0-» LIBRARY icon flute University This is to certify that the dissertation entitled Family Life Events, Perceived Stress, and Social Support Utilized: An Examination of Sex Differences presented by Terri Robinson Williams has been accepted towards fulfillment of the requirements for the.PhD Family Ecology degree in / Major professor Date/d“; é/ff—S/ um”.-- ‘5!— ..L_ A : n-v ll‘ - l 1 . 012771 L‘ __ a a .1 till Willi/I MSU WWW/ll Z. 3 1293 01005 2995 RETURNING MATERIALS: Place in book drop to LIBRARJES remove this checkout from Jun—sun. your record. FINES will be charged if book is returned after the date stamped below. N09133:” 1 50 4 k . -. "in ‘1 ‘2‘! q 5*? 2 5 nun 7 am: EMA—9W FAMILY LIFE EVENTS, PERCEIVED STRESS, AND SOCIAL SUPPORT UTILIZED: AN EXAMINATION OF SEX DIFFERENCES by Terri Robinson Williams A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Family and Child Ecology 1985 ABSTRACT FAMILY LIFE EVENTS, PERCEIVED STRESS, AND SOCIAL SUPPORT UTILIZED: AN EXAMINATION OF SEX DIFFERENCES by Terri Robinson Williams The purpose of this study was to extend understanding of male and female differences in the perceived degree of stress resulting from family life events experienced and from relationships and aspects of life which occur on a routine or daily basis. Another purpose was to examine male and female differences in the utilization of social support in times of stress. A final purpose was to examine the employment status of men and women as it relates to the perceived degree of stress experienced. The research was conducted using a sample selected from the NC-164 regional project entitled, "Stress, COping, and Adaptation in the Middle Years of the Family" which was supported by the Agricultural Experiment Stations of the North Central Region of the United States and the Cooperative Research Service of the United States Department of Agriculture. The sample primarily included couples in which the wife was between the ages of thirty-five and sixty-five and which had at least one child. The sample was selected from four rural counties and one Standard Metropolitan Statistical Area in the State of Michigan. Data were gathered by using a mail questionnaire; each husband and wife had a questionnaire and were encouraged to complete them separately. The results of the study indicate that women report that more family life events have taken place. Although not statistically significant, a difference was found be— tween men and women in the perceived total degree of stress experienced. Law events and events that involved relatives and close friends were reported as significantly more stressful in the perceptions of women. Individual events, such as the piling-up of household chores, were found to be perceived as more stressful for women. No events were found to be more stressful for men and men reported lower perceived stress levels. Day-to-day, routine relationships and aspects of life had an equal impact on men and women. The total degree of perceived stress was significantly different for persons in different employment statuses. Those who were retired, homemakers, and full-time employed reported less perceived stress than the disabled, unemployed and part-time employed. Women utilized more social support than men in times of stress. To My Parents, Robert N. and Lucille Robinson and To My Husband, Donald R. Williams ii ACKNOWLEDGEMENTS Appreciation is extended to my guidance committee, Dr. Margaret Bubolz, Dr. Dennis Keefe, Dr. Lillian Phenice, and Dr. Ellen Strommen for their direction and support through all facets of my doctoral program. I wish to especially thank Dr. Bubolz, the chairperson of my committee, who gave me the time and guidance that I needed to have a successful and positive experience during my program and my dissertation. I wish to thank Dr. Barbara Ames who offered me opportunities to develop as an instructor and provided encouragement and friendship through all phases of my work. Appreciation is extended to the NC-164 project research team who provided me with the Opportunity to do the type of research that I hOped to do.‘ A special thanks is extended to Cynthia Cameron whose research abilities, con- structive suggestions, concern and time were invaluable in the preparation of this dissertation. Also, thanks go to Dr. Dwayne Smith for providing needed information and encouragement. A very special thanks is extended to my dear parents, Robert and Lucille Robinson who nurtured me in such a way that allowed me to dream dreams and make them a reality and to my husband Donald Williams, who unselfishly stepped aside to let me proceed. iii TABLE OF CONTENTS Page LIST OF TABLESIIOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO. Vi LIST OF FIGURESOOOOO00......OOOIOOOOOOOOOOOOIOOOOOO Vii CHAPTER I. INTRODUCTIONIOOOOOOOOOOOIOOOIOOOOOIOOOOOOOOO 1 Identification of the Problem............ Statement of Purpose..................... Importance of Study...................... Theoretical Framework.................... Research Questions and Hypotheses........ 11 Terms and Definitions.................... 17 Assumptions.............................. 23 \IKJINH II. REVIEW OF RELATED LITERATURE................ 25 The Middle Years......................... 25 Stress................................... 31 Sex Differences and Stress............... 43 Coping and Social Support................ 51 Summary.................................. 56 III. METHODOLOGYOOIOOOO...OOOOOOOIOOOIOOOOOOCOIO. 58 Research Design.......................... 58 Research Instrument...................... 59 Family Life Events Scale.............. 61 Day-to-day, Routine Relationships and Aspects of Life................ 63 Social Support Utilization Scale...... 6# Research Sample.......................... 66 Response Rate......................... '69 Characteristics of Sample............. 71 Techniques of Data Collection............ 77 Data Analysis............................ 78 iv CHAPTER Page IV. RESULTSIOOO0...0000.0.0...CIOOOOIOOIOOOIOOOOOO 814 Research Question One...................... 85 Research Question Two...................... 88 Research Question Three.................... 92 Research Question Four..................... 95 Research Question Five..................... 99 V. DISCUSSION.OOOIOOIOCOOOOOOIOOOOOOOICC...III...103 Discussion of Research Question One........ 103 Discussion of Research Question Two........ 105 Discussion of Research Question Three...... 