4 -150“! ‘Llrjagqé; " um w‘ n. 1.. J; ~ ,4-3’.‘ - w‘u L~._.‘;_v .\J __; ”C’.‘=' 'vI‘ -—. —J J ”x”, . V.”V. . J . J. -~:?‘§:v:a—- "Mai-{9,}, 53""? «3.2? 2 -~ WWI “~ ,1 . zrééAZLJ-“o'gf .. ,, Jifu" 4?? . ”.4 ~L'vcf, W']"'-. 1’ .I.0)’ ,JVvyH“, "VIN-«NV 1 - W It'll "15/;7$. £34}: VFW: 731‘: 227/7223 QYATE UN! WERSITY LIBRA:ES“r 2222222222222222 222 22 222 22222 22 3 1293 00600 2350 LlEi’tAfiY Michigan State This is to certify that the dissertation entitled COPING STRATEGIES OF MARRIED WOMEN EMPLOYED IN MALE AND FEMALE-DOMINANT OCCUPATIONS presented by James Donald Lawson has been accepted towards fulfillment of the requirements for PhD degreein Counseling Psychology Major professor Date September 14, 1987 MSU is an Affirmative Action/Equal Opportunity Institution 042771 PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE m» — s .32 MSU I: An Affirmative Action/Equal Opportunity Institution COPING STRATEGIES OF MARRIED WOMEN EMPLOYED IN MALE AND FEMALE-DOMINANT OCCUPATIONS BY James Donald Lawson . $ J H, 1!: trw . 2‘ ‘Oll’3' A DISSERTATION 29 Submitted to 2? Michigan state University in partial fulfillment of the requirements , for the degree of DOCTOR OF PHILOSOPHY Tag; 9t.nea1th Education, Counseling Psychology, -v and Human Performance .,»;1. J u 8 V5r* “7 "‘ 1987 S ST37010— ABSTRACT COPING STRATEGIES OF MARRIED WOMEN EMPLOYED IN MALE AND FEMALE-DOMINANT OCCUPATIONS BY James Donald Lawson This study attempted to learn how married women differ in their use of coping strategies and in their perceptions of well-being as a consequence of employment in male or female- dominant careers. An amended version of Gray's (1980) survey was constructed that included the Personal Attributes Questionnaire, to assess masculinity and femininity, and the Texas Social Behavior Inventory, to assess social competence and social self-esteem. This study was carried out with samples of women in female-dominant fields, e.g., elementary teachers and registered nurses, and women in male-dominant fields, e.g., university faculty and physicians. A total of 219 usable surveys were received that included 148 surveys from married women. Married women employed in male-dominant occupations were higher in masculine traits than women employed in female- dominant occupations, while women employed in female-dominant _ occupations were higher in feminine traits than women employed in male-dominant occupations. Women in male or female-dominant occupational groups did not differ in social competence and social self-esteem. A common factor analysis identified three coping strategy factors underlying Hall's (1972) 16 coping strategies. These were named Balanced Role Redefinition, Direct Role Redefinition, and Planful Role Organization. Married women did not differ as a consequence of occupational status in their use of Balanced Role Redefinition. However, women employed in male-dominant occupations scored significantly higher than women in female-dominant occupations on Direct Role Redefinition, while women employed in female-dominant occupations scored significantly higher than women employed in male-dominant occupations in their use of Planful Role Organization. Women did not significantly differ as a consequence of occupational status in their perceived support from spouses, friends, colleagues and parents in this study. Women employed in male-dominant occupations experienced significantly higher levels of role strain than women employed in female-dominant occupations. Women did not differ as a consequence of male or female-dominant occupational status in satisfaction with coping strategies, or satisfaction with their careers, parental roles, or marriage roles. ACKNOWLEDGMENTS I wish to thank Gloria Kielbaso, whose friendship and encouragement over the past three years has been invaluable. I am also indebted to Rochelle Habeck, who gave freely of her time and energy to help me through the dissertation process, and provided many perceptive analyses of theoretical and methodological impasses. I also wish to thank Bill Hinds and Richard Johnson who provided thoughtful critiques and insights into this research. iii TABLE OF CONTENTS Page LIST OF TABLES........................................vi LIST OF APPENDICES....................................viii Chapter 1. INTRODUCTION.................................l Introduction...............................1 overview.‘000......IO...IIUOIIOOOOIIOOIOOOOS M . REVIEW OF THE LITERATURE.....................8 Role strain................................9 Coping.....................................10 Coping strategies..........................20 Traditional versus nontraditional occupations..............................3l Life satisfaction..........................33 Social competence..........................34 Summary....................................35 3. METHODOLOGY Sampling and data collection procedures....39 Elementary school teachers...............46 Registered nurses........................47 University faculty.......................47 Physicians...............................47 Data collection..........................47 Human subjects review committee..........49 Confidentiality..........................49 Limitations..............................50 Instruments................................50 Personal Attributes Questionnaire........53 Texas Social Behavior Inventory..........55 Hypotheses.................................56 Analysis of data...........................57 statistical procedures.....................58 Hypothesis 1.............................58 Hypothesis 2.............................59 Hypothesis 3.............................60 Page 4. RESULTS AND DISCUSSION.OIOCOICOOODCCOOUCOIOOI62 Demographic information....................64 Age......................................79 Years employed in occupation.............79 Combined salary..........................79 Hours worked.............................79 Length of marriage.......................79 Children in home.........................8o Age of youngest child....................80 Reliability................................80 Hypothesis 1...............................82 Hypothesis 2...............................83 Hypothesis 3...............................89 Post hoc analyses..........................99 Discussion.................................107 Post hoc analyses........................119 5. SUMMARY AND IMPLICATIONS.....................127 Summary....................................127 Conclusions................................135 Limitations of the study...................139 Implications...............................141 Therapy............. ..... ................143 Research... ..... .........................144 Social policy............................148 APPENDICESCCOso.aIcc.Ion...o...000.00.00.00000000153 REFERENCES....00Io...OIno.Io..-col-Ioocoocoooooollee LIST OF TABLES Table Page 1 Michigan State University Women Faculty Members Employed in Male-dominant fields........48 2 Distribution of Returns and Non-returns...........63 3 Age of Married Respondents........................65 Age of Married Respondents........................66 Years Worked in Present Occupation................67 Years Worked in Present Occupation................68 Annual Combined Salary of Wife and Husband........69 0‘!ka Annual Combined Salary of Wife and Husband........7o 9 Hours Worked in a Typical Week....................7l 10 Hours Worked in a Typical Week....................72 11 Length of Marriage to Present Spouse..............73 12 Length of Marriage to Present Spouse..............74 13 Number of Children Presently Living in Home.......75 14 Number of Children Presently Living in Home.......76 15 Age of Youngest Child in the Home.................77 16 Age of Youngest Child in the Home.................78 17 Means and Standard Deviations of Masculinity, Femininity and Texas Social Behavior Inventory Scales................................84 18 Rotated Factor Matrix.............................86 19 Means and standard Deviations of Coping Strategy Distributions for Male and Female-Dominant Groups......................90 vi Table Page 20 21 22 23 24 25 26 27 Couple's Decision Regarding Children..............92 Attitude Towards Child Care Arrangement...........93 Attitude Regarding Effects of Employment on ChildrenOOIOOI0......-IOOCOOIOOIOOOOOOIOIOOOOQ4 Attitude Towards Household Chores.. ..... ..........95 Attitude Toward Family, Home, and Career Roles....96 Coping Strategy Factors in Stepwise Multiple Regression Equations Predicting "Satisfaction With the Way You Have Dealt With Possible Role Strains in Your Life"......................102 Coping Strategy Factors in Stepwise Multiple Regression Equations Predicting Aspects of Role StrainIOIOOOOOOOOOI.I...OIOOIOOOIOIOOOI0000104 Demographic and Personality Variables in Stepwise Multiple Regression Equations Predicting Use of Coping Strategies.............106 vii LIST OF APPENDICES Appendix Page A Letter to Lansing School District Superintendent..........................153 QuestionnaireOIOOCU...OOIIOOOOIIIOODOOIOIOIS4 Personal Attributes Questionnaire.........159 Texas Social Behavior Inventroy...........160 Postcard to Accompany Surveys.............163 B C D E Letter to Accompany Survey................162 F G Unmarried Women's Role Questionnaire......l64 H First Follow-up Postcard..................166 I Second Follow-up Postcard.................167 viii CHAPTER I INTRODUCTION Eng EIQQIEE Married working women occupy the roles of wife, worker, and possibly, parent. Early formulations of the emotional costs and benefits associated with the multiple roles of working married women focused upon the strain inherent in attempting to meet multiple role obligations (Good, 1960). In contrast, a contemporary view suggests that women who engage in more roles are often likely to be the recipients of more "potential sources of self-esteem, stimulation, privileges, social status, and social identity" (Baruch, Biener, & Barnett, 1987, p. 134). However, Baruch et a1. (1987) caution that the "nature and quality of a woman's experiences within a role, not merely role occupancy per se, are critical to understanding the processes affecting her well-being" (p. 133). How is a woman's psychological experience of well-being within the roles of wife, worker, and parent determined? Baruch et a1. (1987) suggest satisfaction within roles is determined by a balance of masculine, agentic and feminine, communal opportunities either across or within roles. A balance of masculine and feminine opportunities for 1 functioning are necessary to optimize well-being and decrease the experience of stress, while an imbalance may increase stress (Baruch et a1., 1987). A more elegant formulation about the emotional consequences of coping with multiple roles is that individuals become vulnerable to the experience of stress as a consequence of an ongoing transaction between the person and the environment (Lazarus & Launier, 1978). "To think fruitfully of stress and coping, these must be seen as a special kind of transaction between a person of a particular sort (i.e., with plans, commitments, hidden agendas, and belief systems) and an environment with its own characteristics (e.g., demands, constraints, and resources)" (Lazarus & Launier, 1978, p. 320). Thus, individuals make choices at different points in their life cycles that determine the environmental areas of work, play and intimacy within which they function, and these choices are primarily a consequence of personality (Lazarus & Launier, 1978). Masculine instrumentality/agency and feminine expressiveness/communion have been described as two cohesive aspects of personality and behavioral roles. Instrumentality is associated with goal-directed roles outside the home that require independence, self-reliance and a cognitive instrumentality, while expressiveness is associated with caring for the physical and emotional needs of the family (Parsons & Bales, 1955). More recently, Bakan (1966) proposed that the core properties of femininity and masculinity are communion and agency, respectively, coexisting female and male principles which are evident in all living organisms. Agency is characterized as a concern for oneself and one's own goals, while communion is an identification with and concern for the welfare of others (Bakan, 1966). Recent research suggests that masculinity and femininity are potent predictors of occupations (Lemkau, 1984; 1983; 1979). High levels of masculine instrumentality within the individual are predictive of male-dominant occupational choice, while high levels of feminine expressiveness predict female-dominant career choice for both men and women (Lemkau, 1984; 1983; 1979). Current sex role inventories are viewed as measures of general tendencies to use masculine (instrumental) or feminine (expressive) coping strategies (Moreland, Gulanick, Montague, & Harren, 1978). Since occupational environments differ in their requirements for instrumental and expressive coping strategies, occupational success and satisfaction may be a consequence of the fit between the person's level of masculinity and femininity, and the requirements of the occupational environment for different levels of masculinity and femininity (Tinsley & Heesacker, 1984). Spence and Helmreich (1978) hold that masculinity and femininity should be viewed as internal response predispositions that are sensitive to situational determinants, and also as sources of self-esteem that provide different types of mastery experiences. Masculine instrumentality represents a cognitive instrumental mode of social adaptation, while feminine expressiveness represents a nurturant mode of social adaptation in both women and men (Spence & Helmreich, 1978). Coping is generally construed as an adjustment to unusual stress that requires the mobilization of effort (White, 1974). Coping is subsumed under the more general construct of adaptation, which refers to routine, modal patterns of adjusting to life demands (White, 1974). Thus, contrary to Moreland et. al.'s (1978) characterization of masculine instrumentality and feminine expressiveness as two general coping strategies, masculininity and femininity are more appropriately viewed as modal patterns of adjusting to the demands of life, and fit with White's (1974) notion of adaptation. Levels of masculinity and femininity in individuals may influence the choice of occupational environments, and serve as modal patterns of obtaining self-esteem, control and mastery in both occupational and home environments. When individuals are required to adjust to unusual stress, including the strain of conflicting role demands, individuals can be expected to choose coping strategies that are consistent with their perceptions of control, and are thus most efficacious, within the context of a given transaction between the person and environment (Folkman, 1984). Thus, masculinity and femininity may influence individuals' decisions to enter male or female-dominant occupations, and may also influence the choice of coping strategies that are used to deal with multiple role conflicts. EEEEQSE 2: £22 §£EQY The purpose of this study is to obtain a clearer understanding of the problems of and solutions to the role conflicts in a contemporary sample of married women who differ in that they are employed in either traditional, female-dominant occupations or in nontraditonal, male- dominant occupations. Although previous research (e.g., Gray, 1980) has documented the role strains and coping strategies of married women employed in male-dominant occupations, a need exists to extend this research to a) compare married women employed in female-dominant occupations to women employed in male-dominant occupations on the dimensions assessed by Gray (1980), i.e., career attitudes, role strains, coping strategies, and satisfaction: and b) to explore whether salient personality deminsions differ with female versus male-dominant career status. This research is important because evidence exists that suggests that women employed in female or male-dominant occupations may differ in personality traits and perceptions of role satisfaction that may make different coping strategies more efficacious. A clerarer understanding of the efficacy of coping strategies associated with female or male- . dominant career status may help clinicians conceptualize client behaviors and attitudes associated with role strain, and to subsequently recommend interventions that are most 13“ suited to women who differ in the environmental contexts and personality variables assessed in this study. mam Several research questions were formulated, including: 1) Do married women employed in female-dominant careers differ from women employed in male-dominant careers on salient personality traits, i.e., instrumentality, expressiveness, social competence and social self-esteem? 2) Do married women employed in female-dominant occupations differ in their use of coping behaviors from women employed in male-dominant occuaptions, e.g., are different coping stragegies utilized to balance work and family roles? 3) Do women employed in female-dominant careers differ from women employed in male-dominant careers in career attitudes, role strain, and satisfaction? Overview In Chapter Two, the relevant literature is reviewed under the following areas: coping, presented primarily from Lazarus and his associates' perspectives; coping strategies used to deal with role conflicts and strains; a comparison of the coping strategies presented by different theorists; variables that predict male or female-dominant occupational choice: variables contributing to life satisfaction; and, social competence, viewed as it relates to the concept of coping competence. Chapter Three describes the methodology of this study, including the sampling scheme, survey procedures, instrumentation, formal statement of hypotheses, and rationale for statistical procedures. In Chapter Four, the results and analyses of the findings are presented. A summary of the study, conclusions and implications for further research are presented in Chapter Five. CHAPTER II REVIEW OF THE LITERATURE In recent years both the popular and scientific literature has been the focus of an explosion of interest in the issues faced by women, and couples, who choose to combine marriage and work roles. The concern that has blossomed in this area is due to the fact that the number of women working in the past decade increased by fifty percent (U. S. Department of Labor, 1982). Professional women no longer view the homemaker and career role as mutually exclusive (Yogev, 1983), and are marrying in increasing numbers (Epstein, 1973). However, the proportion of divorced professional women, compared to divorced professional men and the general population, is disportionately high (Rosow & Rose, 1972; Epstein, 1973). It appears that the combination of work and family roles can provide potent sources of both self-esteem and stress, and the general question that has been, and continues to be asked is: How can women and couples effectively balance work and family roles? The early research in the 1960's in this area tended to be sociological and focused mainly upon changes in sex roles (Rapoport & Rapoport, 1980). The concept of the "dual-career family" emerged in 1969 (Rapoport & Rapoport, 1969), and signaled an attempt to understand the linkages 8 "u" - 'V-v— between men's and women's roles. Case studies and semi- structured interviews were the predominant research method utilized (Rapoport & Rapoport,1980). The constructs that emerged from this primarily inductive period of research included: 1) the early determinants of dual-career families; 2) the importance of social support to the married career woman; 3) sources of stress and gain to the couple; 4) marital satisfaction, conceived as a balance between stress and gain; and, 5) tension lines that demarcate the boundary beyond which non-traditional behavior becomes a threat to a marital relationship (Rapoport & Rapoport, 1980). Transitional research in the early 1970's used cross- sectional studies to inductively test hypotheses, and were followed by the current generation of research which uses diverse research methods, and reflects the emergence of specialized interests, including; academic, social policy, feminist and therapeutic concerns. These deductive studies tend to more precisely define constructs, are more focused in their hypotheses, and direct attention to the management of the strains in a dual-career family (Rapoport & Rapoport, 1980). 391; strain. Goode (1960) defined role strain as the difficulty subjectively experienced in meeting role demands. He noted that role strain is a normal experience that obtains from the obligations of different role relationships. Individuals are faced with the dilemma of allocating energies and skills within their role systems to reduce role strain to 10 a manageable proportion (Goode, 1960). The current focus upon management of role strain is reflected in the interest in the literature upon coping strategies used by married working women to obtain maximum satisfaction by effectively balancing the demands and optimizing the gains available in work and family roles (eg., Hall, 1972; Gray, 1983, 1980; Anderson-Kulman & Paludi, 1986; Suchet & Barling, 1986; Bunker; 1985; Blackwell, 1984; Levo, 1983; Edwards, 1983; Schaper, 1983; Harrison & Minor, 1982; Oates, 1981; Olds, 1980). Within this context, coping has generally refered to responses to external life-strains that serve to prevent, avoid, or influence emotional distress. Coping is therefore inseparable from the life- strains experienced by people and from their inner emotional life (Pearlin & Schooler, 1978). Coping. In general, definitions of the psychological construct of coping share the central theme of a "struggle with external and internal demands, conflicts, and distressing emotions" (Lazarus & Folkman, 1984, p 283). There is converging agreement that the construct of coping is subsumed under the more general concept of adaptation, which refers to routine, modal patterns of adjusting to life demands (White, 1974). Coping, however, involves adjusting to unusual stress, and requires the mobilization of effort (White, 1974). The effectiveness of coping must be gauged against the functions that coping serves (Lazarus & Folkman, 1984). White 11 (1974) maintains that coping serves: 1) to enable the organism to secure information from the environment: 2) to maintain the internal conditions necessary for action and for processing information; 3) and to maintain flexible freedom of movement. Pearlin and Schooler (1978) hold that coping serves three functions: 1) to change the situation out of which the experience of strain emerges; 2) controlling the meaning of experiences before they become stressful; 3) and controlling stress itself. Cohen and Lazarus (1979) maintain that coping fulfills five tasks: 1) reducing harmful environmental conditions and enhancing prospects of recovery; 2) tolerating or adjusting to negative events and realities; 3) maintaining a positive self-image; 4) maintaining emotional equilibrium; and 5) continuing satisfying relationships with others. A central distinction that can be made between coping functions is that coping behavior may be either "problem- focused" or "emotion-focused" (Cohen & Lazarus, 1979; Folkman & Lazarus, 1980). "Problem-focused" coping pertains to "the management or alteration of the person-environment relationship that is the source of stress," while "emotion- focused" coping refers to "the regulation of stressful emotions" (Folkman & Lazarus, 1980, p 223), which may include behavior and expression, physiological disturbance, subjective distress, or a combination of these three manifestations of emotions (Lazarus & Folkman, 1984). Most people commonly use a combination of both problem-focused and 12 emotion-focused coping to deal with stressful encounters, indicating that coping is both consistent, and also variable across situations (Folkman & Lazarus, 1980; Pearlin & Schooler, 1978). Lazarus and his associates' conceptualization of and measurement of problem-focused and emotion-focused coping has evolved with the development of the Ways of Coping Checklist. The 68-item Ways of Coping Checklist was initially used to obtain a self—report of coping thoughts and actions middle- aged women and men used in specific encounters that required coping responses (Folkman & Lazarus, 1980). Factor analytic and correlational analyses lent support to the position that coping consisted of two primary dimensions that were represented within the inventory, e.g., problem-focused and emotion-focused coping. The Ways of Coping Checklist was later modified by deleting items, rewording unclear items, and adding items to obtain a 66-item inventory that assessed the individual's self-perception of coping responses on a 4-point Likert scale (Folkman & Lazarus, 1985). Factor-analytic evidence suggested that problem-focused and emotion-focused coping were represented by eight subscales in the revised Ways of Coping Checklist. Problem-focused coping was represented by a single subscale. Emotion-focused coping was represented by five subscales, including; wishful thinking, distancing, emphasizing the positive, self-blame, tension-reduction, and self-isolation. The eighth subscale, seeking social support, 13 consisted of a combination of problem-focused and emotion- focused coping strategies. The Ways of Coping Checklist was again factor-analyzed, using coping responses of 85 married couples that were collected on five separate occassions at six-month intervals. On the basis of three separate factor analyses, Folkman, Lazarus, Dunkel-Schetter, DeLongis and Gruen (1986) concluded that the Ways of Coping Checklist assessed eight dimensions, including: 1) Confrontive coping. This coping response includes aggressive efforts to change the situation that include a degree of hostility and risk-taking. 2) Distancing. Distancing describes efforts to detach oneself and create a positive outlook. 3) Self-control. Self-control includes efforts to control one's feelings and actions. 4) Seeking social support. This includes efforts to seek both informational and emotional support. 5) Accepting responsibility. This subscale assessed an acknowledgement of one's role in a problem and an effort to rectify the situation. 6) Escape-Avoidance. Escape-avoidance described both wishful thinking and behavioral efforts to escape or avoid stressful situations. 7) Planful problem-solving. This subscale described deliberate problem-focused efforts to change the situation that incorporated an analytic problem-solving approach. A l4 8) Positive reappraisal. This described efforts to extract positive meaning from the situation that could contribute to personal growth. A correlational analysis of the eight subscales indicated that the strongest relationships existed between positive reappraisal and planful problem-solving (r=.39), positive reappraisal and self-controlling (r=.39), and between planful problem-solving and self-controlling (r=.37), suggesting that these three scales share a common coping theme. A recent study that compared disabled athletes who planned to persist in sport with those who did not plan to persist found that those planning to persist used planful problem-solving, positive reappraisal and distancing more often that those not planning to persist (Habeck, Ewing, Overton, 8 Drummer, 1986), suggesting that persistance may partially explain the shared variance between planful problem-solving and positive reappraisal. From Lazarus and his associates' perspective (eg., Lazarus 8 Launier, 1978), coping processes are determined by an interaction of personality and situational factors, and thus involve transactions that are mediated by cognitive appraisals. The three main stress-relevant perceptions are harm-loss, threat, and challenge, which refer to a "balance of forces such that environmental demands tax or exceed the resources of the person" (Lazarus 8 Lanunier, 1978, p 288), and require coping effort. Within this context, the form of coping chosen is influenced by primary appraisals (what is at 15 stake in the encounter), and secondary appraisals (coping options). Folkman et. a1. (1986) assessed the relationship of coping responses in 85 married couples to primary appraisals that included: concern for one's own physical well-being, self-esteem, a goal at work, financial strain, losing respect for another person, and concern for a loved one's well-being. Secondary appraisals included appraisals that the situation; could change or be changed, had to be accepted, was one in which more information was needed before acting, or was one in which one had to hold back from doing what one wanted to do. If threat to one's self-esteem was high, individuals were more likely to use confrontive coping, self-control, accepting responsibility and escape avoidance, but were less likely to seek social support. When a loved one's well-being was at stake, individuals were most likely to use confrontive and escape-avoidance strategies, and less likely to use planful problem-solving and distancing. In stressful encounters involving a goal at work, individuals were more likely to use self-control and planful problem-solving (Folkman et. a1., 1986). Confrontive coping and self-control were likely to be used when loss of respect for someone else was threatened. Confrontive coping and seeking social support were predominant responses to a strain on financial resources, while threats to one's own physical health elicited seeking of social support and escape-avoidance. 