110 Discussion of Research Question Four....... 113 Discussion of Research Question Five....... 116 Sex Role Socialization..................... 120 The Theoretical Fit Between the Boss Model of Family Stress and the Human Ecosystem Approach................ 12h VI. SUMMARY, CONCLUSION, AND IMPLICATIONS......... 129 Summary.................................... 129 Conclusions................................ 130 Limitations................................ 133 Implications............................... 134 APPENDICES A. The Research Instrument....................... 141 B. FILE and F-COPES.............................. 155 C. Sample Selection.............................. 159 D. Coding Procedures............................. 167 E. Characteristics of the Sample................. 169 F. Chi-Square and T Values....................... 173 REFERENCESOO0.......0O...00......OOOOOOOCOOOOIOOIOOO 178 10. 11. 12. LIST OF TABLES Page Alpha Reliability and Test/Retest Correlations of FILE........................... 63 Alpha Reliability and Test/Retest correlations for F-COPESooooooooooooocoo-coocoo 65 women: Employment StatUSI I I I I I I I I I I I I I I I I I I I I I 75 Men: Employment Status........................ 75 Family Income for 1982......................... 77 T-Test of Differences Between Men and Women and the Number of Events Identified by the Nature Of the EventSIIIIIIIIIIIIIIIIIIIIIIIIIII 87 T-Tests of Differences Between Men and Women of Degree of Stress by Nature of Events........ 91 T-Tests of Differences Between Men and Women in the Impact of Day-to-day, Routine Relationships and Aspects of Life.............. 94 T-Tests of Differences Between Employment Status of Men and Women and Total Degree Of Perceived StreSSIIIIIIIIIIIIIIIIIIIIII.II... 96 T-Tests of Differences Between Employment Status and the Total Degree of Perceived Stress Of Men and womenI I I I I I I I I I I I I I I I I I I I I I I I 98 T-Tests of Differences Between Men and Women in Social Support Utilization..................... 101 Family Life Events Perceived as Moderately to Extremely StreSSfulIIIIIIIIIIIIIIIIIIIIIIIIIIII 108 vi LIST OF FIGURES Figure Page 1. Hill's ABC-X Model REvised to Show Degree of Stress and the Alternative Outcomes of Crisis and copingIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII. 10 2. The Boss Model: A Contextual Approach to Family StreSSIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII 10 . The Person/Environment Fit Paradigm........... 3h . The Varying Loads of Stress................... 35 3 L1 5. A Continuum of the Stress Experience.......... 36 6 Design for Analysis........................... 60 7 . A Human Ecosystem Model....................... 125 vii CHAPTER I INTRODUCTION IDENTIFICATION OF THE PROBLEM Stress has become a tOpic of rising concern in recent years for persons from every walk of life including those who study stress in others and those who experience stress in their daily lives. Stress is being investigated from the perspective of many professions and disciplines such as psychology, sociology, physiology, anthropology, business, management, medicine, and family ecology. This illustrates the far reaching nature of stress in the lives of individuals. Concerns focus on the harmful effects of stress on persons which manifest themselves physically, emotionally, and socially and which can have long term or permanent damage. Rabkin and Struening(l976, p. 1014) state that, "Mounting stress precedes sudden cardiac death, myocardial infarctions, accidents, athletic injuries, tuberculosis, leukemia, multiple sclerosis, diabetes, and many minor medical com- plaints." These are a small selection of the physical manifestations; there are other consequences of stress including psychological, emotional, and social effects. Because of its potentially harmful effects, concern arises for its reduction and control. For reduction and control to be accomplished, a more complete understanding of stress is needed. 2 One aspect of the stress experience which has a limited understanding and information base is the difference in the perceived threshold of psychological-emotional stress of men and women, particularly the stress men and women perceive individually as a result of undergoing family life events. If men and women perceive stress differently in life events then there may be varying effects on their lives. This perception may impact upon the entire stress experience. The experience can be internal to the individual emotionally, physically, or psychologically or external to the individual as evidenced in social relationships with family, friends, relatives and others. Socially, men or women may choose to interact with others differently when stress is perceived. Social interaction in times of stress is also an area of , limited information. This study will explore the differences in the stress perceptions that men and women have ofselected family life events and in the utilization of social support. STATEMENT OF PURPOSE The general purpose of this study is to increase knowledge of the stress experience by examining differences in the perception of stress by men and women. The investi- gation will focus on the perception of stress in family life events rather than on the objective existence of stress. The actual measure of physical or psychological stress is not the aim of this study; the aim is to examine differences in the male and female discernment of stress 3 related to certain family life events. Focus is on the distinctive or diverse manner in which the sexes perceive stress. It is the hope of the researcher that this information can be used for testing or expanding the theory on which this study is based and for practical use in the development of methods for the control and reduction of stress in individual lives. The sample for this study itself generates a second overarching purpose for the study. The participants selected are in the middle years of their lives. The body of research on the topic of the middle years of life has been relatively small. Research that focuses on the ex- perience of persons in this time of life has only recently begun to be emphasized. Because of this general lack of information, this particular research effort has as one of its general purposes to generate additional understanding of this period of family life for the individual. Although the topic of stress is relatively specific, the fact that this research deals with family life events makes it broader in scope. Family life events which have taken place in the lives of the participants will be identified giving greater knowledge of potential stressors in the middle years of life. This study has four specific purposes. The first is to identify differences in the number of family life events perceived by men and women as occurring and also to identify sex differences in the nature of the perceived events. LL Family life events are grouped into categories, including events internal to the family, school and work events, events involving relatives and close friend, events related to household finances, and law events. It is the intention of the researcher to explore differences in the number of reported events within each category thereby determining the relationship between the nature of the stressful events and the sex of the respondent. The second purpose is to identify differences in the perceived degree of stress experienced by men and women resulting from the family life events. This purpose goes beyond the examination of the quantity of the events and considers the impact felt by the individual. The greater the perceived intensity of the stress the greater the potential impairment of the individual physically, emotionally, and psychologically. The impact that the employment status of men and women has on the degree of stress experienced is the third specific purpose. The concern is on how varying employment statuses of women(full-time, part-time,homemakers, etc.) impact upon the total degree of stress perceived. The employment status of men will also be investigated. For example, does being laid off or retired relate to the total dgree of stress perceived by men? The purpose is to discover if employment status has an impact on the stress perceived by men and women. A final purpose is to identify sex differences in the 5 social support individuals utilize when experiencing stress. The supports include family, friends, neighbors, and community services. These social supports give aid to the individual and help in the reduction and control of stress. It is a goal of this study to examine differences in the selection of social supports utilized by men and women. In summary, the purpose of this study is to identify and examine types of family life events that are perceived as stressful for men and women in the middle years. The degree to which these events cause stress will be analyzed according to the sex of the respondent. Differences in the stress experienced by employment status will be examined and social supports that middle aged men and women draw upon to cope with stress will be identified. IMPORTANCE OF THE STUDY The research study at hand is important for a wide variety of reasons. A basic reason is to test, expand, or clarify theory. It is the hope of the researcher that this study will contribute to the theoretical foundation on which this study is constructed. This is important because it will extend and enhance the understanding and information base that exists on stress. Since the tepics of stress and the middle years are relatively new research areas of concern, additional information is needed for a more complete understanding. Although much work has been done in the area of sex differences, the research will add a new dimension since little work has been done on stress 6 in the middle years and the influence of sex differences in the stress experience. There are many practical reasons why this research is important. In an age where scholars and scientists from a wide range of fields warn of rapidly increasing levels of stress, it is important to understand stress as completely as possible in order to more accurately intervene to prevent it or alter its effects. Reduction of undue stress is important for optimal health, psychologically, emotionally, socially, and physically. Differences between the perceptions that men and women have of stress in their lives can lead to the development of unique approaches for appropriate intervention with the ultimate goal of stress reduction and elimination where possible. A reduction in stress levels can lead to a decrease in related manifestations in all areas of life but particularly in the physical realm. Appropriate preventative measures can be employed instead of after-the- fact treatment. Programs can be developed that will aid in the preparation for potentially stressful family life events in order to reduce the stress and its effects. With new understanding can come new approaches to building family competencies. Individuals and families can be educated to prepare for, reduce, and eliminate undue stress. Finally, this study is important for those services that offer support in times of stress. New information about sex differences will aid these services in tailoring their programs to meet the needs of their clients. Methods of recruiting clients may change and new services may be developed. 7 THEORETICAL FRAMEWORK The theoretical framework used for this study is the stress theory originally proposed by Hill(l958) and further developed by Boss(l984). Boss's work, which directly builds on Hill‘s Family Under Stress model(l958, p. 143) includes these components; A(the stressor event) interacting with B(the family's crisis-meeting resources) interacting with C(the definition the family makes of the event) leading to X(the crisis). This is called the ABC-X model of family crisis or stress. The A, B, and C factors all combine to produce the crisis. A(the stressor event) is benign: in and of itself, it carries no negative or positive influence. It must be combined with the C factor which is the meaning the family attaches to that event. This affects the crisis or stress experienced because each family perceives the event uniquely. What may be a very disturbing situation for one family may not affect another family in the same negative way. Therefore, the perception of the event will influence the crisis or stress experience. The B factor is also significant in the production of stress and crisis. The B factor is the family's crisis-meeting resources. These can include material resources such as money or family strengths that are emotional, psychological and social. Each family will draw on a unique set of resources in response to the event. If these are strong, they will minimize the crisis or stress experienced. 