16 Secondary appraisal of coping options influenced coping in the following ways. Individuals were more likely to accept responsibility, use confrontive coping, planful problem- solving, and positive reappraisal if they appraised the encounter as changeable. Subjects were more likely to use distancing and escape-avoidance in encounters when the encounter was judged as having to be accepted. When the stressful situation was judged as requiring more information before action could be taken, individuals sought more social support, and used planful problem-solving and self-control. Higher levels of confrontive coping, self-control, and escape-avoidance were used when subjects appraised the situation as requiring that they hold back from doing what they wanted to do. Situational contexts exert a strong influence in determining whether the focus of coping efforts are primarily problem-focused, or emotion-focused (Lazarus 8 Folkman, 1984). Work contexts require more problem-focused coping, while health contexts elicit emotion-focused coping (Folkman 8 Lazarus, 1980). Men use more problem-focused coping than women in work contexts; however, there are no gender differences in emotion-focused coping (Folkman 8 Lazarus, 1980). In general, individuals use problem-focused coping in encounters appraised as changeable, and emotion-focused coping in encounters appraised as unchangeable (Folkman 8 Lazarus, 1980). For example, students use more problem— focused coping during periods of anticipation and preparation fiil 17 for course examinations, than during the period after the exam while waiting to learn about the outcome (Folkman 8 Lazarus, 1985). The possible alternation between problem- focused and emotion-focused coping in behavioral sequences is demonstrated with a sample of married couples who used both confrontive coping and escape-avoidance in encounters that involved self-esteem and concern for a loved one's well being (Folkman et a1., 1986). Folkman et al. (1986) suggest that while engaged in stressful encounters that involve these stakes, individuals may alternate coping styles in a pattern of engagement, disengagement, and reengagement. This interpretation is consistent with the position that coping changes over the course of time, and is therefore a process, rather than a singular event (Folkman 8 Lazarus, 1980). Coping as process is illustrated by the grieving process, which changes from the perception of loss, to its resolution (Lazarus 8 Folkman, 1984). Factors that help to determine the choice of coping activity include both person factors and situational requirements, constraints and available resources (Lazarus 8 Folkman, 1984). In general the more ambiguous the situation, the more inference is required in order to make an appraisal of the situation, and the more influence person factors have in determining the meaning of the situation (Lazarus 8 Folkman, 1984). Although the role of social support in buffering the individual against the adverse effects of stress is well 18 documented (eg., Gentry 8 Kobasa, 1984), it is unclear to what extent social support should be considered a situation or person variable. Bloom (1979), for example, maintained that the coping efficacy of a group of mastectomy patients was determined by social support and information that increased their sense of control and reduced feelings of external control. Lazarus and Folkman (1984) note that the ability to maintain effective social supports may be considered a coping skill. Person factors, or personality traits, that influence appraisal and coping processes, include, but ere not limited to: motivational patterns, including the Type A behavior pattern; belief systems including self-esteem and locus of control; cognitive styles: ego development; intelligence and knowledge; and social competence (Lazarus 8 Folkman, 1984). In general, when faced with health problems, individuals with a more internal locus of control engage in more adaptive forms of coping than do externally focused people. However, it seems likely that under some circumstances, a perception of internal control may be harmful. For example, Type A individuals may engage in an exaggerated struggle to control encounters and suffer damage to their cardiovascular systems as a result of their struggles (Lazarus 8 Folkman, 1984). The relationship between coping and both physiological and psychological outcomes was recently examined. The relationship between somatic health status, psychological symptoms, and personality factors (mastery and interpersonal 19 trust), primary appraisal, secondary appraisal and eight forms of coping measured by the Ways of Coping Checklist were assessed in 150 adults (Folkman, Lazarus, Gruen, 8 DeLongis, 1986). When somatic health status was entered as the dependent variable, and personality factors, primary appraisal, secondary appraisal, and coping were entered as independent variables in multiple regression analyses, the predictor variables failed to account for a significant ' percentage of variance. However, when psychological symptoms were treated as the dependent variable in multiple regression analyses, the predictor variables accounted for 43 percent of the variance. Specifically, mastery, interpersonal trust and concern for a loved one's well—being were negatively associated with psychological symptoms, while confrontive coping, concern about financial security, and concern about one's own physical well-being were positively associated with symptoms. Wrubel, Benner, and Lazarus (1981) view the ability to maintain supportive social relationships as central to most coping activity. Social competence overlaps the concept of coping competence, the presence of generalized skills and resources for coping, which are inferred by summating outcomes of stressful encounters over a period of time. Social competence, therefore, is also "a summation of how well the person functions (social effectiveness) in a variety of social settings" (Wrubel, et a1., 1981, p 70). Social competence is defined as a "high level of coping competence, 20 and indeed, the ability to deal with others with minimal levels of stress may be one of its criteria" (Wrubel, et a1., 1981, p 71). Coping strategies, Although several studies identified methods used by married working women to deal with role conflicts and strains, Hall (1972) was the first researcher to attempt to evaluate the effectiveness of various coping strategies. Hall (1972) noted that women's roles operate simultaneously rather than sequentially, which may increase the demands encountered and the conflict experienced to a greater degree for women than for men. Thus, the primary problem confronting employed women is the interrole conflict that arises from competing demands among multiple roles (Hall, 1972). Hall (1972) identified 16 coping strategies that he categorized into three types of coping: structural role redefinition (Type I); personal role redefinition (Type II); and, reactive role behavior (Type III). These are conceptually aligned with Levinson's (1959) notion. According to Levinson (1959), roles may be viewed from three different perspectives; 1) structurally given demands, which include norms, expectations, taboos, and responsibilies; 2) personal role conceptions; and 3) role behavior. Structural role redefinition (Type I), reduces conflict between roles by changing external demands on one's behavior in.a specific role. Structural role redefinition involves actively negotiating with the 21 environmental transmitters of structurally imposed demands, and arriving at an agreement with the role sender on a new set of expectations. Items identified by Hall (1972) as demonstrating Type I coping included: eliminating activities within roles; hiring outside help to assist with chores; working to change societal definitions of women's roles; family members sharing household tasks and assisting in resolving conflicts between roles; and overlapping roles whenever possible. Personal role redefinition (Type II), involves a change in expectation of role behaviors by the individual. since personal role redefinition does not include a simultaneous alteration in the attitudes and expectations of role senders, Type II coping was perceived as a less permanent solution to role conflicts than Type I coping. Type II coping strategies included: establishing rules and priorities for dealing with roles; reducing personal standards within certain roles; developing new attitudes to reduce conflicts; keeping roles totally separate; rotating attention from home to children to work, depending on which need is most pressing; considering personal interests very important; and eliminating entire roles. Reactive role behavior (Type III), was identified by Hall (1972) as the least satisfactory solution to role conflicts. Reactive role behavior is demonstrated when the person attempts to cope with interrole conflict by 22 working harder in all roles. Type III coping was perceived as involving the greatest amount of strain and effecting no permanent relief since it does not involve the alteration of either internal or external expectations. Type III coping strategies included: scheduling and organizing activities carefully to optimize output; attempting to meet all role demands by doing everything people expect; and having no conscious strategy for dealing with conflicts between roles. In general, Type I coping, structural role redefinition, was perceived as a more stable and permanent resolution of interrole conflicts than Type II and Type III coping strategies. Hall's (1972) conceptualization of coping strategies was initially promising, since the 16 strategies separately defining the three types of coping were derived from Levinson's (1959) role theory. However, subsequent analysis with Hall's (1972) model has yielded inconsistent results. Therefore, the research conducted with Hall's (1972) model of coping strategies will be reviewed in some detail. Hall (1972) obtained his 16 coping strategies from an open-ended questionnaire on roles and role conflicts completed by 250 college-educated women in a pilot study in New Haven, Connecticut. He later assessed the relationship between the 16 coping strategies derived from the pilot study and satisfaction, by surveying 450 women who had graduated from the University of Connecticut. Of this sample, 170 23 married women completed the survey. Satisfaction with coping strategies, measured on a scale that ranged from dissatisfied to extremely satisfied, approached significance with Type I coping, while a combination of engaging in Type I and not engaging in Type III coping was significantly related to satisfaction. Gray (1980) noted that Hall's research suffered from serious methodological problems, since no demographic data other than year of graduation was collected, and only one-third of the women in the sample were employed full-time. It appeared to Gray (1980) that many of the 16 strategies identified by Hall (1972) could fall into more than one category. Accordingly, Gray (1980) designed a survey and data analysis procedure to determine: 1) the role strains and conflicts reported by women; 2) whether the distinct catagories of strategies for dealing with role conflicts identified by Hall (1972) can be identified by factor analysis; 3) whether the variables of profession, age, salary, and children are related to the ways in which women deal with role conflicts; and 4), the coping strategies that are linked with the reported satisfaction with roles. Gray's (1980) survey was completed by 232 married professional women, including 71 medical doctors, 85 lawyers, and 76 professors. The results of her analysis indicated that 77 percent of the women often experienced strains between home and career roles. Gray (1980) subjected a total of 47 items in the survey to a principal components factor 24 analysis with a subsequent rotation to a varimax criterion. Of these 47 items, 16 were the coping strategies identified by Hall (1972). Although 14 factors were extracted in this analysis, none of the factors were clearly defined by the strategies identified by Hall (1972) as representative of Type I, II, or III coping. strategies most associated with satisfaction were: having family members share household tasks (Type I); reducing standards within certain roles (Type II); scheduling and organizing activities carefully (Type III); having family members help resolve role conflicts (Type I); and considering personal interests important (Type II). Strategies that were negatively related to satisfaction were: eliminating activities within certain roles (Type I); keeping roles separate (Type II); attempting to meet the expectations of all (Type III); overlapping roles (Type I); and not having any conscious strategies for dealing with role conflicts (Type III). Hall's coping strategy model has also been tested by other researchers with samples and research methods that differed substantively from that used by Hall (1972) and Gray (1980). For example, Harrison and Minor (1982) administered surveys with an open-ended response format to 50 black women, of whom 13 were nonprofessional single working mothers and 27 were nonprofessional married working mothers. The responses to a question designed to elicit coping strategies for dealing with conflicts between work and family roles were 25 categorized by raters according to Hall's (1972) model of three types of coping strategies. The results of this survey indicated that marital status interacts with coping strategies to predict satisfaction with the mother role and work role. Married employed mothers tended to choose Type II coping, or personal role redefinition, when a conflict arose between the mother and worker role. Most single employed mothers preferred Type III coping, or reactive role behavior, in order to handle mother and worker conflicts. Both groups of mothers were equally satisfied with their mother role. Satisfaction with the worker role was not related to the type of coping strategy used. However, single employed mothers were more satisfied with their worker role than married employed mothers. The implications of this research are that the method used to categorize coping strategies, as well as marital status, may influence the results obtained when satisfaction with work and family roles is assessed. Schaper (1983) tested Hall's (1972) model on 255 married, working mothers in Los Angeles. It was hypothesized that the use of a multi-dimensional measure of role strain would yield information different from that obtained using a measure of general satisfaction. Four role strain factors, and 16 coping strategies were included in the analysis. Results indicated that the more frequent use of coping strategies aimed at changing others expectations (Type I) or attempting to meet all role demands (Type III) were 26 associated with increases in role strain, while changing one's personal concept of role demands (Type II) was associated with decreased role strain. The results of research testing Hall's (1972) model suggest that satisfaction and role strain cannot be considered identical dependent variables. This view is shared by other researchers. For example, Edwards (1983) noted that complex relationships exist between role strain, coping and satisfaction, and Olds (1980) found that androgynous women reported high levels of role conflict, but used effective coping strategies to maximize their well- being. The relationship between career and family role stage, role strain level, coping strategies and satisfaction was assessed by Edwards (1983). Subjects included 115 mothers who were employed 30 or more hours each week in a professional or managerial capacity. Role strain was lower at the plateau level of career development than at the establishment level, but was not related to family role stage. The relationship between satisfaction and coping was positive for personal role redefinition (Type II), negative for reactive coping (Type III), and unrelated to structural role redefinition (Type I). To summarize, the above studies that have tested Hall's (1972) model have failed to yield consistent results. The reported relationship between satisfaction, role strain and coping strategies has varied with the 27 methods used to assess these constructs, and with the samples surveyed. Although the factor analysis used by Gray (1979) to identify coping strategies can be faulted for including items other than Hall's (1972) coping strategies in the analysis, the inconsistent results obtained across studies provide converging lines of evidence that Hall's (1972) model does not adequately predict coherent themes within coping strategies, or the relationship between satisfaction and coping strategies. It would appear that a different model of coping strategies is called for. It may be appropriate to compare Hall's (1972) coping model with a more general coping model developed by Pearlin and Schooler (1978) on a large sample of family members in order to draw parallels between the work of these researchers, and the distinction between problem-focused and emotion-focused coping put forth by Lazarus and his associates. Pearlin and Schooler (1978) attempted to identify sources of role strain, and coping strategies used by family members in their roles as marriage partners, economic managers, parents and workers. A factor analysis yielded 17 coping factors that Pearlin and Schooler (1978) conceptualized as representative of three general coping strategies. Responses that modify the situation were viewed as the most direct way to cope with life-strains, since they were directed at altering or eliminating the source of strains. Coping factors that were associated with this coping style included negotiation in marriage, the use of 28 punitive discipline in parenting, and optimistic action in occupation. This direct action style of coping appears to be most closely aligned conceptually with Hall's (1972) Type I and Type III coping, since both are aimed at directly altering the source of strains, although the former method aims at a more permanent solution than the latter. Both Hall's (1972) Type I and Type III coping, and Pearlin and Schooler's (1978) direct action style of coping, appear to be examples of Folkman and Lazarus' (1980) problem-focused coping, which is aimed at managing or altering the person-environment relationship that is the source of stress. Specific strategies from each of Hall's (1972) catagories appear to fit within Pearlin and Schooler's (1978) direct action style of coping, e.g., having family members share household tasks (Type I); having family members assist in resolving conflicts between roles (Type I); rotating attention from home to children to work, depending on which need is most pressing (Type II); considering personal interests very important (Type II); scheduling and organizing activities carefully to optimize output (Type III); and, attempting to meet all role demands (Type III). Pearlin and Schooler (1978) identified their second coping style as responses that function to control the meaning of the problem. Positive comparisons and selective ignoring in all four roles were associated with this coping style. The substitution of rewards in occupation and the 29 devaluation of money in the area of household economics were also conceptually identified as responses that function to control the meaning of problems. This coping style is conceptually analogous to Hall's (1972) Type II coping style, personal role redefinition, which involves a change in expectations of role behaviors by the individual. Specific strategies that appear to function to control the meaning of problems include: developing new attitudes to reduce conflicts (Type II); and, reducing personal standards within certain roles (Type II). These coping strategies appear to be examples of emotion-focused coping, which functions to regulate stressful emotions (Folkman 8 Lazarus, 1980). Behaviors that aid individuals in accommodating to existing stress without being overwhelmed by it was the third coping style identified by Pearlin and Schooler (1978). This coping style "functions neither to alter the situation- generating the stress-provoking strains nor to create congenial perceptions of problematic experiences within the situation" Pearlin 8 Schooler, 1978 p. 7), and includes several orientations to life-prolems, including; denial, passive acceptance, withdrawal, magical thinking, and avoidance. Factors related to this coping style included: emotional discharge versus controlled reflectiveness in marriage; passive forebearance versus self assertion in marriage; potency versus helpless resignation in the parental role; and optimistic faith in one's financial future. Although this coping style is not conceptually related to any 30 of the three styles identified by Hall (1972), specific strategies that may be related to this style include: eliminating entire roles (Type II); and, having no conscious strategy for dealing with conflicts between roles (Type III). These coping strategies also appear to be examples of emotion-focused coping. Evidence exists which suggests that elements of Pearlin and Schooler's (1978) coping scale may be conceptually and statistically reduced to an active versus passive dimension, that reflect problem-focused and emotion-focused coping, respectively. Olds (1980) surveyed 215 working husbands and wives from corporate and academic organizations in New York City. The final sample consisted of 91 couples, and 33 "half couples." An abbreviated version of Pearlin and Schooler's (1978) coping scale was administered, which included coping items pertaining to marital and occupational roles. These items were subjected to a principal components factor analysis with a varimax rotation that required the extraction of two factors. With minor exceptions, the factor structure that emerged for males and females was identical. Factor 1, labeled “Action," contained six items that dealt with taking action, e.g., "I try to find a fair compromise in our marriage problems." Factor 2 was labeled "Selective Ignoring and Emphasis," (Ignore) and contained 13 items that described a mental strategy of ignoring difficulties and emphasizing the positive side of the situation. Items that were included in this factor had been previously identified by Pearlin and 31 Schooler (1978) as representative of cognitive meaning manipulation and stress management coping styles. Examples included "I wait for time to remedy the situation," "I just keep hurt feelings to myself," and "I try to overlook my spouse's faults and pay attention only to the good points." It appears that the Action factor identified by Olds (1980) is representative of problem-focused coping, while the Ignore factor represents emotion-focused coping. Traditional versus non-traditional occupations. Although women employed in male and female dominant occupations do not differ in their self-reported life satisfaction or self-esteem (Blackwell, 1984), women employed in traditional and non-traditional occupations differ in antecedent life experiences and personality characteristics (Lemkau, 1984; 1983; 1979). Most importantly, women employed in female-dominant occupations are characterized by stereotypically feminine, expressive personality characteristics, while women employed in male-dominant occupations tend to incorporate high levels of sterotypically masculine, instrumental personality characteristics (Lemkau, 1984; 1983; 1979). This relationship extends to females in male and female-dominant college majors (Strange and Rea, 1983; Harren, Kass, Tinsley 8 Mooreland, 1979). _ Lemkau (1979) compared women employed in male-dominant careers to women employed in female-dominant careers for levels of masculinity and femininity, assessed by the Ben Sex Role Inventory (Bem, 1974), under three instructional sets. p-v-v . 32 These sets included their perceptions of masculinity and femininity in 1) social situations, 2) job situations, and 3) the ideal woman. Women employed in female-dominant careers described themselves almost as "feminine" on the job as in a social situation, while women employed in male-dominant careers were significantly less feminine in their occupational roles. When all instructional sets were considered together, women employed in female-dominant careers were significantly more feminine than women employed in male-dominant careers. Although women employed in male- dominant careers had higher masculinity scores in all three instructional sets, these differences were not significant. Research reported in Lemkau's (1983) review that shows significant personality differences between women employed in male-dominant careers to women in general focuses primarily upon differences in scores on the Cattell 16PF. Masculine traits were thus extrapolated from scores on the Cattell 16PF scales, rather than being directly assessed by the masculinity and femininity scales of a contemporary sex role inventory such as the Bem Sex Role Inventory (Bem, 1974) or Personal Attributes Questionnaire (Spence, Helmreich, 8 Stapp, 1975). In general, women employed in male-dominant occupations were more dominant, adventurous, radical, imaginative, confident, resourceful and composed, compared to women in general (Lemkau, 1983). High levels of instrumental characteristics have been associated with social self-esteem and social competence 33 (Spence et. a1., 1975; Ben, 1977), which may predispose individuals to take an active role in solving problems that arise from the conflicting demands of multiple roles. It may be that women employed in male and female-dominant occupations differ in average levels of instrumentality, which in turn may predict an active, problem-focused coping style. This possibility is supported by research demonstrating that sex-role beliefs are related to secondary appraisals (Levo, 1983). Androgynous women had more confidence in their coping resources (higher secondary appraisals) than did feminine women, and tended to select more effective coping strategies to vignettes depicting home- career conflicts. Life satisfaction. Moreland, Gulanick, Montague and Harren (1979) suggest that instrumentality and expressiveness, assessed by current sex-role inventories such as the Bem Sex Role Inventory (Bem,1974), can be conceptualized as a general tendency to utilize instrumental (masculine) or expressive (feminine) coping strategies. Tinsley and Heesacker (1984) maintain that occupations can be characterized by their requirements to utilize instrumental or expressive coping strategies, and that occupational success may be closely related to the goodness of fit between levels of instrumental and expressive characteristics required by occupations and levels of instrumentality and expressiveness in individuals employed in certain occupations. 