8 Although this model could serve as the framework for this paper, the Boss model will be utilized because it extends the framework: it provides a more detailed model which takes into account some important variables and allows for greater understanding and explanation. Boss has developed her model in two parts in order for the processes to be more easily understood. The first.part(see figure 1) includes basic changes that Boss made on Hill's work. The interaction between the A, B, and C factors does not move in one direction; this model shows influence in both directions hence describing the mutual interaction of the factors. The X factor(crisis in Hill's model) is now considered as the degree of stress. The A, B, and C factors do not necessarily combine to form a crisis: the family may be able to successfully employ family resources and/or perceive the event as harmless thereby coping with the situation and avoiding crisis. Coping implies a lesser degree of stress: it is "avoiding crisis by holding the degree of stress to a tolerable level"(Boss, p. 2). This model shows how the degree of stress(the X factor)feeds back into the process by influencing the A, B, and C factors. Boss expands the model beyond the core by adding dimensions that involve the internal and external contexts in which the family finds itself(see figure 2). The ex- ternal context involves six indicators: 1)historical - the time when the event takes place, 2)economic - the state of the economy of the larger environment, 3)developmental - the stage in the life cycle of both the 9 individual and the family itself, 4)constitutional - biological health and physical strength of the members of the family, 5)religious - values and beliefs which influence how a family perceives the event regarding attempts to control or accept a stressful situation, and 6)cultural - canons and mores by which the family defines the event and the coping resources they employ. These six indicators of the external context influence the internal context. The internal context has three dimensions: 1)socio- logical - the structure and function of the family regarding its boundaries, 2)psychological - the family's ability to use defense mechanisms, and 3)Philosophical - the family's values and beliefs. The internal context influences and is influenced by the family responses which are on an emotional, cognitive, and behavioral level. The researcher believes this model most adequately describes the complex nature of the family stress process. Without the additions made by Boss, the process could not be described in its complexity. This study is concerned with one primary aspect of the model. It focuses on the individual's perception of stress related to an event. That is, the participants of this study are asked how they define a certain event that has happened to them. They can define it as disturbing or not disturbing in various degrees. This is their percep- tion of an event, the C factor in the model. Another concern lO ' + . - 4 A <———> [3 <——> C = X Dograo oi Biraaa <1 \\ Coping the come . THE PERCEPTION EVENT nesounces or THE event A B . c Hill '5 ABC-X Model revised to show degree at stress and the alternative outcomes at crisis and coping. Booo.’ Pauline. 'l’amily Strcoo: A Contextual Approach.“ Chapter written for Handbook on Man-la c and the Family. flarvln Suaaaan and Susan Stolmota. Eda.(Pub1TEB—€FT Hanna). (In Press) ' Figure 1 Hill's Model of Family Stress Revised EXTERNAL Maiorlcol COMIEXY If “'"m' Imenmu. CONYEXT MW ' Holloloaa g"... - loamacoa oliha "m. anvaa 'l' AI-f-fi—F. 9", c ' QX az'm Cultural . Anatomy ' """"""' "' (Ialomal’ Ioolaloolfll Behavioral ' i Psychological Enollooa rMoaoahlcalL - u to I 0" W?" .ztcaaa: aeontaatoa! Approach. Chaptoryrl an r ifi Mom). Ioaa. "in"; . hula loco-an and Ioaan Italic-an. “N"IW" (In haul. Figure 2 The Boss Model: A Contextual Approach to Family Stress 11 of this study is social support. Social support is a resource individuals can choose to use in times of stress and is part of the B factor, coping resources. Employment status is part of the external context. Boss does not explicitly include employment status in the model but implicitly,employment status of men and women would be part of the economic context. The middle years of life involve the developmental context of the individual. This model, although not used in its entirety, does provide the necessary theoretical structure for the study. The literature cited in the review section was selected to fit this model. The sections on the middle years and sex differences focus on the alternate perceptions of men and women at middle age. The variables of sex of the respondent and the middle years then influence perception. They also influence the resources that are available. For example, middle aged persons tend to be financially more stable than younger persons, hence giving them an additional resource. The literature review also deals with coping and social support which. in the model, is the level of stress or crisis experienced. RESEARCH QUESTIONS AND HYPOTHESES The research questions for this study follow. 1. Do men and women differ in perception of the number and nature of the family life events experienced? 12 2. Are there sex differences in the degree to which the family life events are stressful? 3. Will the sex of the respondent influence the impact which day-to-day, routine relationships and aspects of life make upon the individual? 4. Will the employment status of the respondent influence the perception of the total degree of stress experienced and will this be related to the sex of the participant? 5. Are the social supports that are selected and utilized during times of stress different for the sexes? From the research questions thirty-six hypotheses have been generated. In the literature reviewed for this study, research hypotheses were examined that indicated a direction for certain of the following hypotheses. For the hypotheses which a direction has been examined in other research, both the alternative and the null hypotheses are presented below. Where no direction has been considered previously, only the null hypothesis is presented. Hypothesis one predicts a relationship between the sex of the respondent and the number of family life events that they have perceived as taking place. Hypotheses two through eight involve a relationship between the sex of the respondent and the number of events perceived as taking place by the nature of the event. H : Men and women will perceive that an equal number of potentially stressful family life events has taken place. 