34 The importance of identifying the relationships between instrumental and expressive characteristics required for success in male and female-dominant occupations and those possessed by incumbents is demonstrated by research indicating that role strain is systematically related to job satisfaction. Anderson—Kulman and Paludi (1986) found that role strain among wife, mother and worker roles is inversely related to job satisfaction, and that job satisfaction is positively related to family cohesion. These authors suggest that satisfaction with one's work may promote positive coping among working mothers and their families (Anderson-Kulman 8 Paludi, 1986). Social competence. Social competence is considered to be a construct that overlaps with coping competence, and refers to the management of any social encounter (Wrubel et a1., 1981). Lazarus and Folkman (1984) note that although a clear positive relationship exists between social support and decreased stress, it is unclear whether some individuals are more socially competent, and thus able to solicit and use social support, compared to others. Research has clearly demonstrated that family support (Edwards, 1983) and spousal support (Gray, 1980) are potent predictors of satisfaction. Pearlin and Schooler's (1978) results indicate that individuals' coping strategies in response to marital role strains significantly reduced the amount of distress experienced. The most.effective responses in marriage were those that involved active problem solving. "In marriage it 35 is a reflective probing of problems, rather than the eruptive discharge of feelings created by the problems, that is among the more effective responses" (Pearlin 8 Schooler, 1978, p 11). Coping strategies had smaller effects in reducing the stress of parental and economic roles, and no effect upon occupational role. Having a variety of coping responses worked to decrease stress and marital role strain. Low self-denigration, mastery (an inner sense of control), and high self-esteem, reduced marital role strain. However, these personality variables did not reduce marital role strain to the same degree as having a wide repetoire of coping responses. These data suggest that the coping strategies identified by Pearlin and Schooler (1978) may be considered coping competencies, and that the most effective of these competencies for reducing marital role strain involve active problem-solving strategies. EEEEEEY The theorists reviewed here have approached the concept of coping with slightly different goals in mind, and their assessment of coping and subsequent theoretical vantage points have been influenced from their different starting points. Lazarus and his associates sought to construct a general model of coping that takes into account the transaction between individuals of particular types and environments with different characteristics. Instrumentation was developed that supported the contention that coping could 36 be broadly construed as consisting of problem-focused and emotion-focused strategies (Folkman 8 Lazarus, 1980). Pearlin and Schooler (1978) sought to identify coping strategies used by family members in their overlapping roles (e.g., as marriage partners, economic managers, parents and workers), and concluded that having a variety of coping strategies decreased stress. Personality variables including self-esteem and mastery also reduced strain, suggesting that personality may influence the choice of coping strategies, and should be considered as a confounding influence when comparing groups of individuals for their use of coping strategies. Hall (1972) perceived the multiple roles of women to operate simultaneously, rather than sequentially, and to create higher levels of interrole conflict for women than for men. Hall (1972) sought to identify how women coped with role conflicts and to relate coping strategies to satisfaction. His theoretical perspective that styles of coping with role conflicts could be rationally reduced to structural role redefinition, personal role redefinition and reactive role behavior was driven by Levinson's (1959) view of roles. Subsequent research with Hall's (1972) model yielded inconsistent relationships between his three coping styles and emotional well-being, assessed by satisfaction and and role strain, due to differences in the populations assessed and differences in assessment techniques. This suggests that complex relationships exist between coping 37 strategies, role strain, and satisfaction, and that the relationships between these variables may be influenced by both personality and environmental variables. Gray (1980) noted that women employed in nontraditional occupations may differ from women in general in their use of coping strategies. She attempted to correct a serious deficiency in Hall's (1972) study by obtaining demographic data from her subjects. Since demographic variables may influence the choice of coping strategies, as well as role strain and satisfaction, Gray (1980) attempted to statistically account for the influence of demographic variables on the dependent variables in her study. Gray (1980) also attempted to test Hall's (1972) idea that his 16 coping strategies represented three broad coping styles by subjecting to a factor analysis a variety of attitudinal and behavioral survey items, including Hall's (1972) coping strategies. The factor analysis was flawed in that it included items other than Hall's (1972) 16 coping strategies. Although Gray (1980) attempted to account for the influence of demographic variables upon dependent variables including role strain, career attitudes, and coping strategies, she did not compare women who work in nontraditional occupations to women working in traditional_ occupations. However, evidence suggests that women who work in nontraditional male-dominant occupations may differ from women employed in traditional female-dominant occupations in personality traits (e.g., masculinity and femininity) 38 and in situational demands specific to male-dominant and female-dominant occupational environments that may influence the choice of coping strategies. This study seeks to extend the work of previous researchers (e.g., Gray, 1980; Hall, 1972) by comparing working married women who differ in that they are employed in either male or female-dominant occupations in their use of coping strategies, career attitudes, role strain and satisfaction. In this study, Hall's (1972) 16 coping strategies will be assessed by common factor analysis in order to determine the appropriateness of Hall's (1972) three coping styles. This study seeks to statistically control demographic/situational variables and personality traits to determine if women differ in their career attitudes, role strain and use of coping strategies as a consequence of their occupational environments (e.g., male or female-dominant). By statistically identifying coherent themes within Hall's (1972) coping strategies, and controlling for the environmental and personality variables that influence the transaction between person and environment, this study seeks to bring into clearer perspective the ways in which women differ in their self-perceptions of well-being and choice of coping strategies as a consequence of male or female-dominant occupational status. CHAPTER III METHODOLOGY In this chapter the methodology of this study is described. Included in the chapter are descriptions of the sampling scheme, survey procedures, instrumentation, formal statement of hypothesis, and rationale for statistical procedures. SAEEILDQ BBQ DALE QQLISQELQB EIQQSQBEQQ The validity of statistical inference, i.e., the accuracy of generalizations drawn about population parameters from sample statistics, depends upon how representative the sample is of the population (Glass 8 Hopkins, 1984). When a sample is randomly drawn from a population, the degree of accuracy of statistical inference is represented by the sampling error of the statistic. That is,-the statistic differs from the estimated parameter only by chance on any variable, and one can estimate the margin of error around a statistic (Glass 8 Hopkins, 1984). Given that inferential statistics is based on the assumption of random sampling from populations, threats to the validity of conclusions drawn from survey research often arise from the method used to obtain samples. For example, when accidental or convenience samples, street-corner polls, readers of a certain magazine, or members of a certain 39 4O organization are used to represent a population that extends beyond the sampled population, it is not possible to estimate the sampling error because these sampling strategies contain unknown types and degrees of bias, or lack of representativeness, in addition to sampling error (Glass 8 Hopkins, 1984). The criteria for conducting a sample survey are that an optimim balance is found between the precision of the estimate obtained from the sample, and the cost of the survey in terms of convenience and dollars (Raj, 1972). Mail surveys are the quickest and least expensive method of collecting information, and additionally eliminate the biases introduced into survey research by interviewers who shape and interpret respondent answers (Raj, 1972). One of the major threats to mail survey research is self-selection bias (Glass 8 Hopkins, 1984), which refers to the fact that individuals who complete and return mail surveys may differ in unkown ways from those who do not complete the surveys (Raj, 1972). When the percentage of nonrespondents to a mail survey surpasses 20 percent, it is probable that self-selection bias exists (Isaac 8 Michal, 1971; Borg 8 Gall, 1971). In general, nonrespondents to mail surveys tend to be from lower socioeconomic levels (Dillman, 1978) and to have achieved less academic success than respondents (Borg 8 Call, 1971). Although this fact of mail survey research is often treated as a given, and seldom addressed in the research 41 literature (e.g., Edwards, 1983; Olds, 1980), Raj (1972) suggests two strategies for decreasing this source of systematic bias. One method is to randomly sample the nonrespondents, and to send interviewers to collect the information from them (Raj, 1972; Isaac 8 Michael, 1971; Borg 8 Gall, 1971). "The interviewers use all the persuasion at their command to obtain a response" (Raj, 1972, p 117). The second method involves sending an abbreviated version of the original survey, containing only the essential elements, to the nonrespondents. The former method can be considered unethical, since it involves the use of coercion, and does not respect the participant's freedom to decline to participate in research, which is mandated in APA Ethical Principle 9 (American Psychological Association, 1982). The latter method does not take into account the strong possibility that it is impossible to identify the "essential" parts of a survey if the preliminary data are systematically biased. A reasonable alternative approach, which respects individuals right to decline to participate in a survey, yet also increases the percentage of survey respondents to a level that often exceeds 80 percent, is to use elements of the Total Design Mail Survey Method (Dillman, 1978). For a total of fifty mail surveys, the average response rate was 74 percent for researchers who used this method (Dillman, 1978). The Total Design Method (TDM) for mail survey research includes the following elements. A cover letter is mailed to 42 survey recipients along with the survey, which communicates an explanation of the aims of the research, its benefit to a group with whom the recipient identifies, and the importance of the respondent to the study's success. The letter has an appropriate letterhead, is accurately dated, and is signed by the researcher. Exactly one week after mailing the letter and survey, a follow-up postcard is sent to all recipients of the first mailing. The note on the postcard thanks those who have already returned their questionnaires, and reminds those who have not. It is signed by the researcher. A second follow-up is mailed to nonrespondents exactly three weeks after the original mailout, which consists of a cover letter than informs them that their questionnaire has not yet been recieved, restates the original appeals of the first letter, includes a replacement questionnaire, and another return envelope. A third and final follow-up is mailed seven weeks after the original mailing, which consists of a cover letter, another questionnaire and return envelope, and is sent by certified mail to the remaining nonrespondents. The use of the follow up methods described above in five statewide surveys increased the return rate percentage of nonrespondents by 24 to 29 percent of the nonrespondents for each of the first two follow ups, and 33 percent by the certified mailing (Dillman, 1978). From 19 to 27 percent of questionnaires were returned prior to the first follow up postcard mailing, and the first follow up increased the percentage of total responses from 15 to 25 percent. The 43 final response rates ranged from 70 to 75 percent. Dillman (1978) notes that the returns from the first two mailings are greater in studies that utilize specialized populations. The Total Design Method relies on personalization throughout the implementation process, which communicates the importance of each individual's participation to the research (Dillman, 1978). The survey method to be described for this study differed from that outlined by Dillman (1978) in two respects. First, the second follow up in this tudy consisted of a postcard, rather than a letter and additional survey. Second, this study did not include a third follow up letter. It should be noted that the survey participants in this study consisted of a specialized population. Dillman (1978) observed that the return rates for specialized populations are at the high end of the range, e.g., at 27 percent for the first mailing, and 25 percent for the first follow up. Thus, it appeared that although the use of a postcard, rather than a letter and additional survey for the second follow up might decrease the return rate, the decrease would be partially offset by the use of a specialized population in this study. Also, an offer of a summary of the results of the study and list of references was extended to women participating in the study, following Dillman's (1978) suggestion that the offer of incentives may increase return rates. A final estimated return rate of 63 percent was calculated for this study, based upon average return rate estimates from five statewide 44 survey studies Dillman (1978) used to estimate return rates. That is, a 23 percent return was expected from the initial mailing, and 20 percent increments in the return rates were anticipated to result from the two follow up mailings. The sampling method decribed for this study incorporated the approaches to sample selection and respondents described above. It additionally incorporated safeguards to protect the confidentiality of all survey respondents. This sampling method aimed to obtain a final sample of 100 married working women as participants in the study. A final sample of 100 participants was needed to ensure the stability of the correlation matrix, and accuracy of statistics generated from the correlation matrix (e.g., factor analysis of coping scale items). From five to seven subjects are needed per item in factor analysis to ensure the stability of the correlation matrix (E. Amoli, personal communication, March 16, 1987; Guadagnoli 8 Velicer, in press). Since Hall's coping scale consists of 16 items, a final sample of 100 would guarantee the stability of the correlation matrix of coping scale items used to generate the factor analysis (E. Amoli, personal communication, March 16, 1987). Researchers vary in their criteria for considering an occupation to be dominated by gender. Lemkau (1983, 1984) considers an occupation to be dominated by gender if 75 percent of the incumbent workers within an occupation are either male or female. Lemkau (1983, 1984) uses statistics 45 provided by the U. S. Bureau of the Census to obtain her estimate of male or female occupational dominance. Strange and Rea (1983) consider college majors to be dominated by gender if two-thirds of the students enrolled with a major are either famale or male. Lemkau's (1983; 1984) criteria of 75 percent same gender dominance, based upon U. S. Bureau of the Census data, was used to determine male and female- dominance of occupations in this study. Married women elementary school teachers and registered nurses represented female-dominant, traditional occupations in this study. Married women university faculty members and physicians represented male-dominant, nontraditional occupations. It should be noted that married women employed as physicians and university faculty were included in this study so that results could be compared to Gray's (1980) study, which utilized women from these male-dominant occupations in her study. Teachers and nurses were included so that results could be compared to those of Blackwell (1984), who used representatives of these female-dominant occupations in her study. Consequently, marrried women in both male and female- dominant occupations were adequately sampled. Elementary school teaching may be considered a female- dominant occupation since 75 percent of the workers in this occupation are women (U. 8. Bureau of the Census, 1983). Registered nurses also represent a female-dominant occupation, because 95 percent of registered nurses are females (U. S. Bureau of the Census, 1983). 46 Postsecondary education cannot be considered to be dominated by either gender, since only 63 percent of this occupation is comprised of males (U. S. Bureau of the Census, 1983). However, the percentage of males within ten of 28 educational specialties (i.e., law, chemistry, physics and political science) exceeds 75 percent (U. S. Bureau of the Census, 1983). Therefore, a sample of women who are employed in postsecondary educational areas dominated by males could be considered to represent women employed in a male-dominant occupation. Women physicians represent a male-dominant occupation because 86 percent of physicians are males in this country (U. S. Bureau of the Census, 1986). Elementary 5211991 teachers... The population sampled consisted of women elementary school teachers in the Lansing School District. A total of 981 women elementary school are employed in the 34 elementary schools in the Lansing School District (C. Nelson, personal communication, February 9, 1987). A mailing list of women elementary school teachers was obtained from the Lansing School District (see Appendix A), and 120 elementary school teachers were randomly selected from the list. It should be noted that this list did not identify women according to marital status. Since the percentage of working married women is about 55 percent, and about 45 percent of all working women are single . (U.S. Department of Labor, 1982), the number sampled was expected to result inia final return of about 41 surveys from married women elementary teachers, which is in excess of the 47 25 surveys needed within this population to ensure the reliability of the statistical procedures used. Eggigpgpgg purses, A total of 2731 registered nurses reside in Ingham County. A current mailing list of registered nurses in Ingham County was obtained from the Michigan Board of Licensing and Regulation. The list was arranged alphabetically and was not broken down by gender or marital status. A total of 120 women were randomly selected from the list of registered nurses residing in Ingham County to receive surveys. Univepsipy fagulty, As of Fall, 1986, 1082 women were employed in faculty and specialist positions at Michigan State University. A total of 118 women faculty members, excluding women physicians employed at Michigan State University, occupied positions in areas that are considered male-dominant, based upon Census Bureau statistics (U. S. Bureau of the Census, 1983) (see Table 1). The entire population of 118 women received surveys. Bhygipigppp A total of 676 physicians reside in Ingham County. A current mailing list of these physicians was obtained from the Michigan Board of Licensing and Regulation. A total of 91 of these physicians were women. The total population of 91 women physicians residing in Ingham County received surveys. . papa gpllgppippp The survey (see Appendices B, C and 0), along with a letter to the survey recipients (see Appendix E) and postcard (see Appendix F) was mailed with a 48 Table 1 Michigan State University Women Faculty Members Employed in Male-dominant Fields Male-dominant field 2 Chemistry 9 Physics 3 Natural Science 15 Economics 3 History 4 Political Science 6 Engineering 3 Medical Science 56 Agriculture and Forestry 19 49 stamped return envelope. One week after the survey was mailed, a follow up postcard was mailed to survey recipients (see Appendix H). Exactly one week after mailing the first follow up postcard, a second follow up postcard was mailed (see Appendix I). Although unmarried women were asked to return the survey questionnaire uncompleted (see Apendices B and E), they were informed that their opinions are valued, and asked to complete an abbreviated version of the survey (see Appendix G). The rationale for asking unmarried women to participate in the survey by completing the abbreviated version of the survey was to avoid the implication that unmarried status is less desirable than married status. Human Sppjggpp Reyigg memippggp In compliance with Michigan State University policy regarding the use of human subjects in research, permission was sought from the University Committee for Research Involving Human Subjects (UCRIHS) to conduct this study. Following review of the research proposal, UCRIHS approved the study. gppfiggppiglipyp The confidentiality of survey respondents was protected by making all returned questionnaires anonymous. That is, neither surveys nor return envelopes were coded in any way that could be used to identify respondents. Survey recipients were advised of the confidentiality of their responses in the letter that will accompanied the surveys (see Appendix E). 50 Lipitatiops. The validity of the generalizations that are drawn from this research are limited to the following populations: married women elementary school teachers employed in the Lansing School District; married women Michigan State University faculty members: married women nurses and physicians residing in Ingham County. I2§££BE§DE§ An amended version of Gray's (1980) protocol was administered to women participating in this study. The principle questionnaire consisted of these sections: background information; personal attitudes toward roles; attitudes of others toward my career; role strain and coping methods; and satisfaction with coping methods (see Appendix B). The questionnaire differed from Gray's (1980) survey in four ways. First certain questions were eliminated from the survey, including questions asking for professional status at time of marriage, periods of time not employed, occupation of father and mother, occupation and income of spouse, educational level of mother, mother's characteristics, negatively phrased coping strategies (total of 13 questions), and the open-ended question that asked for advice for a young woman in combining career and family. Additionally, spaces for comments by survey respondents were eliminated from all sections except personal attitudes towards roles.' 51 Second, the scoring methods were changed by revising the response format from open-ended to fixed choice categories on respondent's income level. Third, new variables were added including: a) marital status; b) three questions assessing satisfaction with career, marital, and mother roles that are derived from the research of Harrison and Minor (1982); and c) three questions that assess role strain of parental, marriage, and occupational roles. The strains derived separately from parental, marriage, and occupational roles were assessed in this study because evidence suggests that the interaction of separate sources of family and work role strains must be examined to learn how each source of role strain affects outcomes associated with well-being (Baruch, Biener, 8 Barnett, 1987). The question asking for occupation reflected the occupations surveyed in the proposed research. Fourth, the first response options to the first two items assessing personal attitudes toward roles were slightly altered in order to eliminate an evaluative bias that appeared in the original response options. The impact of the revisions to Gray's (1980) survey are not known, but were not expected to significantly impact the results of this study, e.g., the alteration of an item from open to close-ended responses should not have altered the accuracy of the response; and the items pertaining to role strain and role satisfaction used the same format as similar items on Gray's (1980) original survey. 52 Gray's (1980) survey questionnaire was developed from a pilot study that involved sending an open-ended questionnaire to a small sample of subjects that included 10 married women doctors, 10 lawyers, and 10 professors. The information from the pilot study served as the basis for developing categories of responses with fixed alternatives to objective questions. The revised questionnaire was reviewed by 24 doctoral psychology students, and by two psychologists with experience in survey research. The final version of the questionnaire contained 64 fixed choice items, and one open-ended question. The first 16 items on the survey, which comprised Part A, asked for background information on occupation, age, salary, and children. Part B, consisting of four questions, asked for personal attitudes toward roles. Part C assessed the respondent's perception of the attitudes of significant others toward her career. Part D, consisting of four questions, used a four-point Likert-scale ranging from Strongly Agree to Strongly Disagree to assess various types of role strains, and 16 coping strategies. Coping strategies consisted of the same 16 coping responses identified by Hall (1972). The next section of the survey consisted of thirteen of these coping strategies, rephrased in order to reflect the opposite meaning of the original statement, in order to assess the internal consistency of this part of the questionnaire. The next question assessed satisfaction with handling roles. The final question, which had an open-ended 53 response format, asked for advice on how a young woman might best combine a demanding career and family. Reliability was determined by test-retest and by computing correlations between opposite statements. Test- retest was performed by having 10 doctoral students complete the entire survey a week apart. No significant differences were found in responses. The internal check of reliability, consisting of correlations between opposite statements of coping strategies, ranged from .45 to .19, with all but three having opposite signs. The validity of the questionnaire was assessed by interviewing 15 of the women who had completed the questionnaire, and comparing the information from the interview with the women's responses to the survey. Gray (1980) used the questionnaire as the basis for the interviews, and concluded that the survey results were validated by information from the focused interviews. Eepsopal Attpibutes Qpestiopnaipe (PAQ), The Personal Attributes Questionnaire (PAQ) was designed to assess masculinity and femininity (Spence 8 Helmreich, 1978). The short form of the FAQ is divided into three eight-item scales, labeled Masculinity (M), Femininity (F), and Masculinity-Femininity (M-F). Each scale of the short form of the FAQ is composed of eight bipolar items describing personal characteristics, on which respondents are asked to rate themselves on a five-point scale (Spence 8 Helmreich, 1978). 54 Each item is scored 0 to 4, a high score on items assigned to the M or M-F scale indicates a masculine response, and a high score on the F scale indicates a feminine responses. The 24 item-form of the PAQ is a shortened version of an original 55-item PAQ. The initial version of the PAQ was composed of items selected from a pool of over 130 items collected by Rosenkrantz, Vogel, Bee, Broverman and Broverman (1968). These itmes were characteristics students rated as differentiating the typical adult male and female. Masculine traits are conceptualized as respresenting what Parsons and Bales (1955) identified as instrumental behavior, while feminine traits refer to expressive behavior (Spence 8 Helmreich, 1978). Spence and her associates asked introductory psychology students to rate each item separately for their hypothesized responses of a) the typical adult male and female, b) the typical college student of each sex, c) the ideal individual of each sex, and finally, d) themselves. The 55 items selected for the PAQ were chosen from those for which both sexes exhibited a consistent stereotype about sex difference for the typical adult, typical student, ideal adult, as well as for themselves. The ideal ratings on the 55 items were used to divide the items into the three scales, M, F, and M-F. Items composing both the M and F scales were defined as characteristics socially desirable in both sexes, but believed to occur to a greater degree in either males or 55 feamles. Items met this criterion if the mean ratings of both the ideal man and the ideal woman fell on the same side of the scale midpoint, but towards the stereotypic masculine or feminine pole. Mean ratings of items assigned to the M-F scale lay on opposite sides of the scale mid-point, suggesting that the characteristic was desirable either in a man or a woman, but not in both (Spence 8 Helmreich, 1978). The short form of the PAQ correlates .93, .93 and .91 for M, F and M-F, respectively, with the full 55-item PAQ. Cronbach alphas for the short form were .85, .82 and .78 for M, F, and M-F, respectively, in a sample of college students. 1533; Sppigl peppyipr Ipyeppppy (TSBI), The Texas Social Behavior Inventory (TSBI) is composed of statements designed to assess the individual's self-confidence and competence in social situations. Respondents rate themselves for each item on a five-point scale, ranging from "not at all characteristic of me" to "very much characteristic of me." Responses are scored 0 to 4, high scores indicating high social self-esteem and competence. Factor analyses of the full 32-item scale yielded three factors that were labeled self-confidence, social dominance, and social competence (Helmreich, Stapp, 8 Ervin, 1974). Correlations of the TSBI with the Marlowe-Crowns social desirability scale in a sample of college students were low (.12 in women and .29 in men) (Spence, Helmreich, 8 Stapp, 1 1975). 56 The TSBI short form, consisting of 16 items (Helmreich 8 Stapp, 1974), correlates .96 with the long form. Cronbach alpha for the short form in a sample of college students was .91. Mesa; The primary questions of interest in this study were: 1) Do married women employed in female-dominant careers differ from women employed in male-dominant careers on salient personality traits, i.e., instrumentaltiy, expressiveness, social competence and social self-esteem? 2) Do married women employed in female-dominant occupations differ in their use of coping behaviors from women employed in male-dominant occupations, e.g., are different coping strategies utilized to balance work and family roles? 3) Do women employed in female-dominant careers differ from women employed in male-dominant careers in career attitudes, role strain, and satisfaction? The following hypotheses were tested to answer these primary research questions. 1. Women employed in male-dominant occupations will score significantly higher than women employed in female-dominant occupations on the Masculinity scale of the Personal Attributes Questionnaire (PAQ), and on the Texas Social Behavior Inventory. Women employed in female-dominant occupations will score significantly higher than women 57 employed in male-dominant occupations on the Femininity scale of the PAQ. 2. Women employed in male-dominant occupations will not differ from women employed in female-dominant occupations in their use of coping strategies. 3. Women employed in female-dominant occupations will not differ from women employed in male-dominant occupations in personal attitudes toward roles, attitudes of significant others towards careers, role strains, or satisfaction. 83:11.11;ng The research strategy was conceptualized as consisting of demographics, personality dimensions (i.e., PAQ and TSBI), occupational status (male or female-dominant), attitudes towards career (self and significant others), role strains, coping strategies, and satisfaction (8. Raudenbush, personal communication, April 2, 1987). In this model, both demographic variables and personality dimensions were viewed as possible causal influences upon type of occupation chosen. Type of occupation was then viewed as influencing career attitudes, role strains, coping strategies and satisfaction. Since the purpose of this study was to learn if occupational status (male or female-dominant) influences career attitudes, role strains, coping strategies, and satisfaction, demographic variables and personality dimensions on which occupational groups (male and female-dominant) statistically differed would be treated as confounding variables. This was accomplished by using covariates in the MANOVA 58 used to test Hypothesis 2. Covariates were included in the ANCOVA analyses used to test for occupational group differences on the dependent variables in Hypothesis 3, excluding personal attitudes toward career, which consisted of categorical response options and was tested with chi square. An additional issue of concern was that the number of dependent variables tested in Hypothesis 3 should be manageable so that substantive interpretations of occupational group differences on these variables whould be feasible (S. Raudenbush, personal communication, April 2, 1987). That is, the construct "Attitudes of Others Toward my Career" had eight items, "Role Strain" had seven items, and ”Satisfaction With Coping Strategies" had four items. Therefore, items were grouped together into scales on the basis of existing theoretical considerations if an acceptable level of internal reliability was obtained for each scale (:!.60 or greater). These scales would then be used as single scores for the construct in further analyses. WW fiyppphesis 1p This hypothesis was tested with Multivariate Analysis of Variance (MANOVA), which allows the effects of multiple dependent variables to be assessed (Norusis, 1985). The dependent variables were scores on 1) the Masculinity scale of the Personal Attributes Questionnaire (PAQ), to assess instrumentality; 2) the Femininity scale of the PAQ to assess expressiveness, and 59 3) the Texas Social Behavior Inventory (TSBI), for which a single score assesses social competence and self-esteem. Following the MANOVA analyses, univariate F tests were employed to determine significant occupational group differences on factor levels only if the multivariate F was significant. Mean scores of married women employed in male- dominant occupations were expected to be significantly higher than mean scores of women employed in female-dominant occupations on a) the Masculinity scale of the PAQ, and b) the TSBI. Mean scores of women employed in female-dominant occupations were expected to be significantly higher than mean scores of women employed in male-dominant occupations on the Femininity scale of the PAQ. fiyppphegig 2; This hypothesis was tested with common factor analysis and MANOVA. Hall's (1972) 16 coping strategies was subjected to a common factor analysis in order to identify psychologically meaningful dimensions underlying the coping strategy items. Common factor analysis, rather than principle component analysis, was utilized because the former procedure is based upon the assumption that observed -correlations between variables are primarily the result of an underlying regularity in the data, e.g., the common determinants, as well as by idiosyncratic, or unique determinants, while the latter procedure makes no assumption about the underlying regularity of the data (Kim, 1975). Since Hall (1972) maintained that the 16 coping items represent strategies used to cope with the strains of 60 multiple roles, it must be assumed that correlations between the 16 coping strategy items are primarily the result of underlying regularities in the data, e.g., coping strategies. The maximum-likelihood method of factor extraction, a common factor analysis procedure (Norusis, 1985), was used to extract factors. Since the number of factors to be extracted can be estimated, based upon Hall's (1972) assumption that the 16 coping strategies represent three categories of role behavior, the number of factors to be extracted was based on this a priori assumption (E. Amoli, personal communication, March 3, 1987; Kim, 1975), and set at three. Factor scores were then generated for coping strategy factors. MANOVA was utilized to determine occupational group differences on factors obtained in the common factor analysis. Multivariate and univariate F tests were used to determine significant occupational group mean differences on coping strategy factors. Hypophesis 3p Contingency tables and chi square tests of significance were used to test this hypothesis for personal attitudes toward roles. Crosstabulation tables assess the relationships between a) occupational groups, and b) the possible responses to each question assessing personal attitudes toward roles (2 X 4) tables. Contingency table analysis is an appropriate test for the relationships between occupational groups and personal attitudes toward roles because the possible response choices to items are categorical . ANCOVA occupational attitudes of satisfaction statistic to for interval variables to 61 was used to test this hypothesis for group differences (male or female-dominant) on others toward my career, role strains, and with coping strategies. ANCOVA is an appropriate use in this analysis because it is appropriate data, and allows the effects of confounding be statistically controlled (Norusis, 1985). HIEEEZ‘J CHAPTER IV RESULTS & DISCUSSION A total of 449 surveys were mailed to prospective participants. A total of 226 surveys were returned. Fourteen surveys were returned by the post office as undeliverable as addressed (see Table 2). Of the total number of surveys received, one was excluded because the respondent indicated she was employed in two of the four occupations sampled (i.e., university faculty and physician), and six were excluded because the respondents indicated they were not currently employed. A total of 219 usable surveys were recieved for a response rate of 51 percent. The final response percentage was calculated by excluding the fourteen undeliverable surveys and the seven unusable surveys from the initial number mailed (449) to obtain an adjusted number mailed (428), and dividing the number of usable surveys (219) by the adjusted number mailed (428), following the procedure used by Gray (1980). Of the 219 surveys that were used in this analysis, 42 were completed by physicians, 68 were completed by elementary teachers, 42 were completed by registered nurses, and 67 were completed by university faculty (see Table 2). The response 62 Table 2 63 Distribution of Returns and Non-Returns Physicians Elementary Registered University Teachers Nurses Faculty Mailed 91 120 120 . 118 Returned by 7 0 7 0 Post Office Unusable l 0 6 0 (Non-working or occupation unclear) Mailed minus 83 120 107 118 excluded Completed 42 68 42 67 surveys Response 50.6 56.6 39.2 56.7 percentage Married 31 46 32 39 women Percentage 73.8 67.6 76.1 58.2 married Single 11 22 10 28 women Percentage 26.2 32.4 23.9 41.8 single 64 percentage for each occupation, calculated with adjusted numbers mailed for each occupation, was: physician (50.6 8); elementary teacher (56.6 %); registered nurse (39.2 t); and university faculty (56.7 %) (see Table 2). A total of 148 (67.5 %) of the 219 usable surveys were completed by married women, and 71 (32.5 8) were completed by single women. The number of usable surveys received from married women within each occupation was: physician 31 ' TE (73.8 8); elementary teacher 46 (67.6 %); registered nurse 32 ‘ (76.1 %); and university faculty 39 (58.2 %). The number of usable surveys recieved from single women within each occupation was: physician 11 (26.2 %); elementary teacher 22 (32.4 %); registered nurse 10 (23.9 %); and university faculty 28 (41.8 8) (see Table 2). WW Demographic data for married women is reported in this section with contingency tables. The relationship of each demographic variable and each of the four occupations is first presented, followed by the relationship of the demographic variable and male and female-dominant occupational status. Demographic variables on which male and female-dominant occupational groups differed significantly were included in subsequent relevant statistical analyses as covariates. Since demographic data were recorded as interval data, ANOVA with a two-tailed significance test was used to determine significant differences between male and female- dominant occupational groups on demographic data. 65 0.00H hva m.w~ mm m.HN mm 0.0n mv H.HN Hm Hmuou 30m vs. V a Hm u an an.sm n m m.v m.m m.oH v.- m.nm m.m~ H.m v.a s 0H 6H an mm mm m m m.m~ o.oH o.om ~.e~ m.- m.- n.nn o.om H.m m.~ m.o~ m.o~ m.o~ m.o~ s.e m.~ m H m m m m n H n.8H o.on m.~H ~.v~ n.8H m.- ¢.v¢ o.om H.n v.m n.m o.m~ o.mH o.m~ m.~a H.n H n m m m m e a H.sm o.on o.mm «.me ¢.Hn o.o~ ~.- 0. m.m s.m m.m H.Hn e.gm m.ma 8.4 o. e n v «a as e m o o. o.on m.~a H.m v.an n.en o. o. o. s.m m.m s.m m.mn s.mn o. o. o n m n as «H o o omuwm mmusm omume mvuav osumm mmuan onnom mmuam Hmuou cecaou huaoosu >uwmuo>aco owns: cmuoumfimom mucous» xnmusmsmam cmaoamanm sowummsooo ucmoumm cacaoo unmoumd 30m ucsoo musmocommmm possum: no mac n OHQMB 66 0.00H bed v.mm hh m.bv or Have» 30“ mm. v a u up oo.m u x N m.¢ m.w m.oa ¢.~m m.mm m.n~ H.m v.H HmuOO s 0H 0H mm mm mm m m cacaoo v.Hb n.6n m.bn 5.0m b.mv m.m¢ h.mm 0.0m ucmcfifiop m.m w.h m.h m.m~ m.om m.mH w.h m.H ImHsEmm m w m mm ma ma m H m.w~ o.ov m.~m m.mn n.vm H.5m m.mn 0.0m ucmcfifiop m.m b.m m.va b.ma H.5N m.mm n.e ¢.H Imam: N e oH Ha ma om n H mcumum Hmcoflummsooo oolom mmlam omlme mvnae ovlmm mmlam omlmm mmIHN unmouom CEDHOU ucmoumm 30m assoc mucmpsommmm coauumz mo ems e OHQMB 67 0.00H hva a.mN on w.HN mm n.Hn ov H.HN an Hmuou 30m woo. v a mH u as mm.en u x m m.v m.m s.mH H.n~ m.m~ e.s~ Hmuou s nH an an on em cesHoo n.4H H.n~ ~.5H m.m~ m.mm m.vn suHsomu m.~ m.s ~.nH s.n~ m.m~ s.n~ suHmuw>Hcs H n m a HH m n.6H m.on ~.sH «.mn m.mH ~.mH muss: H.n m.~H w.mH v.8n m.mH 6.mH smumumHmmm H v m HH 6 m m.~e m.mn H.~m ¢.m~ H.H~ s.s umzommu. m.m m.oH H.mn n.H~ ¢.sH n.e sumucmsmHm n m mH 0H m m m.w~ 5.5 v.m m.HH ~.en m.mn m.w ~.n ~.n m.~H m.H¢ n.~n :mHonsna w H H v nH oH cofiummcooo OHOE no em mmuHm owumH mHuHH OH:m muH :oHummsooo uswmmum ucmoumm cssHoo ucwouom 30m uccou :H pmxuoz mummy m OHQMB 68 o.ooa ova H.nm mm m.m¢ mm Hmuou 30% v a mooo. m u up Hw.HN u x m.v m.m >.mH H.n~ m.mm >.>H Hmuou 5 MH mm em mm mm cscHoo H.5m ~.mm n.mb m.Hm m.mm m.mm ucmcHEOU H.m m.HH m.mm m.wm m.pH o.m ImHmEmm v m mm Hm vH h m.mv m.om h.om m.mm «.mm H.mh ucmsHEOp n.v m.m b.m m.mH m.vm m.>~ ImHm: m e m mH vm mH msumpm Hmcofipmmsooo duos no om mmIHm oman mHuHH OHIm mIH unmoumm sEoHoo ucmonmm 30m assoc coflummsooo ucmmwum CH pwxuoz mass» m OHQMB .q-u.cu!ir\ n!-th~< \: 9.. ~ ellin‘. 69 o.ooa mvH N.@N mm H.N~ mm h.Hn MV o.o~ mm Hugo» 30m mooo. v d mH u an so.e¢ m.mn n.mH m.mH H.nH 6.8 H.v v.H Hmuou mm mm sN mH HH 6 m cssHoo m.mm m.~e ~.- m.mH H.m s.mH o. suHsomu m.mn m.Hn m.mH m.n m.~ m.~ o. suHmum>Hco mH «H m n H H o m.nH s.oH «.8H H.~e m.vm n.nm o.ooH «mus: m.H~ v.m m.~H o.m~ m.mH n.m n.m nmumumHmmm s n v m m m n ~.H~ m.~e o.sn m.mn 4.6m n.mn o. umnommu m.nm H.m~ h.H~ «.mH h.m n.v o. >HsucmEmHm HH NH 0H s v m o m.wn o.n m.m~ n.m o. n.6H o. m.mm e.n H.¢~ e.n o. v.n o. :mHonsnm mH H s H o H o :oHusdoooo macs so mmm.mmu mmm.amu mmm.m¢n mmm.mmn mmm.m~u www.mHa ooo.osm ooo.owm ooo.omw ooo.oem ooo.onm ooo.o~m ooo.on vacuums cssHoo ucmouwm 30m ucsoo panama: can wqu no xumHmm cmcwnaoo Hszcc< b canoe Noo. v Q m MU mh.om H x 70 o.ooa m.mn n.6H 6.6H H.mH 6.8 H.6 6.H Hmuou mvH mm mm hm mH HH 6 m casHoo 6.66 6.66 6.H6 6.6s 6.66 n.66 6.66H uzmcHson 6.66 H.n~ ~.6H 6.sH ~.mH 6.~H H.6 6.6 umHmsmm ms 6H 6H 6H 6H 6H 6 m 6.66 6.66 H.66 H.H~ H.6 6.66 6. ucmcHsos 6.66 p.66 6.6H 6.mH 6.6 6.H 6.6 6. an62 s6 66 6H 6H 6 H m 6 msumum HmcoHummsooo H666» macs no 666.66: 6m6.66: 666.66: 666.mmu 666.6~: 666.6H: 36m 666.6» 666.66 666.66 666.66 666.66 666.6m 666.6H unmoumm cacHou unmouom 30m ucsoo ccmnmcm can wMHz mo >H6Hmm pmansou Hmscc< m OHQMB 71 0.00H hVH m.w~ an w.HN NM w.0n mv H6666 30K H666. v d 6H n 66 66.66 u 6.6 6.6H H.66 5.5H 5.5H H.6 5.5 6H 55 66 65 6H 6 6 6.65 6.66 6.55 6.65 6.6 6. 6. 6.5H 6.66 6.66 6.5H 6.5 6. 6. 5 5H 5H 5 H 6 6 6. 5.6 6.6 6.66 6.66 6.55 6.65 6. H.6 6.6H 6.H6 6.65 6.H5 H.6 6 H 6 6H 6 5 H 6.6H 6.65 6.66 6.6H 5.55 H.HH 6.65 5.5 6.5H 6.66 6.6 6.6 5.5 5.6 H 6 65 6 6 H 6 6.65 5.55 6.55 5.6H 6.55 H.HH 6. 6.6 6.6H 5.66 H.6H H.6H 5.6 6. 5 6 5H 6 6 H 6 OHOE no H6 66-H6 66sH6 66sH6 66aH5 65uHH 6H|H H6666 cssHoo huHoomm >uwmuo>6ca onus: pwuwumHomm Honommu >kucmsmHm :mHonwcm c06ummcooo UCOOHQQ CEfiHOU unwouom 30m uccoo xmwz HmoHd>B 6 2H cwxuoz 65:0: 6 mHnms 72 c.00H bed v.Nm hh 0.5v on H666» 30m 1a.... Hoo. v Q m u up 5H.H5 n 5.6 v.5H 6.65 6.6H 5.5a H.6 6.5 Hana» 6H 55 mm 65 6H m 6 sEsHoc 0.0H 5.55 6.6m 5.5m 5.00 m.wm o.ooa usmsafioc n.H b.HH 5.65 5.5H m.mH v.0H 5.m Imamfimh H 6 m5 6H 5H 6 6 0.0m p.66 5.m6 5.66 5.55 H.HH o. usmsfifioc m.5H h.m5 5.65 H.6H 6.5 6.H o. Imam: 6 6H 65 5H 6 H o msumum HmsOHumsscoo 650E 50 H6 omIHm omIHv olem omIHm ONIHH OHIH usmosws sssHoc uswouwm 30m assoc x663 Hmch>B 6 s6 cmxuoz musom 0H OHQMB 73 Table 11 Length of Marriage to Present Spouse Count Row Percent Column Percent 1-5 6-10 11-15 16-20 21 or years years years years more Occupation 6 9 10 v o 6 Physician 19.4 29.0 32.3 .0 19.4 24.0 27.3 40.0 .0 14.6 2 11 6 10 16 Elementary 4.4 24.4 13.3 22.2 35.6 teacher 8.0 33.3 24.0 43.5 39.0 9 7 2 4 10 Registered 28.1 21.9 6.3 12.5 31.3 nurse 36.0 21.2 8.0 17.4 24.4 8 6 7 9 9 University 20.5 15.4 17.9 23.1 23.1 faculty 32.0 18.2 28.0 39.1 22.0 Column 25 33 25 23 41 total 17.0 22.4 17.0 15.6 27.9 2 x .= 25.11 df = 12 p < .01 Row total 31 21.1 45 30.6 32 21.8 39 26.5 147 100.0 74 Table 12 Length of Marriage to Present Spouse Count Row Percent Column Percent 1-5 6-10 11-15 16-20 21 or years years years years more Occupational status 14 15 17 9 15 Male- 20.0 21.4 24.3 12.9 21.4 dominant 56.0 45.5 65.0 39.1 36.6 11 18 8’ 14 26 Female- 14.3 23.4 10.4 18.2 33.8 dominant 44.0 54.5 32.0 60.9 63.4 Column 25 33 25 23 41 total 17.0 22.4 17.0 15.6 27.9 X = 7.59 df = 4 p < .10 Row total 70 47.6 77 52.4 147 100.0 Table 13 Number of Children Presently Count Row Percent Column Percent 75 Living in Home 0 1 2 3 4 6 Occupation 3 12 14 2 0 0 Physician 9.7 38.7 45.2 6.5 .0 .0 7.9 28.6 26.9 16.7 .0 .0 12 12 15 5 2 0 .Elementary 26.1 26.1 32.6 10.9 4.3 .0 teacher 31.6 28.6 28.8 41.7 66.7 .0 7 7 14 2 1 1 Registered 21.9 21.9 43.8 6.3 3.1 3.1 nurse 18.4 16.7 26.9 16.7 33.3 100.0 16 11 9 3 0 0 University 41.0 28.2 23.1 7.7 .0 .0 faculty 42.1 26.2 17.3 25.0 .0 .0 Column 38 42 52 12 3 1 total 25.7 28.4 35.1 8.1 2.0 .7 x - 18.97 df 15 p < .21 Row total 31 20.9 46 31.1 32 21.6 39 26.4 148 100.0 76 0.00H wva 5.5m mu m.bv Ob Hmuos 30m av. mm.v x 0.5 H.m H.mm 6.55 h.m5 Hmuou H 6 5H 56 56 56 sEsHoc o.ooa o.ooa n.5m w.mm 5.m6 0.0m usmsflfioc 6.H 5.6 0.6 5.56 6.65 6.65 Imamsmm H 6 5 65 6H 6H 0. o. b.H6 5.66 5.6m 0.0m usmsfifioc o. o. H.b 5.55 5.55 5.65 Imam: o o m 65 65 6H msumum Hmsoaummscoo m 6 m 5 H o usmosom sEsHoc usmcumm 30m ussoc 050: :6 msH>HH hHusmmmHm schHHsc mo HmnEsz 65 OHQMB Table 15 77 Age of Youngest Child in the Home Count Row Percent Column Percent Occupation Physician Elementary teacher Registered nurse University faculty Column total Infant 6-10 11-15 16-20 21 or to 5 older 18 2 3 4 1 64.3 7.1 10.7 14.3 3 6 36.0 9.5 17.6 222.2 25 0 12 10 5 5 2 35.3 29.4 14.7 14.7 5.9 24.0 47.6 29.4 27.8 50.0 9 4 6 5 1 36.0 16.0 24.0 20.0 4.0 18.0 19.0 35.3 27.8 25.0 11 5 3 4 0 47.8 21.7 13.0 17.4 .0 22.0 23.8 17.6 22.2 .0 50 21 17 18 4 45.5 19.1 15.5 16.4 3.6 X = 11.02 df = 12 p < .52 Row total 28 25.5 34 30.9 25 22.7 23 20.9 110 100.0 Table 16 78 Age of Youngest Child in the Home Count Row Percent Column Percent Infant 6-10 11-15 16-20 21 or to 5 older Occupational status 29 7 6 8 1 Male- 56.9 13.7 11.8 15.7 2.0 dominant 58.0 33.3 35.3 44.4 25.0 21 14 ll 10 3 Female- 35.6 23.7 18.6 16.9 5.1 dominant 42.0 66.7 64.7 55.6 75.0 Column 50 21 17 18 4 total 45.5 19.1 15.5 16.4 3.6 = 5.75 df 4 p < .21 Row total 51 46.4 59 53.6 110 100.0 79 Age; The majority of the respondents in this study were between 31 and 45 years of age (70 %) (see Tables 3 and 4). Male-dominant (M - 39.71) and female-dominant (M - 40.75) occupational groups did not differ in age £(1,145)-.65, p>.05. xggpp emplpygd in Qppppgpipnp The highest percentage of participants had been employed for six to ten years (25.9 %) (see Tables 5 and 6). Male-dominant (M - 10.17) participants were employed for significantly fewer years than individuals in the female-dominant (M - 14.68) occupational group 3(1,145)-15.34, p<.0001. Qombined salapy. The largest proportion of respondents reported that their combined salaries, i.e. the salary of the wife and husband combined, exceeded $70,000 (35.9 %) (see Tables 7 and 8). The combined salaries of women employed in male-dominant occupations were higher (M - 60,300) than were the combined salaries of women employed in female-dominant (M - 49,500) occupations F(1,l43)=19.20,.p<.0001. Mpppg wopked. A large percentage of the married women in this study work from 41 to 50 hours per week (36.1 %) (see Tables 9 and 10). Women employed in male-dominant occupations (M - 49.1) work more hours per week than women employed in female-dominant (M = 38.81) occupations F(1,145)-20.5, p<.0001. Lepgph pf mappiage. A large proportion of women reported being married for more than 20 years (27.9 8) (see Tables 11 and 12). While the mean length of time women 80 employed in female-dominant occupations (M - 16.12) had been married was slightly longer than women employed in male- dominant (M - 14.13) occupations £(l,l45)-1.15, p>.05, the difference was not statistically significant. gpilgpgp 1p appep A large proportion of the married women reported having two children presently living in the home (35.1 8) (see Tables 13 and 14). Women in female- dominant occupations (M s 1.49) had slightly more children living in their homes than did women employed in male- dominant (M - 1.20) occupations fi(1,146)-2.63, p>.05, although the differences between occupational groups was not significant. Agg pf youpgest child, The age of most of the children living in the homes of married women was five years or younger (45.5 %) (see Tables 15 and 16). Children of women employed in male-dominant (M - 7.12) occupations were' significantly younger than children of women employed in female-dominant (M = 9.71) occupations E(1,108)-3.83, p<.05. It should be noted that the age of the youngest child in the home was not included as a covariate in later analyses since it would have restricted the analyses to women with children living in the home. ' 8211221112! The reliability of separate scales within the survey was computed with Cronbach alpha. Three scales included in the survey questionnaire are known to contain an acceptable level of internal reliability, i.e., the Masculinity scale and 81 Femininity scale of the Personal Attributes Questionnaire (PAQ), and the Texas Social Behavior Inventory (TSBI) (Spence 8 Helmreich, 1978). Cronbach alpha for the Masculinity and Femininity scales of the PAQ in this study were .78 and .77, respectively. Cronbach alpha for the TSBI in this study was .83. Eight items included in "Attitudes of Others Toward my Career" were used by Gray (1980) to assess a support dimension. While six of these items appear to have face validity for the assessment of perceived support from significant others, i.e. survey items 14, 15, 16, 19, 20, and 21, two items appear to assess perceived competitiveness between the respondent and her spouse (items 17 and 18). Since Gray (1980) presented no evidence that support and competitiveness can be represented as a bipolar dimension, items 17 and 18 were excluded from the scale labeled Support in this analysis. Cronbach alpha for the six items included in the Support scale was .62. Seven survey items were used to assess role strain (items 22 to 28). One item (item 26) pertained only to women with children, and was excluded from the scale labeled Role Strain. Cronbach alpha for the six items used to construct the Role Strain scale was .84. Four items assessed "Satisfaction With Coping Strategies" (items 45 to 48). Item 47 pertained only to women with children, and could not be included with the other three 82 items to construct a scale. Cronbach alpha for the remaining three items assessing satisfaction was .55. Separate scale scores were assigned to participants for scales on which an acceptable level of internal reliability was obtained (.60 or greater). These scales were the Masculinity and Femininity scales of the PAQ, the TSBI, and the Support and Role Strain scales. Mypothesis 1 This hypothesis stated that women employed in male- dominant occupations will score significantly higher than women employed in female-dominant occupations on the Masculinity scale of the PAQ, and on the TSBI. Women employed in female-dominant occupations will score significantly higher than women employed in male-dominant occupations on the Femininity scale of the PAQ. This hypothesis was tested with Multivariate Analysis of Variance (MANOVA), which allows the effects of multiple dependent variables to be assessed (Norusis, 1985). The dependent variables were scores on 1) the Masculinity scale of the PAQ; 2) the femininity scale of the PAQ; and 3) the TSBI. Male and female-dominant occupational groups significantly differed on the dependent variables p(3,145)-4.27, p<.006. Following the multivariate analysis, univariate analyses were performed on the individual measures. As predicted, women employed in male-dominant occupations (M - 3.78) scored significantly higher than women employed in female-dominant 83 occupations (M - 3.54) on the Masculinity scale £(1,145)-6.44, p<.01 (see Table 17). Women employed in male- dominant occupations (M - 3.64) scored higher than women employed in female-dominant occupations (M - 3.52) on the TSBI. Although the mean scores of the two occupational groups were in the predicted direction on the TSBI, male-and female- dominant groups did not significantly differ on this measure E(l,145)=1.64, p>.05. (see Table 17). As predicted, women employed in female-dominant occupations (M - 4.02) scored significantly higher than women employed in male-dominant occupations (M - 3.82) on the Femininity scale £(1,145)=6.19, p<.01 (see Table 17). In summary, this hypothesis was partially supported. Individual scores on the Masculinity and Femininity scales of the PAQ were included in subsequent relevant analyses as covariates in order to provide statistical control for the confounding effects of masculine, instrumental traits and feminine, expressive traits upon dependent variables. We}. This hypothesis stated that women employed in male- dominant occupations would not differ from women employed in female-dominant occupations in their use of coping strategies. This hypothesis was tested with common factor analysis and MANOVA. Hall's (1972) 16 coping strategies were subjected to a common factor analysis in order to identify psychologically meaningful dimensions underlying the coping 84 Table 17 Means and Standard Deviations of Masculinity, Femininity and Texas Social Behavior Inventory Scales Occupational Status Male-dominant group Female-dominant group Masculinity Scale * 5 3.782 3.547 s_ .535 .583 n 70 77 Femininity Scale * g 3.823 4.024 s9 .479 .498 p 70 77 Texas Social Behavior Inventory M 3.638 3.524 s2 .559 .521 Q 70 77 Note: * p < .01 6H3 85 strategy items. The maximum-likelihood method of factor extraction, a common factor analysis procedure (Norusis, 1985), was used to extract factors. The number of factors to be extracted was set at three, based upon Hall's (1972) assumption that the 16 coping strategies represent three categories of role behavior. A rotated factor solution was obtained, in which all three factors had eigenvalues greater than 1.0. The respective eigenvalues for Factors 1, 2 and 3 are 1.69, 1.51, and 1.01. Factors 1, 2, and 3 accounted for 10.6 percent, 9.5 percent, and 6.3 percent of the total variance in the analysis, respectively. A total of 26.4 percent of the total variance was accounted for by the three factors. Seven coping strategy items had substantial loadings (2|.3OI) on Factor 1 (see Table 18). These items, followed by their respective factor loadings, are: "Family members share household tasks with me" (.32); "Family members help me resolve conflicts between my roles" (.30); "I have established rules and priorities for dealing with roles (e.g., if my child is sick, I stay home from work)" (.65); "I have reduced my standards within certain roles (e.g., I no longer care if the house is spotless)" (.49); "I have developed new attitudes to reduce conflicts (e.g., I no longer think I can be "superwoman" and do everything well)" (.67); "I attempt to meet all role demands by doing everything people expect of me" (-.35); and, ”I have no 86 Table 18 Rotated Factor Matrix Factor Coping Strategy Item Balanced Direct Planful I II III "...eliminated activities..." (29) .20 i .96 * -.18 "...hired outside help..." (30) .18 .37 * .02 "...change societal..." (31) .00 .09 -.11 "...share household tasks..." (32) .32 * .16 .05 "...resolve conflicts..." (33) .30 * .13 .05 "...overlap roles..." (34) .08 .12 -.10 "...established rules..." (35) .66 * -.13 .24 "...reduced standards..." (36) .49 * .15 -.23 "...new attitudes..." (37) .67 * .08 -.01 "...roles totally separate..." (38) .03 -.13 .30 * "...rotate my attention...” (39) .25 .11 -.03 "...consider interests..." (40) .04 -.00 .41 * - "...tend to schedule..." (41) .04 .16 .78 * "...doing everything..." (42) -.36 * .24 .04 "...eliminated entire roles..." (43) -.19 .20 -.02 "...no conscious strategies..." (44) -.38 * -.01 -.23 Note: * a factor loading 3 |.30| Balanced - Balanced Role Redefinition Direct - Direct Role Redefinition Planful - Planful Role Organization 87 conscious strategies for dealing with conflicts between roles” (-.38). Coping strategy Factor 1 was labeled Balanced Role Redefinition since the content of items defining this factor appeared to reflect a balanced approach to role redefinition aimed at altering the person-environment relationship creating role conflicts, as well as attempting to regulate stressful emotions. Two items had substantial loadings on coping strategy Factor 2: "I have eliminated certain activities within roles (e.g., I no longer do the cooking)" (.96): and, "I have hired outside help to assist me with chores" (.37). This factor was labeled Direct Role Redefinition because it appeared to reflect a direct approach to coping with role conflicts that included an active restructuring of personal roles. Three items loaded substantially on Factor 3. These items were: "I keep my roles totally separate (e.g., I never bring work problems home or home problems to work)" (.30); "I consider my personal interests very important (e.g., I make time to play the piano)" (.41); and, "I tend to schedule and organize activities very carefully to maximize