01 13 H : Women will perceive that more potentially stressful family life events have taken place than men. H02: Men and women will perceive that an equal number of events that are internal to the family has taken place. H2: Women will perceive that more events internal to to the family have taken place than men. H03: Men and women will perceive that an equal number of school events has taken place. HO“: Men and women will perceive that an equal number of work events has taken place. H : Men and women will perceive that an equal number of events involving relatives and close friends has taken place. . Women will perceive that more events involving 5 relatives and close friends have taken place than men. H06: Men and women will perceive that an equal number of health events has taken place. H6: Women will perceive that more health events have taken place than men. 07' Men and women will perceive that an equal number of household finance events has taken place. H08: Men and women will perceive that an equal number of law events has taken place. Hypotheses nine through sixteen make predictions concerning the degree of stress perceived from family life events and the sex of the respondent. Hypothesis nine involves a mean degree of stress score developed by measuring the disturbance from all of the events. For hypotheses ten through sixteen, a mean stress score is cal— culated for each particular category of family life events based on the nature of the events. These hypotheses in- volve the mean scores for the categories of events and the sex of the respondent. 09' 011‘ 012’ 013 13‘ Ho1u‘ Hlu: H s 015 H016‘ 14 There is no sex difference in the perceived degree of stress resulting from family life events. Women will perceive a higher degree of stress resulting from family life events than men. Life events that are internal to the family will be perceived as equally stressful for men and women. Life events that are internal to the family will be perceived as more stressful for women than for men. There is no sex difference in the perceived degree of stress resulting from school events. There is no sex difference in the perceived degree of stress resulting from work events. Life events involving relatives and close friends will be equally as stressful for men and women. Life events involving relatives and close friends will be more stressful for women than for men. There is no sex difference in the perceived degree of stress resulting from life events involving health. Women will perceive a higher degree of stress resulting from life events involving health than men. There is no sex difference in the perceived degree of stress resulting from life events involving household finance. Life events involving the law will be equally stressful for men and women. Hypotheses seventeen through thirty-four make pre- dictions regarding the impact of day-to-day, routine relationships and aspects of life and the sex of the respondent. These are routine activities and relationships that an individual encounters daily. H 017 There is no sex difference in the impact that children make daily on the respondent. 17‘ 018‘ 019‘ 020‘ 021‘ 022‘ 023' 024‘ 025‘ 026‘ 027‘ 028‘ 15 The daily impact of children will be greater for women than for men. There is no sex difference in the impact that parents make daily on the respondents. There is no sex difference in the impact that the spouse makes daily on the respondents. There is no sex difference in the impact that the ex-spouse makes daily on the respondents. There is no sex difference in the impact that the in-laws make daily on the respondents. There is no sex difference in the impact that brothers/sisters make daily on the respondents. There is no sex difference in the impact that friends make daily on the respondents. There is no sex difference in the impact that neighbors make daily on the respondents. There is no sex difference in the impact that work makes daily on the respondents. in the impact that respondents. There is no sex difference leisure makes daily on the in the impact that on the respondents. There is no sex difference transportation makes daily There is no sex difference in the impact that health makes daily on the respondents. The daily impact that health makes on the respondents will be greater for women than for men. There is no sex difference in the impact that meals make daily on the respondents. There is no sex difference in the impact that household chores make daily on the respondents. There is no sex difference in the impact that finances make daily on the respondents. There is no sex difference in the impact that pets make daily on the respondents. There is no sex difference in the impact that errands make daily on the respondents. 16 H03“: There is no sex difference in the impact that 1me use makes daily on the respondent. The employment status of the individual involves res- ponsibilities and pressures unique to the status. Employ- ment status may potentially influence the perceived degree of stress reported by the respondent. However, the relation— ship between employment status and sex of the respondent as it influences the total degree of perceived stress has not been previously examined. Therefore, no prediction is made about the direction of the relationship. H : Employment status and the sex of the respondent 035 are not related to the total perceived degree of stress. Social support is utilized by many to mediate between the individual and the stress arising from life events. Hypothesis thrity-six predicts a relationship between the sex of the respondent and the social support that they utilize. H Men and women utilize social support equally in times of stress. 036‘ H36: Women will utilize more social support in times of stress than men. The testing of these hypotheses will provide em- pirical evidence toward answers to the stated research questions concerning the perception of stress in the lives of men and women. 