my output" (.77). Coping strategy Factor 3 was labeled Planful Role Organization because it appeared to reflect a planful, organized approach to dealing with strains between roles. Factor scores were then generated for each married woman 88 for each factor using the regression method, which resulted in three unit-normal distributions of scores for all subjects on the three coping strategy factors. MANOVA was then utilized to determine if male-and female-dominant occupational groups significantly differed on the coping strategy factors, using the three unit-normal distributions of factor scores as the data set. The three dependent variables in this MANOVA were regressed factor scores on Factor 1 (Balanced Role Redefinition), Factor 2 (Direct Role Redefinition), and Factor 3, Planful Role Organization. Covariates were included in the MANOVA in order to statistically control for the confounding effects of demographic variables and personality dimensions upon which male and female-dominant occupational groups had been shown to significantly differ in this study. Covariates included were: combined salary; years employed in present occupation; hours worked in a typical week; masculinity; and femininity. Male-dominant, compared to female-dominant occupational groups, differed significantly on the coping strategy factors used as dependent variables F(3,117)-3.20, p<.02. Since the multivariate F test was significant, separate univariate F tests were performed to determine significant differences between male-and female-dominant occupational groups on each of the three coping strategy factors. Married women employed in female-dominant occupations (M = .046) scored higher than married women employed in male- dominant occupations (M - -.048) on Balanced Role 89 Redefinition £(1,ll7)-.32, p>.05. (see Table 19). Married women employed in male-dominant occupations (M - .31) reported using Direct Role Redefinition significantly more than married women employed in female— dominant occupations (M - -.28) £(l,ll7)=4.63, p<.03 (see Table 19). Married women employed in female-dominant occupations (M = .19) scored significantly higher than married women employed in male-dominant occupations (H - -.18) on Planful Role Organization £(1,117)-4,95, p<.02 a (see Table 19). ‘ 5222:5551; 2 This hypothesis stated that women employed in female- dominant careers would not differ from women employed in male-dominant careers in "Personal Attitudes Towards Roles," Attitudes of Others Toward My Career," "Role Strain," and "Satisfaction With Coping Strategies." Contingency tables and chi square tests of significance were used to assess the relationships between male- and female-dominant occupational affiliation and "Personal Attitudes Towards Roles" (see Tables 20 to 24). The crosstabulation of a joint frequency distribution of cases according to the two classificatory variables (i.e., occupational affiliation and the four catagorical response options to survey questions 9, 10, ll, 12, and 13) made it possible to determine if the two variables were statistically independent. Table 19 Means and Standard.Deviations of Coping Strategy Distributions 90 for Male and Female-Dominant Groups Occupational Status Male-dominant group Balanced Role Redefinition '3 lg lg Direct Role Redefinition * l3 '8 l3 Planful Role Organization ** -.O48 .820 60 .314 .906 60 -.184 .780 60 .046 .889 64 -.287 .993 64 .192 .871 64 Female-dominant group 91 Male and female-dominant occupational status was not significantly related to couple's decisions about having children (see Table 20). Women's attitudes towards child care arrangements were not related to male and female- dominant career status (see Table 21). Women did not differ as a consequence of male or female-dominant occupational status in their attitudes regarding the effects of employment on children (see Table 22). Male and female-dominant occupational affiliation was significantly related to attitudes toward household chores x2(3, u - 136) - 20.07, p<.0002 (see Table 23). A large percentage of married women employed in male-dominant occupations reported they hired outside help to take care of all the major chores (40.6 %), and a large percentage of married women employed in female-dominant occupations shared with their family responsibility and concern about getting chores done (44.4 %) (see Table 23). Male and female-dominant occupational affiliation was not related to women's attitudes toward family, home and career roles (see Table 24). ANCOVA was used to determine if male-dominant, compared to female-dominant occupational groups, significantly differed on "Attitudes of Others Toward my Career," "Role Strain," and "Satisfaction With Coping Strategies." Covariates included in each of the ANCOVA reported in this study were: combined salary; years employed in present Table 20 Couple's Decision Regarding Children Count Row Percent Column Percent Now have Hope to Don't Don't child have know want Occupational status 55 7 4 4 dominant 44.4 63.6 80.0 50.0 69 4 1 4 Female- 88.5 5.1 1.3 5.1 dominant 55.6 36.4 20.0 50.0 Column 124 11 5 8 total 83.8 7.4 3.4 5.4 Our decision relative to children is that: a) We now have a child (or children) b) Although we do not yet have children, we hope to have them in the future c) We don't know if we want children d) We don't want children Row total 70 47.3 78 52.7 148' 100.0 93 Table 21 Attitude Towards Child Care Arrangement Count Row Percent Column Percent Feel Feel Ambiv- Doesn't Row fine guilty alent apply total Occupational status 31 2 20 16 69 Male- 44.9 2.9 29.0 23.2 47.9 dominant 50.8 22.2 45.5 53.3 30 7 24 14 75 Female- 40.0 9.3 32.0 18.7 52.1 dOminant 49.2 77.8 54.5 46.7 Column 61 9 44 30 144 total 42.4 6.3 30.6 20.8 100.0 — 3.04 df = p < .38 When someone else takes care of the children ( or took care of them, if they are grown): a) I feel fine about the care I have arranged for my children ’ b) I feel guilty and anxious about not being with them c) I have ambivalent feelings about the situation d) Not applicable .Table 22 Attitude Regarding Effects of Employment on Children Count Row Percent Column Percent Only Mostly Same , Doesn't positive positive amount apply Occupational status 8 32 11 17 Male- 11.8 47.1 16.2 25.0 ,dominant 47.1 45.7 36.7 58.6 9 38 19 12 Female- 11.5 48.7 24.4 15.4 dominant 52.9 54.3 63.3 41.4 Column 17 70 30 29 total 11.6 47.9 20.5 19.9 = 2.89 df = p < Row total 68 46.6 78 53.4 146 100.0 Do you feel that your employment has had any good/bad effects on your children? a) It has had only positive effects (e.g. they are more independent, etc.)ture b) It has had more positive effects than negative effects c) It has probably had about the same amount of positive and negative effects (e.g. they are more independent and also somewhat neglected) d) Not applicable 95 Table 23 Attitude Towards Household Chores Count Row Percent Column Percent Worry Hired Family Do my Row a lot help helps best total Occupational status 7 26 16 15 64 Male- 10.9 40.6 25.0 23.4 47.1 dominant 36.8 81.3 33.3 40.5 12 6 32 22 72 Female- 16.7 8.3 44.4 30.6 52.9 dominant 63.2 18.8 66.7 59.5 Column 19 32 48 37 136 total ,14.0 23.5 35.3 27.2 100.0 X = 20.07 df 3 p < .0002 My feelings about household chores can best be described as follows: a) I worry a great deal about managing to get the chores done b) I have hired outside help to take care of all the major chores c) My family and I share responsibility and concern about getting chores done d) I do my best and don't worry if the chores don't get done - Table 24 Attitude Toward Family, Home, and Career Roles Count Row Percent Column Percent Occupational status Male- dominant Female- dominant Column total 96 Career Family Home- Impos- comes comes comes sible first first first to rate 2 43 0 25 2.9 61.4 .0 35.7 50.0 44.3 .0 55.6 2 54 1 20 2.6 70.1 1.3 26.0 50.0 55.7 100.0 44.4 4 97 1 45 2.7 66.0 .7 30.6 = 2.47 df = p < Row total 70 47.6 77 52.4 147 100.0 If I had to rate the relative importance of my family, home, and career roles, I would say that: a) My career definitely comes first b) My family definitely comes first c) My homemaker role definitely comes first d) It is impossible to rate these roles 97 occupation; hours worked in a typical week; masculinity; and femininity. Six of the eight items included in "Attitudes of Others Toward my Career" were used to construct a Support scale (survey items 14, 15, 16, 19, 20, and 21) that was analyzed with ANCOVA. Women employed in female-dominant occupations (M a 3.34) reported receiving more support, as assessed by the Support scale, than women employed in male-dominant occupations (M - 3.29) £(l,125)=2.9, p>.05. The two survey items (items 17 and 18) that were not included in the Support scale because they appeared to assess perceived competitiveness with one's spouse were analyzed separately with ANCOVA. Item 17 asked the degree the respondent agreed with the statement "My husband and I have often felt in competition in terms of careers." A high score on items 17 and 18 indicates high competitiveness. Male- dominant occupational affiliation (M - 1.67) was associated with higher competitiveness than female-dominant occupational affiliation (M - 1.45) £(l,135).65, p>.05. Item 18 asked the degree "My husband and I sometimes compete in areas not related to our careers (who is‘a better parent, tennis player, etc.)." Married women employed in male-dominant dominant occupations (M - 1.98) were more likely to compete with their spouses than were married women employed in female-dominant occupations (M - 1.74) £(l,136)-.92, p>.05. Six of the seven survey items that assessed "Role Strain" were used to construct a Role Strain scale (items 22, 98 23, 24, 25, 27, and 28) that was analyzed with ANCOVA. Male- dominant occupational status (M - 3.17) was associated with significantly higher levels of role strain than female- dominant occupational status (M = 2.91) E(l,l35)-4.93, p<.02. Item 26, which asked the degree the respondent agreed with the statement that "I often feel a strain in trying to fulfill my parental obligations" was analyzed separately with ANCOVA. Women employed in male-dominant occupations "53 (M - 2.93) reported feeling more strain than women employed in female-dominant occupations (M - 2.75) E(l,llO)-1.05, p>.05. The four survey items that assessed "Satisfaction With Coping Strategies" (items 45 to 48) were analyzed separately with ANCOVA. Item 45 asked the degree the respondent was satisfied with coping strategies, i.e. "How satisfied are you with the way you have dealt with possible role strains in your life?" A high score on the four items assessing satisfaction indicates a high level of satisfaction. Female- dominant occupational affiliation (M =2.9) was associated with higher levels of satisfaction with coping strategies, as assessed by item 45, than male-dominant occupational affiliation (M I 2.83) fi(l,136)-1.03, p>.05. Item 46 asked, "Overall, how satisfied do you feel with your career?" Women employed in male-dominant careers (M - 3.2) were slightly more satisfied with their careers than were women employed in female-dominant occupations (M - 3.17) I(1,135)-.68, p>.05. 99 Item 47 asked, "Overall how satisfied do you feel as a mother?" Women employed in male-dominant occupations (M - 3.5) were slightly more satisfied with their parental roles than were women employed in female-dominant occupations (M I 3.45) £(1,112)-.45, p>.05. Item 48 asked, "Overall how Satisfied do you feel as a marriage partner?" Female-dominant occupational affiliation (M - 3.22) was associated with higher levels of marital satisfaction than male-dominant occupational status (14 - 3.14) 2(1,135)-.24, p>.05. Epgp Mpg Apalyses A series of multiple regression equations were undertaken in order to address the issue of whether different coping strategies are more efficacious for individuals in the four different occupations sampled in this study, and for the combined group of individuals, irrespective of occupational status. The first series of regression analyses sought to determine how coping strategies are related to subjective perceptions of emotional states, i.e., satisfaction and role strain. The relationships between occupational affiliation and the emotional outcomes associated with different coping strategies was viewed as important since it could then serve as the basis for assessment and intervention. . The second series of regression analyses sought to identify the situational/demographic and personality variables of individuals in each of the four occupations that 100 predict use of the coping strategies. Since different coping strategies might be associated with greater likelihood of different types of role strain within different occupations, these predictor variables would serve as risk factors that predict negative emotional outcomes. Situational/demographic and personality variables that predict coping strategies found to be related to satisfaction, however, could serve as the basis for prescriptive interventions. . Stepwise multiple regression was used in the following analyses. Separate regression analyses were completed for each occupation, and for all individuals irrespective of occupational status. In the first series of regression analyses, separate regression analyses were performed with each item assessing satisfaction (total of four items) and each item assessing role strain (total of seven items) serving as dependent variables, and factor scores on Balanced Role Redefinition, Direct Role Redefinition and Planful Role Organization serving as independent variables in the regression equations. A total of 55 separate regression analyses were performed in this first series of statistical tests (5 groups X 11 dependent variables). Results are reported only for significant relationships, e.g., predictor variables significant at the .05 level or lower. Four survey items assessed "Satisfaction With Coping Strategies" (survey items 45, 46, 47, and 48). Significant relationships between coping strategy factors and satisfaction were obtained only with survey item 45, which 101 asked, "How satisfied are you with the way you have dealt with possible role strains in your life?". Planful Role Organization significantly predicted satisfaction with dealing with possible role strains for the combined group, elementary teachers, and for university faculty (see Table 25). Seven survey items assessed "Role Strain' (survey items 22, 23, 24, 25, 26, 27, and 28). Significant relationships were obtained between coping strategy factors and all seven role strain items (see Table 26). Role strain survey item 22 stated, "I often feel a strain in trying to fulfill both home and career obligations." Planful Role Organization was negatively related to role strain, while Direct Role Redefinition positively predicted role strain for the combined group (see Table 26). Planful Role Organization was negatively related to strain between home and career roles for elementary teachers: that is, elementary teachers who used Planful Role Organization had lower levels of strain between home and career roles. Role strain item 23 stated, "I often feel a strain between my personal needs and career obligations." Direct Role Redefinition predicted conflict between personal needs and career obligations for the combined group, for physicians and for registered nurses (see Table 26). Role strain item 24 stated, "I often feel a strain between my personal needs and home obligations.” Direct Role 102 .umaoH Ho H0>0H mo. as» u6 us6cflu6sva mH 0Hn6an6> uouoHcoum ms» £0623 s6 math6s6 sonmchmH 0HmHuHsE chsHm 6 no muHsmmu as» musommscms 660506 moHHuso o>os6 ms» no somm soHu6uHsmmuo oHom Hsus6Hm u HsusmHm "muoz 65. 666. 6H.6 66. H6666H6 suHoouu suHmuo>Hos 6H. mHo. 66.5 mm. Hsus6Hm Hmsosou assusosmHm 66. 6666. 66.6 65. HsucmHm acoHuoooooo HH< m m w 6uom 0Hn6HH6> msouc 5 =0uHH uso» s6 msH6Hum oHom oHnHmmom squ uH6mo 0>6m so» >63 mss squ soHu06umHU6m: msHuchon msoHu6svm sonmmummm 0HmHuHsz omH3Qmum s6 muouonm hooumuum msHmoc mm OHQMB 103 Redefinition predicted conflict between personal needs and home obligations for physicians, while Planful Role Organization was negatively related to this type of role strain for elementary teachers (see Table 26): that is, elementary teachers who used Planful Role Organization had lower levels of strain due to conflicts between personal needs and home obligations. Role strain item 25 stated, "I often feel a strain because of the lack of time to do all the things I want to do." Planful Role Organization was negatively related to strain attributed to lack of time for the combined group (see Table 26); that is, individuals who used Planful Role Organization experienced less strain due to time limitations. Role strain item 26 stated, "I often feel a strain in trying to fulfill my parental obligations.” Direct Role Redefinition predicted strain in the parental role for the combined group (see Table 26). Role strain item 27 stated, "I often feel a strain in trying to fulfill my obligations as a marriage partner." Planful Role Organization was negatively related to strain in the marital role for the combined group, for elementary teachers, and for university faculty (see Table 26). Role strain item 28 stated, "I often feel a strain in trying to fulfill my career obligations." For the combined group, Planful Role Organization was negatively related to strain in the career role, while Direct Role Redefinition was positively related to career role strain. Direct Role 104 Table 26 Coping Strategy Factors in Stepwise Multiple Regression Equations Predicting Aspects of Role Strain Role strain item Group Variable Beta 5 p 32 "...home and career..." (22) All occupations Planful -.27 -3.21 .001 .12 Direct .19 2.30 .02 Elem. teacher Planful -.43 -2.83 .007 .19 "...needs and career..." (23) All occupations Direct .30 3.54 .0006 .09 Physician Direct .43 2.49 .01 .19 Reg. nurse Direct .43 2.52 .01 .19 "...needs and home..." (24) Physician Direct .46 2.66 .01 .21 Elem. teacher Planful -.36 -2.31 .02 .13 "...lack of time..." (25) All occupations Planful -.18 -2.12 .03 .03 "...parental obligations..." (26) All occupations Direct .18 1.99 .04 .03 "...marriage partner..." (27) All occupations Planful -.23 -2.62 .009 .05 Elem. teacher Planful -.43 -2.78 .008 .18 Univ. faculty Planful -.37 -2.30 .02 .14 "...career obligations..." (28) All occupations Planful -.22 -2.61 .01 .09 Direct .17 2.05 .04 Physician Direct .37 2.07 .04 .14 Note: Balanced - Balanced Role Redefinition; Direct - Direct Role Redefinition; Planful - Planful Role Organization 105 Redefinition predicted strain in the career role for physicians (see Table 26). The second series of regression analyses assessed the degree situational/demographic and personality variables predicted use of coping strategies. The three coping strategy factors served as separate dependent variables in the regression analyses. Situational/demographic variables entered as independent variables included: age, years worked in present occupation, combined salary, hours worked in a typical week, length of marriage to present spouse, number of children living in the home, and Support, as assessed by the Support scale previously described. Personality variables entered as independent variables included: Masculinity and Femininity, as assessed by the PAQ, and social competence and social self-esteem, assessed by the TSBI. A total of 15 regression equations were computed (5 groups X 3 dependent variables). Only significant relationships between predictor variables and dependent variables are reported. Balanced Role Redefinition was predicted by social competence and social self-esteem, assessed by the TSBI, for the combined group. For physicians, Balanced Role Redefinition was predicted by the presence of support (see Table 27). Direct Role Redefinition was predicted by the absence of support, total hours worked, and social competence and social self-esteem for the combined group. The absence of support also predicted Direct Role Redefinition for university Table 27 106 Demographic and Personality Variables in Stepwise Multiple Regression Equations Predicting Use of Coping Strategies Coping Strategy Factor 2 Group Variable Beta 5 p B Balanced Role Redefinition All occupations TSBI .19 2.16 .03 .04 Physician Support .43 2.38 .02 .18 Direct Role Redefinition All occupations Support -.30 -3.37 .001 .14 Hours .21 2.40 .018 TSBI .21 2.33 .021 Physician Masculinity .51 3.01 .005 .27 Reg. nurse Salary .43 2.47 .02 .19 Univ. faculty Support -.39 -2.33 .02 .16 Planful Role Organization All occupations Hours -.22 -2.40 .017 .07 Masculinity .19 2.01 .046 Univ. faculty Masculinity .53 3.39 .001 .28 Note: TSBI a Texas Social Behavior Inventory 107 faculty. The use of Direct Role Redefinition was predicted by masculinity in physicians, and by a higher combined salary for registered nurses (see Table 27). Planful Role Organization was negatively related to number of hours worked for the combined group, but positively predicted by masculinity both the combined group and for university faculty (see Table 27). 21mm Hypothesis I stated that male-dominant, compared to female-dominant occupational groups, would differ in levels of instrumentality, expressiveness, and social competence and social self-esteem. Married women employed in male-dominant occupational groups (physicians and university faculty) were expected to have higher levels of masculine, instrumental traits than married women employed in female-dominant occupational groups (registered nurses and elementary school teachers). Married women employed in female-dominant groups were expected to have higher levels of feminine, expressive traits than married women employed in male-dominant occupational groups. This hypothesis was partially supported by the results of this study. Married women employed in male- dominant occupations percieved higher levels of masculinity, asssesed by the Masculinity scale of the Personal Attributes Questionaire (PAQ), in their self-concepts than did married women employed in female-dominant occupations. Conversely, married women employed in female-dominant occupations reported that they incorporated higher levels of femininity, 108 assessed by the Femininity scale of the PAQ, than did married women employed in male-dominant occupations. However, married women employed in male-dominant occupations did not significantly differ from married women employed in female- dominant occupations in levels of social competence and social self-esteem, assessed by the Texas Social Behavior Inventory (TSBI). The results of this study are consistent with previous research that has found male-dominant occupational status positively related to masculine, instrumental traits and female-dominant occupational status positively related to feminine, expressive characteristics in women (Lemkau, 1983, 1979). This research supports previous research (e.g., Lemkau, 1983, 1979) by demonstrating with a contemporary sample of married women that male-dominant occupational affiliation is associated with higher levels of masculinity than female- dominant occupational status, while female-dominant occupational affiliation is associated with higher levels of femininity than male-dominant occupational affiliation. Previous researchers extrapolated from the Cattell 16PF what they considered to be instrumental or expressive traits in women (Lemkau, 1983), or failed to obtain women's self- perceptions of their instrumentality and expressiveness in a role-free context (Lemkau, 1979). This study extends previous research by demonstrating the positive relationship between male and female-dominant occupational affiliation, 109 and masculinity and femininity, respectively, with the Personal Attributes Questionnaire, a sex role inventory that is specifically designed to assess individual's self- perceptions of instrumental, masculine and expressive, feminine personality traits (Spence, Helmreich, & Stapp, 1975). Contrary to Hypothesis I, women who differed in male versus female-dominant occupational status did not also differ in overall levels of social competence and social self-esteem, assessed by the TSBI. Although women employed in male-dominant occupations were slightly higher than women employed in female-dominant occupations in social competence and social self-esteem, this difference did not reach significance. A partial explanation for this finding is suggested by the observation that social competence and social self-esteem, assessed by the TSBI, is positively related to both masculinity (£3.83) and femininity (;-.30), assessed by the PAQ, in college women (Spence, Helmreich, & Stapp, 1975). Thus, the TSBI is positively related to both masculinity and femininity. Male-dominant, compared to female-dominant occupational group differences on personality dimensions appear to be most salient on the PAQ, which assesses masculinity and femininity separately, than upon the TSBI, which measures a personality dimension that is related to both masculinity and femininity. Hypothesis II stated that women employed in male- dominant careers would not differ from women employed in 110 female-dominant careers in their use of coping strategies. Use of coping strategies was assessed by asking women the degree they used Hall's (1972) 16 coping strategies. Based upon Hall's (1972) conclusion that the 16 coping strategies could be reduced to three types of role behavior, e.g., structural role redefinition, personal role redefinition, and reacive role behavior, common factor analysis was used to identify three coping strategy factors. Occupational group differences upon the three coping strategy factors was then assessed with MANOVA, using variables upon which male- dominant, compared to female-dominant occupational groups, differed in this study as covariates, e.g., combined salary, years employed in present occupation, hours worked in a typical week, masculinity and femininity. The rationale for selecting the null hypothesis of no differences in coping strategies between male-dominant, compared to female-dominant occupational groups, was that differences in coping strategies would be accounted for by confounding variables that included both situational/demographic and personality components that would be controlled for by using covariates in the MANOVA. Coping strategy Factor 1 was most strongly identified by coping strategies identified by Hall (1972) as reflective of personal role redefinition. These coping strategies, followed by their respective rotated factor loadings, were: "I have developed new attitudes to reduce conflicts...." (.67): "I have established rules and priorities for dealing with 111 roles...." (.65): and "I have reduced my standards within certain roles...." (.49). These coping strategies appear to correspond to a coping style that functions to control the meaning of the problem (Pearlin & Schooler, 1978), and appear to be examples of emotion-focused coping (Folkman & Lazarus, 1930). Two coping strategy items that identify coping strategy Factor 1, however, are examples of Hall's (1972) structural role redefinition, e.g., "Family members share household tasks with me" (.32): and "Family members help me resolve conflicts between my roles" (.30). These strategies appear to be oriented towards soliciting social support in an effort to solve role conflicts, which involves both problem-focused and emotion-focused coping (Folkman & Lazarus, 1985). . Two coping strategies have negative loadings on coping strategies Hall (1972) identified as reactive role behavior, e.g.: "I have no conscious strategies for dealing with conflict between roles" (.-.38): and "I attempt to meet all role demands by doing everything people expect of me" (-.35). The former strategy suggests that coping strategy Factor 1 reflects a conscious pattern of behavior, while the latter coping strategy, considered with the other coping strategies that identify this factor, implies an adjustment to stressful role demands that balances personal interests with role demands. Coping strategy Factor 1 was labeled Balanced Role Redefinition because it is most strongly identified by coping strategies that exemplify personal role redefinition, and 112 uses social support (including both problem-focused and emotion-focused coping) to adjust to stressful role conflicts. Coping strategy Factor 2 was identified by coping strategies Hall (1972) labeled structural role redefinition, e.g.: "I have eliminated certain activities within roles...." (.96): and, "I have hired outside help to assist me with chores" (.37). These coping strategies appear to reflect planful problem-solving (Folkman, Lazarus, Dunkel-Schetter, DeLongis, & Gruen, 1986) as well as a direct action approach (Pearlin & Schooler, 1978) to solving stressful role conflicts. Coping strategy Factor 2 was labeled Direct Role Redefinition to reflect these converging influences. Coping strategy Factor 3 was most strongly identified by a coping strategy that reflects Folkman and Lazarus' (1985) notion of planful problem-solving. However, it is also defined by an example of Hall's (1972) reactive role behavior, e.g., "I tend to schedule and organize activities very carefully to maximize my output" (.78). And further, the other two coping strategies identifying coping strategy Factor 3 are examples of Hall's personal role redefinition. "I consider my personal interests very important...." (.41) suggests an agentic focus in solving role conflicts, while "I keep my roles totally separate...." (.30) suggests that roles are compartmentalized in a mixture of instrumental and internal planned attempts to solve role conflicts. Coping 113 strategy Factor 3 was labeled Planful Role Organization to reflect these influences. In summary, coping strategy Factors 1, 2, and 3 were labeled Balanced Role Redefinition, Direct Role Redefinition, and Planful Role Organization, respectively. Balanced Role Redefinition was defined by a mixture of Hall's (1972) three methods of solving role conflicts, and thus does not clearly support his notion that the 16 coping strategies reflect three separate styles of solving role conflicts, e.g., structural role redefinition, personal role redefinition, and reactive role behavior. Planful Role Organization is similarly defined by more than one of Hall's (1972) three coping styles. Direct Role Redefinition is composed of two of the six coping strategies Hall (1972) identified as structural role redefinition, and thus lends partial support to his idea that the 16 coping strategies can be reduced to three coping styles. The finding that male-dominant, compared to female- dominant occupational groups did not differ on coping strategy Factor 1, Balanced Role Redefinition, may be partially due to the balance of both emotion-focused and problem-focused coping reflected in this coping dimension. Individuals are more similar in their coping than they are different, in that they use a mixture of both problem-focused and emotion-focused cOping (Folkman & Lazarus, 1980). This' coping strategy factor also appears to be sensitive to social support in the family for solving role conflicts, as well as 114 social competence necessary for soliciting help from family members in solving role conflicts. To the extent that social competence overlaps the notion of coping competence (Wrubel, Benner, & Lazarus, 1981), this coping strategy may be reflective of a general coping strategy dimension upon which individuals who differ in respect to male-dominant, compared to female-dominant occupational status, are more similar in their coping than they are different. Married women employed in male-dominant occupations reported using coping strategy Factor 2, Direct Role Redefinition, to a greater extent than married women employed in female-dominant occupations. Although demographic and personality dimensions upon which occupational groups differ have been statistically controlled in this study, women employed in male-dominant groups, compared to women employed in female-dominant groups, reported significantly higher levels of role strain in this study, and lower levels of support. Neither role strain nor support were included as covarites in the MANOVA. It may be that higher levels of role strain, combined with characteristics of male-dominant occupational environments that may include lower levels of support, contribute to married women physicians' and university faculty members' greater use of Direct Role Redefinition coping, e.g., eliminating activities within roles and hiring outside help to assist with chores. Married women employed in female-dominant occupations reported greater use of coping strategy Factor 3, Planful 115 Role Organization, than married women in male-dominant occupations. Significantly lower levels of role strain, combined with slightly higher levels of support may create occupational and home environments that allow individuals the luxury of scheduling and organizing activities carefully, and keeping roles totally separate (e.g., by not bringing work problems home or home problems to work). In summary, Hypothesis II was partially supported by a lack of occupational group differences on coping strategy Factor 1, Balanced Role Redefinition. The finding of occupational group differences on coping strategy Factor 2, Direct Role Redefinition, and coping strategy Factor 3, Planful Role Organization, suggests that occupational groups may differ in their use of coping strategies as a consequence of variables related to home and occupational environments that were not assessed in this study. Hypothesis III stated that women employed in male- dominant occupations would not differ from women employed in female-dominant occupations in career attitudes, role strain, and satisfaction with coping strategies. Individuals differed as a consequence of occupational status on only one item assessing "Personal Attitudes Towards Roles," e.g., attitudes toward household chores. The finding that married women employed in male-dominant occupations hired outside help to take care of all the major chores is consistent with the finding that women with higher incomes were most likely to hire outside help to take care of major 116 chores (Gray, 1980). Women employed in male-dominant occupations in this study had significantly higher joint incomes than women employed in female-dominant occupations, and thus could be expected to be more likely to hire outside help. Conversely, women employed in female-dominant occupations may share responsibility and conern about getting chores done with their families because they are less able to afford to hire outside help. The married women in this study did not differ as a consequence of occupational status in "Attitudes of Others Toward my Career," which was assessed by a Support scale, and two separate items assessing competition with one's spouse. Previous research has not directly questioned whether women employed in male-dominant occupations differ from women employed in female-dominant occupations in levels of support. The relatively moderate level of internal consistency within the Support scale used in this study suggests that support from spouses, friends, colleagues and parents may be a less unitary psychological dimension than other dimensions assessed in this study, such as role strain and social competence. It may be important to obtain reliable assessments of support perceived to be received separately from spouses, friends, colleages and parents in order to determine if women differ on these dimensions as a consequence of male-dominant or female-dominant career status. 117 Women employed in male-dominant occupations reported significantly higher levels of role strain than women employed in female-dominant occupations. Several factors may contribute to this finding. Individuals are buffered against the debilitating effects of stress by support available in their home and occupational environments (Gentry & Kobasa, 1984). Support from spouses is positively related to women's success in solving home-work role conflicts (Reardon, 1982: Gray, 1980), while role strain in working women is partially predicted by the degree of conflict within the family (Anderson-Kulman & Paludi, 1986). Although the difference in support received by women in male-dominant versus female- dominant occupational groups was not significant, women employed in male-dominant occupations, compared to women employed in female-dominant occupations in this study, reported slightly higher levels of competition with spouses, and slightly lower levels of support from spouses, friends, parents and colleagues. Thus, it may warrant further study to determine whether women employed in male-dominant occupations that demand significantly greater time committments are made more vulnerable to the experience of role strain by receiving slightly lower levels of support, compared to women in female-dominant occupations. Finally, role strain in working women has been found to be significantly lower at the plateau level of career development than at the establishment level (Edwards, 1983). Women in the male-dominant occupational group had worked 118 significantly fewer years than women in female-dominant careers. Although years worked in present position was statistically controlled by including this variable as a covariate in the ANCOVA analysis, the establishment of a career may impose sources of stress that were not assessed in this study, including job insecurity and self-imposed high performance standards. Although the difference was not significant between women employed in male-dominant occupations and women employed in female-dominant occupations on role strain in trying to fulfill parental obligations, women in male- dominant occupations scored higher than women in female- dominant occupations on this variable. Since these results were in the same direction as the significant results between occupational groups on the Role Strain scale, this finding may be a consequence of a low level of power in this statistical test, i.e., the reliability of a single survey item cannot be assessed, and is therefore questionable. Perhaps role strain that is a consequence of the role conflicts related to parental obligations could be reliably assessed with several questions in future research. The finding that women employed in male-dominant occupations did not differ from women in female-dominant occupations in their "Satisfaction with Coping Strategies" is consistent with Hypothesis III. However, the finding that the three items assessing satisfaction (e.g., with role strains, career, and as a marriage partner) did not reach an 119 acceptable level of internal consistency to be treated as a single scale suggests that women may perceive satisfaction with these separate sources differently. Since the reliability and precision of single items assessing satisfaction cannot be determined, future research should use scales that uses multiple questions to assess each source of satisfaction separately and thus create separate subscales for each source of saisfaction. In summary, Hypothesis III was partially supported. The finding of significant occupational group differences on Role Strain, which had the highest internal consistency of the scales constructed from this amended version of Gray's (1980) survey, suggests that the finding of no differences between male-dominant and female-dominant occupational groups on Support and the separate items assessing "Satisfaction with Coping Strategies" should be interpreted cautiously. If higher levels of internal consistency were obtained on scales assessing support and satisfaction, female-dominant and male- dominant occupational group differences could be expected to emerge as a consequence of greater power (Cohen, 1969) if occupational group differences do indeed exist. Egg; hQQ analyses; The first set of multiple regression analyses sought to determine how coping strategies are related to subjective perceptions of satisfaction and role strain so that the emotional outcomes associated with coping strategies could serve as the basis for assessment and intervention. A set of regression analyses was then performed 120 to identify situational/demographic and perosn variables that predict use of different coping strategies. In line with Lazarus and Launier's (1978) notion that coping strategies must be viewed as transactions between persons and environments, the second set of regression analyses sought to identify the situation and person variables common to individuals in all of the occupations sampled in this study , that predicted use of coping strategies, as well as r- identifying the situation and person variables unique to I married women in each of the four occupations that predicted use of coping strategies. Planful Role Organization predicted satisfaction with dealing with role strains for the combined group of the four occupations sampled in this study, elementary teachers and university faculty (see Table 25). This suggests that satisfaction with dealing with role strains is positively related to transactions between the person and environment that allow the individuals in this study, regardless of occupation, to organize their roles so as to maximize their output and to devote energy to personal interests by compartmentalizing roles. The ability to engage in Planful Role Organization is predicted by working fewer hours and by masculinity for working women, irrespective of occupation (see Table 27). For university faculty, masculinity accounts for 28 percent of the variance in this equation. These results indicate that when the individuals sampled in this study are considered together, working fewer hours each week 121 is a situational event that allows individuals with higher levels of masculinity to engage in Planful Role Organization. As a consequence of this transaction between the person and environment, individuals are satisfied with how they deal with the role strains in their lives. When university faculty are considered by themselves, the transaction is more heavily weighted on the person side of the equation, since , masculinity alone accounts for the variance in the equation r1 (see Table 27). ' Strains between home and career obligations were ‘ negatively related to Planful Role Organization, but positively predicted by Direct Role Redefinition for working women regardless of occupation (see Table 26). Although the transaction between person and environment is easily understandable for Planful Role Organization, e.g., it is predicted by working fewer hours and having sufficient personal resources in the form of masculine instrumentality, a variety of situation and person variables unique to each occupation predict Direct Role Redefinition (see Table 27). When working women are considered irrespective of occupation, the absence of support, and greater number of hours worked create the situation for individuals with high social competence to engage in Direct Role Redefinition by eliminating activities within roles and hiring help to assist with chores (see Table 27). The transaction is clearly weighted on the person side of the equation for women physicians, however, for whom masculinity accounts for 26 122 percent of the variance. For registered nurses, the situational advantage of having a high combined salary predicts Direct Role Redefinition, while the absence of support weighs the transaction similary with a situational determinant for university faculty (see Table 27). Strains between personal needs and career obligations are predicted by Direct Role Redefinition for individuals regardless of occupation, and for physicians and registered nurses (see Table 26). Physicians and registered nurses are employed as health care providers, which creates a situational context in which health care recipients are demanding of high levels of both personal attention and professional competence (Maslach, 1982). Women physicians who are high in masculinity may instrumentally cope with strains between their personal needs and career obligations by eliminating activities within roles, such as no longer doing all of the cooking, and by hiring outside help. Registered nurses who have sufficient combined incomes may similary cope with conflicts between personal needs and career obligations by Direct Role Redefinition. Thus, Direct Role Redefinition appears to be used by both women physicians and by registered nurses to cope with conflicts between personal needs and career obligations. It may be that women physicians and registered nurses share a common occupational context in which health care recipients are demanding of attention and professional competence which diverts attention and energy away from 123 meeting personal needs. For physicians, the use of this coping strategy is predicted by masculinity, a personality trait, while the use of Direct Role Redefinition is predicted by combined salary for registered nurses, a situation variable. Thus, a common occupational situation may be dealt with by the same coping strategy by both registered nurses and physicians. However, registered nurses appear to depend more upon environmental resources to effectively cope with conflict between personal needs and career obligations, while physicians draw upon their inner resources to cope with this role conflict. Women physicians also experience a strain between their personal needs and home obligations, which is predicted by Direct Role Redefinition (see Table 26). Elementary teachers are protected against strains between personal needs and home obligations by engaging in Planful Role Organization (see Table 26). Since the ability to devote energy to personal interests and plan role activities is predicted by working fewer hours, and by higher levels of masculine instrumentality for individuals regardless of occupation (see Table 27), elementary teachers with sufficient levels of masculinity who work fewer hours appear to reap the benefits of their occupational situation by being protected against strains between personal needs and home obligations. Individuals reported less strain due to a lack of time to do the things they want to do when they also engaged in Planful Role Organization (see Table 26), a coping strategy 124 that requires fewer hours worked on the situation side of the transaction, and higher levels of masculinity for individuals in all occupations (see Table 27). Working women, irrespective of occupation, who were parents felt a strain in trying to fulfill their parental obligations if they also engaged in Direct Role Redefinition (see Table 26). Since Direct Role Redefinition is predicted by the absence of support, working more hours, and by higher levels of social competence for individuals considered collectively in this study (see Table 27), parental role strain appears to be a consequence of an environmental situation defined by less support and working more hours, and by the personal attribute of social competence. Working mothers cope with parental role strain by eliminating activities within roles and hiring outside help. Planful Role Organization protected individuals in all occupations, as well as elementary teachers and university faculty, against the experience of strain in fulfilling obligations as a marriage partner (see Table 26). Working fewer hours and having sufficient levels of masculine instrumental resources define the situation and person variables that predict Planful Role Organization for individuals considered irrespective of occupation. In addition, university faculty appear to report less marital strain when they also incorporate high levels of masculine instrumentality into their self-concepts (see Table 27). Since masculinity also predicts male-dominant occupational _....... a... F 125 status for university faculty (Table 17), women faculty employed in male-dominant fields appear to cope in the occupational environment with a cognitive instrumentality that is also used to guard against the experience of marital role strain in the home environment. Direct Role Redefinition predicted strain in trying to fulfill career obligations for individuals regardless of occupation, and for women physicians considered separately. However, Planful Role Organization protected working women in Wi‘ifl-I' ‘- all occupations against the experience of career strain (see Table 26). Women in all of the occupations sampled in this study who work longer hours are more likely to use Direct Role Redefinition. Women who work fewer hours are more likely to engage in Planful Role Organization (see Table 27). Thus, the experience of career strain appears to be related to the number of hours worked, while the coping strategy used is a consequence of the time spent working each week, as well as other person and environmental variables (see Table 27). Although Balanced Role Redefinition did not predict role strains or satisfaction with coping strategies, it should be noted that the use of this coping strategy is predicted by social competence when individuals are considered collectively, and by the availability of support for women physicians (see Table 27). Since women using Balanced Role Redefinition rely upon family members to share household tasks and to help resolve conflicts between roles, both the availability of support and social competence have face 126 validity as variables on the situation and person sides of transactions between persons and environments that predict the use of this coping strategy. CHAPTER V SUMMARY & IMPLICATIONS The purpose of this study was to obtain a clearer understanding of the problems of and solutions to the role conflicts in a contemporary sample of maried women who differ in that they are employed in either traditional female- ‘ dominant occupations or in nontraditional, male-dominant occupations. In this chapter, a summary of the study is presented, followed by conclusions that can be drawn from this research, limitations of the study, and implications for research, therapy, and social policy. Summary Researchers have attempted to learn how working married women's multiple roles of worker, wife, and possibly parent affect their well being. Although an early position held that multiple roles operated simultaneously for women, and were therefore sources of emotional strain, more recent views point to the positive benefits of increased sources of self- esteem and status derived from multiple roles. A more sophisticated approach is that the emotional outcomes of role strain and satisfaction are a consequence of an ongoing transaction between the person and the environment. In this formulation, individuals who differ in personality choose occupations with different characteristics in which 127 128 preexisting modal behavior patterns are most adaptive. That is, occupations are chosen in which styles of behavior are valued, and are thus potential sources of self-esteem, while behavior patterns are also efficacious in terms of offering control and mastery when adjusting to occupational demands. Several areas of literature were reviewed, including Lazurus and his associates' general model of coping, coping research and theory aimed more specifically at coping strategies used to adjust to role conflicts, variables that predict male or female-dominant occupational choice, variables contributing to life satisfaction, and social competence, viewed as it relates to the concept of coping competence. Coping is viewed as an effortful adjustment to stress, including role strain derived from the conflicts of multiple roles. Individual personality differences that are related to occupational choice, including masculine instrumentality and feminine expressiveness, are seen as influencing the coping strategies chosen by women when dealing with role conflicts. Since personality differences that influence occupational choice may also influence women's choice of coping strategies, perceptions of role strain and satisfaction, these personality differences are confounding variables whose effects upon dependent variables must be statistically controlled in order to learn how women who occupy different occupational enviornments also differ in 129 their choice of coping strategies, and perceptions of well being. This research attempted to a) identify personality variables that differ for women who are employed in occupations that are dominated by women or men, b) identify demographic variables that differ for women employed in male-dominant or female-dominant occupations, and c), to statistically control these variables in order to learn if women differ in their use of coping strategies, perceived support, role strain, and satisfaction as a consequence of working in male-dominant or female-dominant occupational environments. An amended version of Gray's (1980) survey was constructed that included the Personal Attributes Questionnaire (PAQ), to assess masculinity and femininity, and the Texas Social Behavior Inventory (TSBI), to assess social competence and social self-esteem. This study was carried out with samples of women in traditional female-dominant fields, e.g., elementary teachers and registered nurses, and women in nontraditional male- dominant fields, e.g., university faculty and physicians. Women recieved mail surveys that were accompanied by a cover letter that explained the research. Two follow-up postcards were sent at one-week intervals after the initial survey to encourage participation. A total of 219 usable surveys were received for an overall response rate of 51 percent. A total of 148 of the 219 usable surveys 130 were completed by married women, and 71 were completed by single women. Married women who completed usable surveys occupied the following occupations: elementary teachers (46), registered nurses (32), university faculty (39), and physicians (31). Statistical analyses were performed only upon the surveys completed by married women. A MANOVA was computed with scores on the Masculinity and Femininity scales of the PAQ, and a single score for the TSBI as dependent variables. Univariate F tests indicated that married women employed in male-dominant occupations were higher in masculine instrumental traits than women employed in female-dominant occupations, while women employed in female-dominant occupations were higher in feminine expressive traits than women employed in male-dominant occupations. Women in male or female-dominant occupational groups did not differ in social competence and social self- esteem. These results were interpreted as both supporting and extending previous research that indicated women who occupy occupational environments dominated by males or females differ in levels of masculine instrumentality and feminine expressiveness. The results of ANOVA computed upon demographic data demonstrated the following statistically significant differences between married women employed in male-dominant and female-dominant occupations: women in female-dominant occupations had been employed longer than women in male- ‘5‘"! 131 dominant occupations: women in male-dominant occupations had higher combined salaries than women employed in female- dominant occupations: women employed in male-dominant occupations worked more hours per week than women employed in female-dominant occupations: and the age of the youngest child of women in male-dominant occupations was less than the age of the youngest child of women in female-dominant occupations. In subsequent relevant statistical analyses, with one exception, personality and demographic variables upon which male-dominant and female-dominant groups differed were included as covariates in order to provide statistical control for the confounding effects of these variables upon dependent variables. The age of the youngest child was not included as a covariate because it would have limited the analyses to women who were mothers. The reliability of separate scales within the survey was computed with Cronbach alpha. The Masculinity and Femininity scales of the PAQ, the TSBI, a Role Strain scale and a Support scale were found to have acceptable levels of internal consistency to serve as construct measures in subsequent analyses. A common factor analysis identified three coping strategy factors underlying Hall's (1972) 16 coping strategies. These were named Balanced Role Redefinition, Direct Role Redefinition, and Planful Role Organization. A MANOVA was computed that included relevant covariates with .w‘i' 132 scores on the coping strategy factors as the three dependent variables. Married women did not differ as a consequence of occupational status in their use of Factor I, Balanced Role Redefinition. However, women employed in male-dominant occupations scored significantly higher than women in female- dominant occupations on Factor II, Direct Role Redefinition, while women employed in female-dominant occupations scored significantly higher than women employed in male-dominant occupations in their use of Factor III, Planful Role Organization. Balanced Role Redefinition appeared to reflect a balance of emotion-focused and problem-focused strategies. Balanced Role Redefinition appears to be a general coping strategy that is predicted by social competence (see Table 27), and thus lends credence to Wrubel, Benner, & Lazarus' (1981) notion that social competence overlaps the concept of coping competence. Since women did not differ in their use of Balanced Role Redefinition as a consequence of occupational status, it appears that women are more similar than they are different in their use of coping strategies. The greater use of Direct Role Redefinition by women employed in male-dominant occupations was interpreted as possible consequences of the higher levels of role strain and lower levels of support reported by women in male-dominant occupations than by women in female-dominant occupations in this study. The greater use of Planful Role Organization by women in female-dominant occupations compared to women in A 1 133 male-dominant occupations was similary interpreted as possible consequences of the lower levels of role strain and higher levels of support reported by women employed in female-dominant occupations than by women in male-dominant occupations. The results of a chi square analysis indicated that women differed as a consequence of occupational status in their attitudes toward household chores. The finding that _ 116‘-, women employed in male-dominant occupations hired outside help to take care of all the major chores was interpreted as l being consistent with the higher incomes of this group of women. The finding that women employed in female-dominant occupations share responsibility and concern about getting chores done with their families was similary interpreted as a consequence of the lower combined incomes of women employed in female-dominant occupations. The results of ANCOVA indicated that women did not significantly differ as a consequence of occupational status in their perceived support from spouses, friends, colleagues and parents in this study. This finding was interpreted as a possible consequence of the relatively moderate internal consistency of the Support scale. The results of ANCOVA demonstrated that women employed in male-dominant occupations experience significantly higher levels of role strain than women employed in female-dominant occupations. This finding was interpreted as possible consequences of several factors, including the lower level of 134 support and higher levels of spousal competitiveness reported by women in male-dominant occupations, and the fewer number of years worked by women employed in male-dominant occupations that may impose sources of stress that were not assessed in this study, including job insecurity and self- imposed high performance standards. The results of ANCOVA indicated that women do not differ ) as a consequence of male or female-dominant occupational Fl status in satisfaction with coping strategies, or :5 satisfaction with their careers, parental roles, or marriage I roles. Since each of the four satisfaction dependent variables was assessed by a single survey question, it was noted that the precision of these satisfaction constructs could not be determined, and it was suggested that these findings should therefore by cautiously interpreted pending the development of reliable scales that assess satisfaction. Post hoc analyses were conducted to determine how coping strategies are related to subjective perceptions of satisfaction and role strain, and to identify situational/demographic and person variables that predict use of different coping strategies. Planful Role Organization predicted satisfaction with dealing with role strains, and was negatively related to six of the seven sources of role strain assessed. Direct Role _ Redefinition was positively related to five of the seven sources of role strain assessed. The results of post hoc analyses demonstrated that the efficacy of these two coping 135 strategies depended upon situation and person variables that were in some cases common to individuals in each of the four occupations sampled, and in other cases determined by situation and person variables unique to different occupations. 