17 TERMS AND DEFINITIONS Throughout this study particular terminology is used. To establish common meanings, the following definitions were established for this research study. Middle Years/Middle Age/Mid-life Conceptual definition: These terms refer to a position in the life cycle as well as a range of chrono- logical ages. This particular position is generally labeled as the years of thirty-five through sixty-five although these vary among researchers in the field such as thirty—five to fifty(Cuber and Harroff, 1965) and twenty-eight to fifty(Frenkel-Brunswick, 1968). Not only does this time period correspond with a chronological age, but also with certain developmental tasks or events which are unique to this stage in life. Operational definition: Participants were selected for this study if the female of a married couple was age thirty-five to sixty-five which met the requirement of being in the middle years of the life cycle. They are placed in this category through self—report on the research instrument(Appendix A). Life Event Conceptual definition: A life event is a situation that arises in daily living within the family or other environment or a situation that occurs at a certain time within the life cycle. Although every individual is 18 unique in his or her experience of the number and timing of the events, most individuals will experience many of the same events as other members of their cohort. Kaplan (1981, p. 70-71) supplements this by stating, "Life events are a change or continuity in personal experience". Change involves normative and nonnormative events while continuity involves chronic problems. Life events are manifested as loss, addition, or redefinition of a social position. "Life events directly influence the experience of subject distress. Some events induce stress by virtue of being jundged intrinsically undesirable. The event reflects the anticipation or experience of the persons inability to fulfill felt obligations or otherwise satisfy felt needs."(Kap1an, 1981, p. 71). Stress may result from the consequences of this situation being perceived as negative. Operational definition: The following are the collective titles for the life events explored in this study: examples of events from each category are included. a. Life Events Internal to the Family 1. death of a member 2. marriage of a member 3. family member moves out of home b. Work 1. major wage earner quits or loses job 2. family member given promotion 0. School 1. family member drops out of school before completing training d. Relatives and Close Friends 1. death of brother or sister 2. death of close friend or confidant 3. member breaks up with close friend or confidant e. Health 1. major wage earner experiences serious illness or accident 2. member experiences memopause 19 3. aged parent committed to institution or placed in nursing home f. Household Finance 1. cut in total income 2. expenses exceed total family income requiring going into debt g. Law 1. family member involved with courts: robbed or assualted: arrested for crime or minor mis- demeanor, jailed, or involved in lawsuit h. Day-to-day, Routine Relationships and Aspects of Life 1. children 2. meals 3. spouse Stress Conceptual definition: Stress has a wide variety of definitions dependent on the particular context in which it is being employed. The definition of stress used in this particular study is one developed by the researcher: it is a compilation of key components of existing definitions. Stress is the loss of desired homeostasis that requires energy which is not readily availab1e(Antonovsky, 1980) and causes disruption of the social unit or system(Monat and Lazarus, 1977). Stress is a personal experience which puts demands on the individual which tax or exceed the available resources(Lazarus and Cohen, 1977: Scott, 1981). Stress for this study is that which is experienced by the individual and is not an attibute of the family. The definition of stress used here is similar to the popular usage and refers to distress. Operational definition: Stress will be identified by the respondent's self-report as to whether an events has caused disturbance. This is measured solely on the per- ception of the respondent,that is, the degree to which he or she is aware of the stress. A response of "not 2O disturbing" is then interpreted as the individuals' lack of perception that stress was experienced. Perception of the Event Conceptual definition: Perception of the event is that which is in the consciousness of the respondent: that of which the respondent is aware, has noticed, detected, or ob- served. To perceive an event is to recognize that it has happened and to recognize the degree of disturbance exper-' ienoed. The perception is the meaning that the individual attaches to the event. Operational definition: The respondent has perceived that an event has taken place if he or she has indicated "yes" or "no" on the instrument. The respondent has per- ceived the degree of disturbance experienced by indicating the degree of intensity on the instrument. Disturbance/Disturbance Conceptual definition: An individual experiences dis- turbance when there is a sense of losing a settled, tranquil, steady state. It is an emotional or mental upset or troub- ling felt within the individual. Disturbance is a percep- tion that reflects the meaning attached to the experience. Operational definition: Respondents will rate the de- gree of disturbance that an event has caused by marking on the questionnaire one of the following: "not", "slightly", ”moderately", quite", or "extremely". Day-to-day, Routine Relationships and Aspects of Life Conceptual definition: Day-to-day, routine relationships 21 and aspects of life are those persons and situations that an individual encounters on a daily basis such as children, pets, and household chores. Unlike the family life events that take place for a particular time period, these relationships are aspects of life which occur routinely. Operational definition: The questionnaire identifies eighteen routine relationships and aspects of life. These are children, parents, spouse, ex-spouse, in-laws, brothers/ sisters, friends, neighbors, work, leisure, transportation, health, meals, household chores, finances, pets, errands, and time use. Impact Conceptual definition: Impact is the effect that relationships and aspects of life have upon the respondent; it is the evaluation of the influence of the relationships and aspects of life. These effects can be positive, neutral, or negative. Operational definition: The questionnaire contains a five point scale on which the respondent will identify the effect of the day-to-day, routine relationships and aspects of life. The scale contains the following effect evaluations: "very negative", "moderately negative", "little effect", "good", and "very good". Coping Conceptual definition: Coping is an "effort to master conditions of harm, threat, or challenge when routine or automatic response is not readily available . . 