92m The finding that women employed in male-dominant ' and female-dominant occupations described themselves as more ‘Cf17 masculine or feminine, respectively, is consistent with the view that individuals select occupational environments in which their behavioral styles are most adaptive, and in which their values and interests are most compatible. Thus, women who perceive themselves as behaving in more masculine, instrumental ways are more likely to select occupations in which a cognitive instrumental approach that rewards inner- directedness is both valued and efficacious in terms of occupational performance. Women who perceive themselves as behaving in more feminine, expressive ways are more likely to select occupations in which a nurturant expressiveness and empathic concern for others is valued and efficacious. It could be argued that occupational environments differ in their requirements for instrumentality and expressiveness, and thus individuals develop instrumental or expressive behaviors as a consequence of occupational demands. However, research demonstrates that choice of male or female-dominant college major for both women and men is predicted by positive relationships between masculinity, femininity, and 136 male or female-dominance, respectively (Strange & Rea, 1983: Harren, Kass, Tinsley, & Mooreland, 1979). Hall's (1972) description of his 16 coping strategies as reflecting structural role redefinition, personal role redefinition, and reactive role behavior was not clearly supported by the factor analysis. The three coping strategy factors that emerged in the factor analysis were substantive in that married women employed in male and female-dominant occupations differed in their use of Direct Role Redefinition and Planful Role Organization, and in that Balanced Role F-EE;7J Redefinition appeared to assess a general coping strategy. To the extent that Balanced Role Redefinition requires social competence, this coping strategy may also assess Wrubel et. al.'s (1981) notion of coping competence. The finding that women did not differ as a consequence of occupational status in their use of Balanced Role Redefinition reflects the mixture of both emotion-focused and problem-focused strategies underlying this coping strategy dimension, and supports the notion that individuals are more similar than they are different in their use of coping strategies (Lazarus & Folkman, 1984). ° Women in the male-dominant occupations sampled in this study reported experiencing higher levels of role strain than women employed in female-dominant occupations. The women in this study did not differ as a consequence of occupational status in their perceptions of satisfaction with their coping strategies, careers, parental or marital roles. These 137 findings suggest that aspects of male-dominant occupational environments that women physicians and university faculty work in create higher levels of role strain than do female- dominant occupational environments. Women employed in male-dominant occupations reported slightly higher levels of spousal competiveness and lower levels of support from spouses, friends, parents and colleagues than women in female-dominant occupations. A possible explanation for the higher role strain of women in male-dominant occupations, but similar levels of satisfaction to women in female-dominant occupations, is that the male- dominant occupational environments occupied by women physicians and university faculty offer social arenas in which these women find fulfillment, e.g., in terms of status, and sources of instrumental mastery, control and self-esteem. However, the demands of these occupations in terms of time and committment may make this group of women more vulnerable to experiencing role strain because less support is available in the occupational environment, and there is less opportunity to obtain spousal support because time and energy are claimed by these demanding professions. Other researchers have pointed to the need to identify coping strategies working women use to manage time, and to determine which strategies are associated with the least amount of role strain (Anderson-Kulman & Paludi, 1986). This study suggests that the number of hours worked each week serves as a potent situation variable that predicts use of a 138 coping strategy that is positively related to the experience of strain in all occupations (e.g., Direct Role Redefinition), and is negatively related to Planful Role Organization, a coping strategy that predicts satisfaction with dealing with role strains, while also offering protection against the experience of some forms of role strain. V The results of post hoc analyses were important because AA .1 “‘5. . the multiple regression analyses identified a) coping ‘ strategies that were significant predictors of satisfaction 1 and role strain, and b) the situation and person variables that predicted use of coping strategies that had either positive or negative emotional consequences. These analyses demonstrated that the efficacy of the coping strategy chosen depended upon a transaction between the person and environment that varied in the weight given to situation or person variables as a consequence of the occupation studied. For example, the use of Planful Role Organization was predicted by masculinity for faculty members, while the use of this coping strategy was predicted by both masculinity and working fewer hours when all occupations were considered together. In some cases the situation and person variables that predicted the use of a coping strategy were common to women in each of the four occupations studied, while in other cases these variables differed for occupations. In this study the situation and person variables differed between occupations in their predictive 139 relationships to coping strategies that were associated with either satisfaction or role strain. This demonstrates that each occupational environment should be considered unique in that it contains a particular set of characteristics that are perceived differently depending upon the personalities and situational constraints (e.g., a lack of social support) and resources (e.g., a high combined salary) of women who occupy particular occupational environments. l Wflmm f)- The findings of this study should be considered in light M of its limitations. First, the validity of the results of the study cannot be generalized beyond the four populations sampled in the study, e.g., married women who were: employed as elementary school teachers in the Lansing School District: registered nurses residing in Ingham County: physicians residing in Ingham County: and Michigan State University faculty employed in male-dominant fields. Since the overall response rate for this study was 51 percent, there is a strong possibility of self-selection bias (Raj, 1972). The individuals who responded to this survey may have perceived it as more relevant to their lives than individuals who did not respond. Individuals who were aware of experiencing higher levels of some sources of strain due to the conflicting demands of multiple roles may have responded in greater numbers than those experiencing less strain. This possibility is supported by the disportionately high number of married women physicians and registered nurses 140 compared to single women physicians and registered nurses who returned completed surveys (see Table 1). Although the survey instrument was an amended version of Gray's (1980) survey, the impact of the revisions upon the validity of the survey are not known. However, the level of internal consistency of scales that assessed the constructs relevant to the study was determined, and gives some indication of the precision of the measurement of constructs assessed in this survey instrument. The major findings of this study have been described in terms of differences between populations of women based on their membership in male-dominant careers and female-dominant careers. However, women in the particular occupations studied also differed in that these male-dominant careers chosen for the study had more education (i.e., doctorates and medical degrees) than was required for the female-dominant careers chosen for the study. Thus, the socioeconomic requirements for entry into these male-dominant and female- dominant careers substantively differed. Therefore, a limitation of this study is that women in male and female- dominant occupations were not matched for education, a primary determinant of socioeconmic status. Thus, differences between women in the male and female-dominant careers studied may be due to differences in education rather than to differences that are attributed to either occupations or to gender dominance of occupations. ‘ :3 “E“ 141 In this study it is assumed that the particular gender dominance within occupations has important psychological consequences for women: and thus, the study is designed to detect these differential effects. It may be, however, that other aspects of the particular occupations chosen could account for these effects. One could argue that the use of male control groups would have enabled a direct test of this assumption and controlled for this factor. It is possible F? that the responses of married men physicians and university ‘ faculty to dependent variables including role strain, coping strategies, and satisfaction would be more similar to their married women colleagues, than to the responses of men in female-dominant occupations (eg., registered nurses and elementary school teachers). Thus, a limitation of this study is that the differences between women in the male and female-dominant occupations studied may be incorrectly attributed to consequences of gender rather than to differences in consequences of occupations. Misha There are several implications of this research that warrent discussion. Women did not differ as a consequence of occupational status on the question that assessed their attitudes toward family, home, and career roles (see Table 24). In fact, women in both male-dominant (61.4 %) and in female-dominant (70.1 %) occupations rated their family role as most important. This supports Baruch et al.'s (1987) notion that married women's family role is the most important 142 source of self-esteem and feminine identity. Baruch et al. (1987) note that family role stressors have been linked to psychological distress in women to a greater extent than work-related stressors, suggesting that vulnerability to stress in other roles and overall well-being are primarily determined by satisfaction within the family role. Since women are socialized to assume primary responsibility for the ' well-being of others, especially children, the extent to ....-. L... which they perceive themselves as adequately meeting this responsibility (the support received from others, ‘ particularly their husbands), may be crucial in determining their emotional and physical well-being (Baruch et a1., 1937). While the women in this study did not statistically differ as a consequence of occupational status on the levels of support they recieved from their husbands, friends, parents and colleagues, women in male-dominant careers reported receiving less support and having more competitive relationships with their spouses. Since women in male- dominant careers reported significantly greater role strain than women in female-dominant careers, it is possible that the statistically nonsignificant differences in overall support and spousal competitiveness had important practical implications in terms of increased role strain for women physicians and women university faculty, compared to elementary teachers and registered nurses. 143 A related issue is determining the importance of the finding that women in male-dominant careers experience greater role strain than women in female-dominant careers. High levels of depression and suicide prevail for women physicians (Pitts, Schuller, Rich, & Pitts, 1979), suggesting that role strain that contributes to women's psychological vulnerability is indeed an important area for study and concern. It is unclear to what extent internal person variables and external environmental variables determine susceptibility to role strain and negative psychological consequences. These issues will be more fully addressed in later parts of this chapter. IEEIQRX The results of this study offer information that may be used by therapists to identify situational factors that put women at risk of experiencing role strain. The clearest finding is that women employed as university faculty and physicians experience more role strain than women employed as registered nurses and elementary teachers. Women who work many hours each week are more likely to use Direct Role Redefinition, e.g., by eleminating activities within roles and hiring outside help, and to experience several sources of role strain. Women who work fewer hours are more likely to use Planful Role Organization and to feel satisfied with their coping strategies. Thus, therapists should question what aspects of individuals' occupational environments, including demands for time and energy, are contributing to 144 role strain, and attempt to help clients arrive at an informed decision regarding how to structure their time. The finding that women employed in male-dominant occupations receive less support and are more competitive with their spouses suggests that this group of women is not buffered from the experience of stress by social support to the same degree as women in female-dominant occupations. Therapists may question whether their clients are psychologically disadvantaged in social environments dominated by men who offer less support than might be available in female-dominant occupational environments, or whether pre-existing socially dysfunctional aspects of masculinity serve as internal barriers that restrict the client's ability to obtain support, and push her to prove herself by working longer and harder than her colleagues. Research Future research should be directed at developing reliable instruments that assess a) support derived from spouses, friends, parents and colleagues, and b), satisfaction with coping with role strains, as well as with career, parental and marital roles. The development of reliable subscales that assess these different dimensions of support and satisfaction would enable researchers to avoid making Type II errors, e.g., concluding that a difference between populations does not exist when it does indeed exist, since greater precision would result in greater power (Cohen, 1969). The development of more powerful measurement 145 instruments would also enable researchers to address the issues posed by Baruch, Biener, and Barnett (1987), by more accurately assessing the complex relationships between the qualitative experience within roles, and emotional outcomes of engaging in multiple roles. Future research may also attempt to utilize the revised Ways of Coping Checklist (Folkman, Lazarus, Dunkel-Schetter, DeLongis, & Gruen, 1986) to learn how women are similar and also different in their coping as a consequence of occupational status. Thus use of the Ways of Coping Checklist would enable researchers to assess how women use the eight coping strategy factors identified by Folkman et al. (1986) to deal with stressful encounters as a consequence of the stakes and resources appraised by women to be relevant to role conflicts. Hall's (1972) list of 16 coping strategies should also be utilized in future research since it would allow researchers to validate their results against the findings of previous researchers whose focus has been upon the role conflicts of working married women. The use of common factor analysis with Hall's (1972) coping strategies, and subsequent MANOVA should also be incorporated into future research in order to identify differences in coping between individuals who are grouped according to classificatory schemes such as occupational status. An anomalous finding in this study was that masculinity predicted the use of coping strategies that had different ’F’e: ‘1 146 emotional consequences for two of the occupations studied. That is, masculinity predicted the use of Direct Role Redefinition for women physicians, a coping strategy that predicted several types of role strain. Masculinity predicted the use of Planful Role Organization in university faculty, a coping strategy that predicted satisfaction with coping strategies. In attempting to untangle the relationships between :1 masculinity, occupation, coping strategies, role strains and “1 satisfaction, it is impossible to arrive at definite 1 conclusions regarding the causal links among variables. For example, the high level of stress experienced by women physicians as a consequence of the demands of the profession may contribute to role conflicts and strain, and a cognitive instrumental approach to solving role strains may be reflected in Direct Role Redefinition. However, researchers who have noted the high level of affective disorders and high suicide rate for women physicians suggest that the antecedent conditions in these women's lives which lead them to select a medical education and career also make them susceptible to affective disorders (Pitts, Schuller, Rich, & Pitts, 1979). Research that investigated the antecedent family experiences of masculine sex-typed women has yielded anomalous findings. Spence and Helmreich (1978) found a pattern of high encouragement of traditional female behavior in the context of low warmth and cognitive involvement from both parents, while Kelly and Worrell (1976) found moderate 147 parental warmth in the context of the encouragement of cognitive competence, and in some cases, parental permissiveness in masculine sex-typed young women. These differences in the family histories of masculine sex-typed women may be reflected in differences in positive and negative components of masculinity in women. Although the positive contributions to self-esteem made by masculinity and femininity have been emphasized in the literature (e.g., Spence et. a1., 1975: Bem, 1977), it has gi£:f“m7 ‘-v also been noted that masculinity and femininity contain negative as well as positive components (Spence, Helmreich, & Holahan, 1979). The socially dysfunctional aspects of masculinity include hostility, arrogance, cynicism, and extreme egocentrism (Spence et. a1., 1979). Masculine sex- typed women score higher than other women, and higher than sex-typed men, in these socially dysfunctional masculine traits (Spence et. a1., 1979). It may be that the Masculinity scale of the PAQ that was used in this study assessed both positive and negative components of masculinity, and that more positive cognitively instrumental components are reflected in the positive relationship between masculinity and Planful Role Organization for university faculty, while more negative socially dysfunctional components are reflected in the positive relationship between masculinity and Direct Role Redefinition for women physicians. It is possible that the dysfunctional aspects of masculinity, e.g., extreme egocentrism and hostility, make it 148 difficult for some women in male-dominant occupations to obtain the social support needed to buffer the stress of occupational demands and role conflicts. Direct Role Redefinition may be used by some women with high levels of dysfunctional masculine traits because these individuals experience high levels of role strain that is a consequence of an inability to obtain social support. Future research should address this possibility by relating the antecedent family experiences of women employed in different occupations, including medicine, to both the positive and negative components of masculine instrumentality and feminine expressiveness. It would be instructive to ask how women who differ in antecedent life experiences also differ in: their choice of occupation: levels of positive and negative components of masculinity and femininity: their choice of coping strategies: their use of social support: and perceived role strain. Social Policy Researchers have noted that male physicians are more likely to be the products of unhappy childhoods than socioeconomically matched controls in occupations other than medicine (Vaillant, Sobowale, & McArthur, 1972). Physicians involved in direct patient care are most likely to have had unhappy childhoods and to experience poor marriages and high drug use (Vaillant et. a1., 1972). Although the suicide rate among male physicians is high, it is surpassed by the suicide rate of female physicians, which is four times that for white 149 American women of the same age (Pitts et. a1., 1979). Moreover, the level of affective disorders among women physicians is estimated to be 65 percent, and it is possible that both the high suicide rate and high level of affective disorder among women physicians can be traced to antecedent life experiences that lead some women to choose medicine as a career (Pitts et. a1., 1979). The results of the present study demonstrated that women physicians and university faculty experienced higher levels of role strain than women elementary teachers and registered .'kfl__: “1 V nurses. It is unclear whether women employed in these male- dominant occupations fields are less capable of obtaining soical support than women in female-dominant careers, or if less social support is available from male colleagues. It has been noted that physicians are highly susceptible to stress because they are socially isolated in highly demanding occupations (Mawardi, 1983). The social isolation of physicians is attributed both to pre-existing personality components that lead them to select medicine as a career, to highly competitive medical educations, and to the aspects of the occupation itself, including competition with colleagues and the isolation of private practice (Mawardi, 1983). It would be inappropriate to screen out of medical schools individuals who have suffered unhappy childhoods that may make them vulnerable to increased social isolation and stress as professionals, since these early developmental experiences may also make physicians more empathetic and 150 committed (Mawardi, 1983). However, it would be appropriate to create conditions in medical school that foster cooperation rather than competition (Mawardi, 1983). In general, it would also be appropriate to identify the antecedent family experiences of women and men who select occupations in which the risk of stress is high. Likewise, it would be valuable to learn how individuals in these occupations cope with stress including role strain through experimental study. For example, the results of manipulated conditions such as offering or withholding social support, availability of access to external supports, e.g., subsidized child care, tax credits for in-home assistance independent of combined income, and reduced work hours could be causally related to choice of coping strategies and emotional outcomes including role strain and satisfaction for women in different occupations. Findings from this research could then be incorporated into the graduate curriculum of individuals most at risk of social isolation and increased strain in their later professional lives, or to support social policy that addresses needed resources. Preventive interventions, including psychotherapy, could be offered to those most at risk, and possibly, to those who have the most to offer in terms of their empathic concern for others and committment to their careers. Although the benefits that accrue to young women who choose high paying, prestigious nontraditional occupations are emphasized in the literature (e.g., Fitzgerald & Crites, 151 1980), the results of this study demonstrate that women physicians and university faculty who earn significantly more than women registered nurses and elementary teachers do not differ in their satisfaction with their careers, marriages, or parental roles. Since the effects of combined salary, hours worked and length of time employed upon dependent variables including satisfaction and role strain were » statistically controlled, it cannot be concluded that having a higher income does not lead to higher satisfaction. However, women physicians and university faculty experienced significantly more role strain than women registered nurses and elementary teachers, suggesting that nontraditional occupational choice includes negative as well as positive emotional consequences. In order to make informed occupational choices, young women who are in the process of weighing the costs and benefits of different occupations should be made aware of both the emotional costs and benefits women experience in nontraditional occupations. In conclusion, this research has addressed the issue of how women cope with multiple role demands from the general perspective that coping is a consequence of transactions between persons and environments. The importance of masculine instrumentality and feminine expressiveness as person variables that are positively related to male-dominant and female-dominant occupational incumbency, respectively, was demonstrated. Women in male and female-dominant occupations differed in their use of coping strategies when the effects 152 of masculinity and femininity were statistically controlled, suggesting that the efficacy of coping strategies is at least partly determined by differences in male and female-dominant occupational enviornments. The results of post hoc analyses demonstrated that the use of coping strategies depended upon situation and person variables that were in some cases common to individuals in each of the four occupations sampled, and in other cases determined by situation and person variables unique to different occupations. The finding that the . efficacy of coping strategies is determined by both person I and situation variables that vary according to occupational ' environments suggests that future researchers may profitably frame their questions in terms of how different individuals in different occupational environments cope in order to maximize their emotional well-being. APPENDICES 153 Appendix A Letter to Lansing School District Superintendent Dear Doctor Halik: I am writing on behalf of a doctoral student in counseling psychology at Michigan State University, Don Lawson, for whom I chair his dissertation committee. Mr. Lawson's dissertation study deals with role conflicts of married professional women. The study is designed to focus on some of the ways that married women physicians, school teachers, university faculty, and nurses in the Lansing area are dealing with the conflicts that sometimes result from heavy role demands. I am sure you would agree that this is an important and timely topic for which much information is lacking, but is needed in order to assist professional women who choose to combine marriage and occupational roles in coping with the stress of these multiple roles. Your assistance is needed in helping to provide a suitable sample of women elementary school teachers for Mr. Lawson's study. He has proposed to randomly select a total of 120 women from the elementary schools in the Lansing School District. However, we need your permission to do so, as well as a mailing list of women elementary school teachers in the Lansing School District. You have both my and Mr. Lawson's assurance that the mailing list would be used only for the purposes of this research, and would be returned to the Lansing School District. Participation in this survey research would, of course, be voluntary, and the annonymity of all respondents would be assured. We feel especially committed to protecting the annonymity and confidentiality of all survey respondents, and therefore have chosen to protect their confidentiality by not coding either the surveys or return envelops in any way that could be used to identify respondents. An additional safeguard to protect confidentiality is that all demographic information that is requested on the survey is very general, and could not be used to identify any particular respondent. Individuals who are asked to participate in this research would benefit by receiving the results of this study, as well as a bibliography for further reading, if they so request. I would appreciate it if you would call me, or write, in order to clarify what steps need to be taken in order to obtain a mailing list of school teachers in the Lansing School District. Sincerely, Rochelle Habeck, Ph.D. 154 Appendix B Questionnaire Instructions: Please complete this survey questionnaire if you are presently married. If you are not presently married, please complete the survey labeled, ”Unmarried Women's Role Questionnaire," and return both surveys in the enclosed envelope. A. BACKGROUND INFORMATION 1. Occupation (circle one) ! Physician / Elementary School Teacher ‘ Registered Nurse / University faculty ‘ 2. Age “ 3. How many years have you worked in your present occupation? 4. Annual combined salary of you and your husband (circle one) ' $10,000-19,999 / $20,000-29,999 / $30,000-39,999 $40,000-49,999 / $50,000-59,999 / $60,000-69,999 $70,000 + 5. How many hours do you work in a typical week. (Include work brought home at night) 6. Length of marriage to present spouse e one) 7. How many children are presently living in your home? 8. Ages of children in home? 155 B. PERSONAL ATTITUDES TOWARD ROLES Please put an X next to the ONE statement that best describes your felings. 9. Our decision relative to children is that: a) We we now have a child (or children). b) Although we do not yet have children, we hope to have them in the future. c) We don't know if we want children. d) We don't want children. Comments 10. When someone else takes care of the children (or took care of them, if they are grown): a) I feel fine about the care I have arranged for my children b) I feel guilty and anxious about not being with them. c) I have ambivalent feelings about the situation. d) Not applicable. Comments 11. Do you feel that your employment has had any good/bad effects on your children? a) It has had only positive effects (e.g., they are more independent, etc.) b) It has had more positive effects than negative effects. c) It has probably had about the same amount of positive and negative effects (e.g. they are more independent and also somewhat neglected). d) Not applicable. Comments \A‘ 156 12. My feelings about household chores can best be described as follows: a) I worry a great deal about managing to get the chores done. b) I have hired outside help to take care of all the major chores. c) My family and I share responsibility and concern about getting chores done. d) I do my best and don't worry if the chores don't get done. Comments 13. If I had to rate the relative importance of my family, home, and career roles, I would say that: a) My career definitely comes first. b) My family definitely comes first. . c) My homemaker role definitely comes first. d) It is impossible to rate these roles. Comments C. ATTITUDES OF OTHERS TOWARD MY CAREER Please circle ONE answer for each statement. (SA - Strongly Agree, A - Agree, D a Disagree, SD - Strongly Disagree) Rating 14. My husband provides sufficient emotional support in terms of my career. SA A D SD 15. I consider my husband's career and my career to be equally important. SA A D SD 16. My husband considers his career and my career equally important. SA A D SD 17. My husband and I have often felt in competition in terms of careers. SA A D SD 18. My husband and I sometimes compete in areas not related to our careers (who is a better parent, tennis player, etc.) SA A D SD 19. My friends have been totally supportive of my career. SA A D SD 157 20. My parents have been totally supportive of my career. SA A D SD 21. My colleagues have been totally supportive of my career. SA A D SD D. ROLE STRAIN AND COPING METHODS Please circle ONE answer for each statement (SA - Strongly Agree, A - Agree, D - Disagree, SD - Strongly Disagree) Rating 22. I often feel a strain in trying to fulfill both home and career obligations. SA A D SD 23. I often feel a strain between my personal needs and career obligations. SA A D SD 24. I often feel a strain between my personal needs and home obligations. SA A D SD 25. I often feel a strain because of the lack of time to do all the things I want to do. SA A D SD 26. I often feel a strain in trying to fulfill my parental obligations (if applicable). SA A D SD 27. I often feel a strain in trying to fulfill my obligations as a marriage partner. SA A D SD 28. I often feel a strain in trying to fulfill my career obligations. SA A D SD 29. I have eliminated certain activities within roles (e.g., I no longer do the cooking) SA A D SD 30. I have hired outside help to assist me with chores. SA A D SD 31. I am working hard to change societal definitions of women's roles. SA A D SD 32. Family members share household tasks with me. SA A D SD . 33. Family members help me resolve conflicts between my roles. SA A D SD 158 34. I overlap roles whenever possible (e.g., I sometimes bring my child to the office, or I work with my husband) SA A D 35. I have established rules and priorities for dealing with roles (e.g., if my child is sick, I stay home from work). SA A D 36. I have reduced my standards within certain roles (e.g., I no longer care if the house is spotless). SA A D 37. I have developed new attitudes to reduce conflicts (e.g., I no longer think I can be "superwoman" and do everything well). SA A D 38. I keep my roles totally separate (e.g., I never bring work problems home or home problems to work). SA A D 39. I rotate my attention from home to children to work, depending on which need is most pressing. SA A D 40. I consider my personal interests very important (e.g., I make time to play the piano). SA A D 41. I tend to schedule and organize activities very carefully to maximize my output. SA A D 42. I attempt to meet all role demands by doing everything people expect of me. SA A D 43. I have eliminated entire roles as I could not handle so many different demands on my time (e.g., I decided not to be a mother because my career is so important to me). SA A D 44. I have no conscious strategies for dealing with conflicts between roles. SA A D E. SATISFACTION WITH COPING STRATEGIES Please circle ONE answer for each statement. (ED - Extremely Dissatisfied, MD - Mildly Dissatisfied, MS - Mildly Satisfied, ES - Extremely Satisfied) SD SD SD SD SD SD SD SD SD SD SD Rating 159 45. How satisfied are you with the way you have dealt with possible role strains in your life? ED MD Ms ES 46. Overall, how satisfied do you feel with your career? ED MD Ms ES 47. Overall how satisfied do you feel as a mother (if applicable). ED MD MS ES 48. Overall how satisfied do you feel as a marriage partner? ED MD MS ES Appendix C F. PERSONAL ATTRIBUTES QUESTIONNAIRE The items below inquire about what kind of a person you think you are. Each item consists of a pair of characteristics, with the letters A-E in between. For example: Not at all A0.IOOBOOOOOCOOOOODOOOOOE very Artistic Artistic Each pair describes contradictory characteristics -- that is, you cannot be both at the same time, such as very artistic and not at all artistic. The letters form a scale between the two extremes. You are to choose a letter which describes where you fall on the scale. For example, if you think you have no artistic ability, you would choose A. If you think you are pretty good, you might choose D. If you are only medium, you might choose C, and so forth. ’ 49. Not at all independent A...B...C...D...E Very independent 50. Not at all emotional A.6.B...C...D...E Very emotional 51. Very passive A...B...C...D...E Very active 52. Not at all able to A...B...C...D...E Able to devote devote self self completely completely to others to others 53. Very rough A...B...C...D...E Very gentle 54. Not at all helpful A...B...C...D...E Very helpful to others to others 160 55. Not at all A...B...C...D...E Very competitive competitive 56. Not at all kind A...B...C...D...E very kind 57. Not at all aware of A...B...C...D...E Very aware of feelings of others feelings of others 58. Can make decisions A...B...C...D...E Has difficulty easily making decisions 59. Gives up very easily A...B...C...D...E Never gives up easily 60. Not at all A...B...C...D...E Very self- confident confident 61. Feels very inferior A...B...C...D...E Feels very superior 62. Not at all A...B...C...D...E Very understanding understanding of others of others 63. Very cold in A...B...C...D...E Very warm in relations to others relations to others 64. Goes to pieces under A...B...C...D...E Stands up well pressure under pressure Appendix D G. TEXAS SOCIAL BEHAVIOR INVENTORY The Social Behavior Inventory asks you to describe your reactions and feelings when you are around other people. Each item has-a scale, marked with the letters A, B, C, D, and E, with (A) indicating "not at all characteristic of me" and (E) "very much characteristic of me," and the other letters, points in between. . A B C D E Not at all Not Slightly Fairly Very much character- very characteristic istic of me of me For each item, choose the letter which best describes how characteristic the item is of you. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74; 75. 76. 77. 78. 79. 80. 161 I am not likely to speak to people until they speak to me. I would describe myself as self-confident. I feel confident of my appearance. I am a good mixer. When in a group of people, I have trouble thinking of the right things to say. When in a group of people, I usually do what the others want rather than make suggestions. When I am in disagreement with other people, my opinion usually prevails. I would describe myself as one who attempts to master situations. Other people look up to me. I enjoy social gatherings just to be with people. I make a point of looking other people in the eye. I cannot seem to get others to notice me. I would rather not have very much responsibility for other people. I feel comfortable being approached by someone in a position of authority. I would describe myself as indecisive. I have no doubts about my social competence. THANK YOU! Rating AOOBOOCOODOOE AOOBOOCOODOOE AOOBOOCOODOOE AO OBO OCO ODO OE AO OBO OCO ODO OE AO OBOOCO ODO OE A..B..C..D..E AO OBO OCO ODO OE AO OB. OCO ODO OE A. OB. OCO ODO OE AO OBO OCO ODO OE A. OBO OCO ODO OE AO OBO OCO ODO OE AO OBOOCO ODO OE AO OBOOCO ODO OE AO OBO OCO ODO OE «‘ 162 Appendix E Letter to Accompany Survey Dear Survey Recipient: The purpose of this letter is to invite you to participate in an important study, the Working Professional Woman's Role Study. There is evidence to suggest that the multiple roles of working married women (i.e., wife, career, and possibly parent) may be potent sources of life satisfaction, but also sources of strain. The purpose of this research is to learn more about the relationships among these variables in professionally employed women. Women employed as elementary school teachers, physicians, university faculty, and _ registered nurses are being asked to participate in this research. As one of the few women who we are asking to assist us in this research, your response is essential. However, your partcipation is voluntary and you may decline to participate or withdraw your participation at any point without penalty. Enclosed is a survey questionnaire that we are asking you to complete and return in the enclosed envelope. The primary focus of this research is upon married working women, however we are also very interested in the opinions of unmarried women. Therefore, we ask that unmarried women complete the survey labeled "Unmarried Women's Role Questionnaire." Please return both surveys in the enclosed stamped, envelope. We estimate that it will take about 15 minutes to complete the full questionnaire. The information you provide will be used for research purposes only (i.e., Don Lawson's dissertation study). We have taken steps to assure the annonymity and confidentiality of all respondents by not coding the enclosed survey or return envelope in any way that could be used to identify you. Additionally, the demographic information requested is quite general, and could not be used to identify individuals who complete the surveys. We will be happy to provide a summary of the results of this research to any interested participant, as well as a bibliography of sources to read on this important topic. In order to request this information, complete the enclosed postcard, and drop it in the mail. We would appreciate it very much if you could fill out the questionnaire and return it in the enclosed stamped, self-addressed envelope by June 1, 1987. Thank you very much for your time and interest. Sincerely, Rochelle Habeck, Ph.D. Don Lawson, M.A. 163 Appendix F Postcard to Accompany Surveys Working Professional Woman's Role Study Yes, I would like to recieve: ___ Summary of research results ___ Bibliography of readings Please complete: Name: Address: 164 Appendix G Unmarried Women's Role Questionnaire A. BACKGROUND INFORMATION 1. Occupation (circle one) Physician / Elementary School Teacher Registered Nurse / University faculty 2. Age 3. How many years have you worked in your present occupation? 4. Annual salary (circle one) $10,000-19,999 / $20,000-29,999 / $30,000-39,999 $40,000-49,999 / $50,000-59,999 / $60,000-69,999 $70,000 + 5. How many hours do you work in a typical week. (Include work brought home at night) 6. How many children are presently living in your home? 7. Ages of children in home? B. ROLE STRAINS Please circle ONE answer for each statement (SA - Strongly Agree, A - Agree, D - Disagree, SD - Strongly Disagree) Rating 8. I often feel a strain in trying to fulfill both home and career obligations. SA A D SD 9. I often feel a strain between my personal needs and career obligations. SA A D SD 10. I often feel a strain between my personal needs and home obligations. SA A D SD 11. I often feel a strain because of the lack of time to do all the things I want to do. SA A D SD 12. I often feel a strain in trying to fulfill my parental obligations (if applicable). SA A D SD 165 13. I often feel a strain in trying to fulfill my career obligations. SA A D SD C. SATISFACTION WITH COPING STRATEGIES Please circle ONE answer for each statement. (ED - Extremely Dissatisfied, MD - Mildly Dissatisfied, MS - Mildly Satisfied, ES - Extremely Satisfied) Rating 14. How satisfied are you with the way you have dealt with possible role strains in your life? ED MD MS ES 15. Overall, how satisfied do you feel with your career? ED MD MS ES 16. Overall how satisfied do you feel as a mother (if applicable). - ED MD MS ES THANK YOU! 166 Appendix H First Follow up Postcard Dear Survey Recipient: Just a reminder! About a week ago we mailed a survey to you that dealt with role conflicts of married professional women. If you have already completed and returned the questionnaire, thank you for your help with the Working Professional Woman's Role Study. If you have not yet completed the survey, please take a few moments today to complete and mail it. If you have misplaced the survey, you may request another survey by contacting us at Michigan State University, 513 Erickson Hall, East Lansing, MI 48824-1034, or by calling Don Lawson at 353-4397. Sincerely, Rochelle Habeck, Ph.D. Don Lawson, M.A. Project Coordinator Project Assistant I-EE;” 167 Appendix I Second Follow up Postcard Dear Survey Recipient: Just another friendly reminder. About two weeks ago, we mailed you a survey that focused on the role conflicts of married professional women. We have no way of knowing if you have filled out and mailed the survey. If you have, great! We appreciate your assistance with the Working Professional Woman's Role Study. If you have not completed and mailed the survey, please do so today. If you have misplaced the survey, you may request another survey by contacting us at Michigan State University, 513 Erickson Hall, East Lansing, MI 48824-1034, or by calling Don Lawson at 353-4397. Sincerely, Rochelle Habeck, Ph.D. Don Lawson, M.A. Project Coordinator Project Assistant .- ' REFERENCES 168 References American Psychological Association. (1982). Etpiaal principles in paa aopdacp pf :eseargh Kipp haaap participants, Washington D.C.: American Psychological Association. Anderson-Kulman, R. E., 8 Paludi, M. A. (1986). Working mothers and the family context: Predicting positive coping. goapnal pf Vocational Behaviop, zap 241-253. Bakan. 0- (1966) - The duality 9.: 11mm existence. Chicoago: Rand McNally. Baruch, G. E., Biener, L., 8 Barnett, R. C. (1987). Women and gender in research on work and family stress. Aaapipap Esayhplogist, 521 130-136. Bem, S. L. (1974). The measurement of psychological androgyy. Journal a; Qonsultipg and Clinical Psychology. 42, 155- 162. Bem, S. L. (1977). On the utility of alternative procedures for assessing psychological androgyny. iaapaal p; QQDEBLLLDQ and Clinical 2§2282192¥1 £51 196'205- Blackwell, J. M. (1984). The relationship between sex role strain and measures of well-being among career women and homemakers. Dissertatiop apstracts Inpappapippala gap lOOS-B. (University Microfilms No. DA8413867) Bloom, J. R. (1979). Psychosocial measurement and specific- hypotheses: A research note. Jouppal a: Qppaplpipg and 91181221 EEYEQQLQQYI All 637-639- V 1 M 169 Borg, W. R., 8 Gall, M. D. (1971). Eapaapippal,zaaaapapl New York: David McKay Company. Bunker, J. W. (1985). Dynamics and components of social support: A study of dual-career couples. sse apappaapa lppappapippall gal 3201-8. (University Microfilms No. DA8522117) Cohen. J- (1969) - 52:12:12.2; 2m; enamels. fer the papayippal sgiepaas, New York: Academic Press. Cohen, R., 8 Lazarus, R. S. (1979). Coping with the stresses of illness. In G. C. Stone, F. Cohen, 8 N. E. Adler (Eds.), Haalph payphplagyl San Fancisco: Jossey-Bass. Dillman, D. A. (1978). Mail apa telephone sugyays: Tag papal daaigp maphaal New York: Wiley. Edwards, C. T. (1983). Role development, role strain, coping and satisfaction among professionally employed mothers. new Abstracts W 4.5.. 668-3- (University Microfilms No. DA8405225) Epstein, C. F. (1973). Wapaa's 216221 Berkeley: University of California Press. Fitzgerald, L. R., 8 Crites, J. O. (1980). Toward a career psychology of women: What do we know? What do we need to know? Jpappal pf Qpapaalipg zsygaology, 211 44-62. Folkman, S. (1984). Personal control and stress and coping processes: A theoretical analysis. apappal a: Eapappalipy apg,§paial Esyghology, gal 839-852. 170 Folkman, 8., 8 Lazarus, R. S. (1980). An analysis of coping in a middle-aged community sample. Qaappal pf fiaalpa apa fiesial Behaxierl.211 219-239. Folkman, S., 8 Lazarus, R. S. (1985). If it changes it must be a process: Study of emotion and coping during three stages of a college examination. lapppal a; Papappalipy aaa SQQlal Egyghology.‘1§l 150-170. Folkman, S., Lazarus, R. 8., Dunkel-Schetter, C., DeLongis, A., 8 Gruen, R. J. (1986). Dynamics of a stressful encounter: Cognitive appraisal, coping, and encounter outcomes. apappal a; Eepsopality and Spaial Esycaalogy, gal 992-1003. Folkman, S., Lazarus, R. S., Gruen, R. J., 8 DeLongis, A. (1986). Appraisal, coping, health status, and psychological symptoms. lpappal p1 Eapappalipy and 522161 Esxshelegxl 591 571-579. Gentry, W. D., 8 Kobasa, S. C. O. (1984). Social and psychological resources mediating stress-illness relationships in humans. In W. D. Gentry (Ed.), fiapappak a: paaayipral pagiaipal New York: Guilford. Glass, G. V, 8 Hopkins, K. D. (1984). §£Q£i§§1221.M§LDQQ§ in Educatipp apg Baaaazana New Jersey: Prentice-Hall. Goode, W. J. (1960). Theory of role strain. Aaapiaap Sociolagical Review, gal 483-496. Gray, J. D. (1980). Role conflicts and coping strategies in married professional women. Qiaaarpapipp apappaapa 171 lppappapippall gal 3781A. (University Microfilms No. 7928135) Gray, J. D. (1983). The married professional woman: An examination of her role conflicts and coping strategies. 2222221221 2! K2222 9222222121 11211 235’243- Guadagnoli, E., 8 Velicer, W. F. (in press). Stability of component patterns: Simulation study. 2212221221221 2% Habeck, R. V., Ewing, M. E., Overton, S. R., 8 Dummer, G. M. (1986, August). Qapipg sppategias pf papsops yiph physiaal disabilipias in apprpl Paper presented at the Annual Meeting of the American Psychological Association, Washington, D. C. Hall, D. T. (1972). A model of coping with role conflict: The role behavior of college-educated women. d s ve 2212822 9222222111 111 4717485- Harren, V., Kass, R., Tinsley, H., 8 MOoreland, J. Influence of gender, sex role attitudes, and cognitive complexity on gender-dominant career choices. goappal pf Counsaling gsyagology, gal 227-234. Harrison, A. O., 8 Minor. J. H. (1982). Interrole conflict, coping strategies, and role satisfaction among single and married employed mothers. Esyghology pf meap Qaarpaplyl alaly 354-360. Helmreich, R., 8 Stapp, J. (1974). Short forms of the Texas Social Behavior Inventory, an objective measure of self- 172 esteem. 83.119219. 9f the W misty... 51 473- 475. Helmreich, R., Stapp, J., 8 Ervin, C. (1974). The Texas Social Behavior Inventory (TSBI): An objective measure of self-esteem or social competence. £282 Qapalpg a; $_l___d.e ecte W in mm 4. 79- Isaac, S., 8 Michael, W. B. (1971). nanabopk in reseapgn apa evalaation, San Diego: Robert R. Knapp. Kim, J. (1975). Factor analysis. In Nie, N. H., Hull, C. H, Jenkins, J. G., Steinbrenner, K., and Bent, D. H. (Eds-L 22.82:. Statistical p__geacka fer the 5221.11 sciapgas, New York: McGraw-Hill. Lazarus, R. S., 8 Folkman, S. (1984). Coping and adaptation. In W. D. Gentry (Ed.), Handbpok a; paaavioral pediaiae. New York: Guilford. Lazarus, R. S., 8 Launier, R. (1978). Stress-related transactions between person and environment. In L. A. Pervin 8 M. Lewis (Eds.), Eepspectivas la ippapaapippal psycholagy, New York: Plenum. Lemkau, J. P. (1979). Personality and background characteristics of women in male-dominated occupations: A review. Psycaalogy a; Wopep Quazperly, 1y 221-240. Lemkau, J. P. (1983). Women in male-dominated professions: Distinguishing personality and background characteristics. Psychalogy 91 Women Qpartaplyl 8121, 144-164. Lemkau, J. P. (1984). Men in female-dominated professions: 173 Distinguishing personality and background features. Jeurnal ef.Yeeatienel,Behexier1 211 110-122. Levinson, D. (1959). Role, personality, and social structure in the organizational setting. aparpal pf Abppppal apa Seeial Bexehelegxl 581 170-180. Levo, L. M. (1983). An investigation of the efficacy of competing explanations of how women cope with stressful home-career conflicts. pissappatiop apstragta lppappagippall gal 1967-8. (University Microfilms No. DA8322618) Meslach. C- (1982)- Beraeeei The eeet ef eerinel Englewood Cliffs, N.J.: Prentice-Hall Mawardi, B. H. (1983). Aspects of the impaired physician. In B. A. Farber (Ed.), Strass ap§,pprppa; ip 222 aamap aazyiaa ppafaaaippal New York: Pergamon. Moreland, J. R., Gulanick, N., Mantague, E. K., 8 Harren, V. A. (1978). Some psychometric properties of the Bem Sex- Role Inventory- Applied Eexeheleeieel Meeeeremenel.21 249-256. (1985)- 2222 eexeaeee eeetieeiee eeieel New York: McGraw Hill. (1981). Role conflict and sex role concepts in Norusis, M. J. Oates, M. H. selected groups of multiple role women. Qiaaappapipp beetreete Internetienell 121 2506-A- (University Microfilms No. 8127391) Olds, D. E. (1980). Coping with the stress of a dual-career marriage. Qissaptapiop apappaapa lpparpational, 211 174 4746-8. (University Microfilms No. 8110766) Parsons, T., 8 Bales, R. F. (1955). Eaailyl 22212112221221 apa ippapaapiap 22222222 Glencoe, 111.: Free Press. Pearlin, L. I., 8 Schooler, C. (1978). The structure of coping- ieernal 2: Health age eeeiel Behexierl 121 2- 21. Poloma, M. M. The married professional woman: An empirical examination of three myths. Qiaaappapiap apappaapa lppappapiapall gal 563-A. (University Microfilms No. 7119042) Pitts, F. N., Schuller, A. 8., Rich, C. L., 8 Pitts, A. F. (1979). Suicide among U.S. women physicians, 1967-1972. 82221222 1222221 21 22222122211 122121 594-695- Raj, D. (1972). Tag gaaigp a; aaapla aapyayal New York: McGraw-Hill. Rapoport, R., 8 Rapoport, R. N. (1969). The dual-career family. Hanan Ralapippal 22y 3-30. Rapoport, R., 8 Rapoport, R. N. (1980). Three generations of dual-career family research. In F. Pepitone-Rockwell (Ed.), 2221222222: capplaal Beverly Hills: Sage. Rosenkrantz, P., Vogel, 8., Bee, H., Broverman, I., 8 Broverman, D. M. (1968). Sex role stereotypes and self- concepts in college students. apappal 91 Qppaaltipg and 21121221 22222212211 221 287-295. ' Rosow, I., 8 Rose, D. K. (1972). Divorce among doctors. leernel ef Merrieee 222 the 2221121 211 587-598. Schaper, G. L. (1983). Married working mothers: An alnalysis 175 of role strain as a function of coping strategies, sex- role attitudes and work/family characteristics. Diasertatiop Abstpacts lppappapippall gal l977-B. (University Microfilms No. DA83238005) ' Spence, J. T., 8 Helmreich, R. L. (1978). Masculinipy apa femininity; Their 2222221221221 eimeneienel eerreletee and aptecendents2 Austin: University of Texas Press. Spence, J. T., Helmreich, R. L., 8 Holahan, C. K. (1979). Negative and positive components of psychological masculinity and femininity and their relationships to self-reports of neurotic and acting out behaviors. Jeerael ef Bereenalifx and 292121 22122219221 311.1673- 1682. Spence, J. T., Helmreich, R. L., 8 Stapp, J. (1975). Ratings of self and peers on sex role attributes and their relation to self-esteem and conceptions of masculinity and feminity. Jpappal a; Pepsopalipy apg Soaial anghology.‘agl 29-39. Strange, C. C., 8 Rea, J. S. (1983). Career choice considerations and sex role self-concept of male and female undergraduates in nontraditional majors. gaprpal 91 2222212221 Bahaviop, 212 219-226. Suchet, M., 8 Barling, J. (1986). Employed mothers: Interrole conflict, spouse support and marital functioning. 2222221 21 922222212221 2222212221 11 167-178- Tinsley, H. E. A., 8 Heesacker, M. (1984). Vocational 1 176 behavior and career development, 1983: A review. Jppppal 22.!222212221 222221221 222 139-190- U. S. Bureau of the Census. (1983). Detailed occupation and years of school completed by age, for thecivilian labor force by sex, race, and Spanish origin. 1222 Qapapa a; pppalaaippl (Supplementary report number PC80-Sl-8). Washington D. C.: U. S. Government Printing Office. U. s. Department of Labor. (1982). 29,2aapa pp ypaap yorkazs. Washington D. C.: U. 8. Government Printing Office. Vaillant, G. E., Sobowale, N. C., 8 McArthur, C. (1972). Some psychologic vulnerabilities of physicians. Ina Nay 2221222 2221221 2: 14.92.3122. 2.8—2.7: 372-375. White, R. (1974). Strategies of adaptation: An attempt at systematic description. In G. V. Coelho, D. A. Hamburg, 8 J. E. Adams (Eds.), Cpping apa agaptatippy New York: Basic Books. Wrubel, J., Benner, P., 8 Lazarus, R. S. (1981). Social competence from the perspective of stress and coping. In J. D. Wine 8 M. D. Smye (Ed§.), fipaial mepapapaal New York: Guilford. Yogev, S. (1983). Judging the professional woman: Changing research, changing values. Payaaalpgy pf flppap 222222211; 11 219-233- -fi