22 environmental demands must be met with new behavioral solutions or old ones must be adapted to meet the current stress"(Monat and Lazarus, 1977, p. 8). Coping is a process by which stress is dealt with by direct action behaviors or intrapsychic behaviors(defense mechanisms). Coping can be functional or dysfunctional. Coping Resources Conceptual definition: The Boss(l984, p. 21) definition of coping resources is used here. Coping resources are the sociological, economic, psychological, emotional, and physical assets upon which an individual can draw in response to a single event or an accumulation of events. Social Support Conceptual definition: Social support is a social resource. Social support is defined by Cobb(l976, p. 300) as "information leading the subject to believe that he is cared for and loved, esteemed, and a member of a network of mutual obligations". Kahn and Antonucci(1980) concur by including in their definition of social support, affect involving respect, affirmation, love, a sense that behaviors and attitudes are appropriated and also is ‘ aid/assistance given to the individual. Operational definition: Social support utilization will be measured by responses on the questionnaire. The responses will identify with whom the participant interacts in times of stressful events and challenges. 23 Employment Status Conceptual definition: The degree to which an in- dividual is involved in work or in the work place. a. full-time - responent works at paid job for thirty-five or more hours per week part-time — respondent works at a paid job less than thirty-five hours per week unemployed - respondent does not work for pay, is laid off, is looking for work, is retired, or is disabled. full-time homemaker — respondent is not employed outside the home but daily domestic tasks are considered full-time work. Operational definition: Employment status is measured by self-report on the research instrument. ASSUMPTIONS The following assumptions are made with regard to the research study. 1. The participants of this study responded to the questions on the research instrument to the best of their ability and awareness or perception: they answered as completely and fully as possible to the limits of their consciousness. It is also assumed that they answered independently of the influence of another person or persons. Stress is experienced by all individuals. The degree may vary and the individual may not perceive it, but all persons experience stress. Potentially stressful life events of various types are an inevitable part of normal everyday life for individuals. Stress is potentially dangerous to an individual's health, psychologically, emotionally, physically, and socially. Negative stress or distress(as opposed to eustress) should be reduced for optimal health. Participants responded that an event was disturbing when the event had negative effects: positive stress, or eustress, was not identified. The intensity or degree of perceived stress is associated with the context of the event and 24 dependent on the individual: this context is assumed in this study. The middle years do not necessarily contain more or less stressful life events than other times in the life cycle: they offer challenges and opportunities similar to other phases. New challenges and Opportunities may arise during this period of life. The middle years are not considered to be a time of crisis as is popularly reported. Instead, it is seen as a time with events unique to the stage of life: these events do not necessarily result in a higher degree of stress than the stress experienced in another stage of the life cycle. Social support is a resource that can be used to cope with stress: it is an option available to respondents. CHAPTER II REVIEW OF RELATED LITERATURE The review of literature is organized around four major topics. These are: 1)the middle years, 2)stress, 3)sex differences, and 4)coping and social support. The research findings in these areas provide the basis for the research questions and hypotheses of this study. What is included here is a summary of the literature reviewed. THE MIDDLE YEARS A section in the review of literature on the topic of the middle years is included because of the relationship between the events identified as stressors on the question- naire and the events that are considered normative for a person in the middle years of life to experience. Most of the events on the instrument generally take place in middle-age. Examples of these are, "adult child has trouble achieving independence", "death of husband's or wife's parents", and "aged parent becomes seriously ill or disabled requiring direct care". Research has been conducted in the area of the middle years which adds to the understanding of this study. It provides a foundation from which to work and produces information about the normative mid-life experience. 25 26 As stated in the preceding definition, middle-age is operationally defined in this study as the chronological ages between thirty-five and sixty-five. Researchers, however, differ on what chronological ages mark the boundaries of the middle years. Cuber and Harroff(l965) place the boundaries at thirty-five and fifty. Frenkel- Brunswick(l968) states that mid-life begins at about the age of twenty—eight and lasts through the age of fifty with age forty-eight appearing to be the climax of mid- life. More recent researchers have lessened their concern for placing chronological limits to middle—age and centered their definitions on various aspects of the lives of individuals. Happenings in the physiological and psycho- logical realms of persons are considered by some to be more accurate measures of the onset of middle-age. In the physiological realm, Brim(l976) describes basic ednocrinological changes in men that could be considered as a signal for the beginning of mid-life. Physical changes in women, labeled as menopause, are primarily due to changes in the hormonal output. For many women, this signals that they have entered mid-life. Another way to define middle age is to consider certain life cycle markers. These are life events or transitions that are experienctd'at a particular time in life which are classified by society as events that are dealt with by middle aged adults(Neugarten, 1968). These events could relate to the family such as the aging and death of an elderly parent or an adult child leaving home. Careers may also be in transition or involve significant 27 events. Mid-life is a time when men begin to question whether career goals are being met(Rosenberg, 1976). Kimmel(l980) discusses a "career clock" which is society's method of signalling whether an individual is "on time" or "behind" in the world of work. This "clock" becomes very significant to middle-aged men particularly when they have determined that they are "behind". This causes a degree of stress. Neugarten(l968) agrees with the idea that events or situations are better indicators of middle- age when she states, "Middle-aged pe0ple look to their positions within different life contexts - body, career, family - rather than chronological age"(p. 93-94). In this study many of these positions or transitions, which the researcher identifies as life events, are included in the questionnaire. It will be determined if these are significant events which are stress producing for those experiencing mid-life. The stress that arises in the middle years may be unique to the time period. Because of the events or transitions that must be faced, many believe it is a time of crisis or extreme stress(Cuber and Harroff, 1965: Frenkel-Brunswick, 1968: Kimmel, 1980). Others believe it is a crisis for some, but for most it is a time of transition containing both high and low points: these experiences are dependent on each individual's particular situation(Neugarten, 1968: Lowenthal and Chiriboga, 1972: 28 Brim, 1976: Livson, 1976; Rosenberg, 1976: Barber, 1981). For many individuals, the stage is seen quite favorably (Deutscher, 196A). Depending on the individual, certain life events will trigger stress and crisis while for others the events have little effect and they proceed through them in a relatively undisturbed manner. Whether the events are seen as a crisis or not, similar transitions will take place for many persons proceeding through the middle years and some degree of stress will be generated. The timing of these events or transitions is important to consider. Events can be anticipated or unanticipated. Those that are anticipated generally produce less stress. Some examples of these are the launching of children and menopause. Other events are unanticipated and result in higher levels of stress. An example of this is the death of a child. These "off-time" events are labeled idiosyncratic events and are found to have the most impact. That is, they are most stress producing and harder to adjust to or overcome (Neugarten, 1968: Lowenthal and Chiriboga, 1975: Kimmel, 1980). The "on-time" or normative events are generally less difficult to overcome because they are anticipated. Because this study is concerned with the sex differences in perceived stress, it is important to note what are considered normative experiences for men and women in the middle years. Are men and women different in their experience of mid-life? Does mid-life necessarily 29 bring two different experiences for the sexes? Researchers have found some differences in how men and women experience middle-age. Barber(l981) and Deutscher(l964) found that women were more extreme, on either end of the continuum, about the favorableness of the middle years. While men were virtually neutral in their views, women were either much more or much less favorable about the mid-life experience. Although many were very positive, some women were moderately to extremely pessimistic and negative(Lowenthal and Chiriboga, 1972: Barber, 1981). The fact that women are more involved in childrearing and have a greater emotional investment may account for their more intense experience of the empty nest. They seem to experience more loss, discontinuity, and disequilibrium(Barber, 1981). Block(l981) also has found women to experience more loss than men during the middle years. This loss can involve the mother role, a spouse, health, or self-esteem. Women define their age by timing events in their family life cycle(Neugarten, 1968). Although the menopause experience receives much attention as a major mid-life experience for women, they themselves perceive it as relatively insignificant(Neugarten, 1976). Finally, women moving through this time period take on character- istics that are considered to be more masculine(Brim, 1976). They become less sentimental and dependent, and more aggressive and domineering. A shift occurs from an inward trajectory to an outward trajectory. 30 For men, the mid-life experience is somewhat different. Barber(l981) states that the empty nest is viewed as positive,with ambivalence,or is not experienced at all. Brim(l976) states that a crisis is not the emphasis for men but this is a definite time of transition. Some reasons men give for this transiton are: 1)aspira- tions-achievement gap, 2)a resurgence of "The Dream" for their life, 3)endocrine changes, 4)stagnation vs growth, 5)confrontation with death, 6)relationships within the family, and 7)social status/role change. Men do appear to have a decrease in testosterone and corticol production as well as a gradual decrease in androgen secretions. Men at this age compared to younger years appear to be more diffusely sensual, more sensitive to the incidentals, less aggressive, more interested in love than power, more affiliative and more present than future oriented(Brim, 1976). Kimmel(1980) and Robertson (1976) state that careers and occupations are the primary concern of men at this time and Neugarten(l968) states that men define age in terms of their work setting situation. Many researchers are finding that men and women actually become more similar in the middle years. Brim (l976)states that men and women move towards a unisex position and that there is a shift toward sameness. Women take on more male characteristics, such as being more aggressive and domineering, while men take on more 31 female characteristics, such as being more sentimental and having an inward trajectory. Lowenthal and Chiriboga (1972) state that both men and women have more high than low points in their lives. The empty nest appears to bring an increase in life satisfaction for both(Deutscher, 1964): children and the family bring pride and satisfaction. Barber(l981) states that it is not sex that influences the mid-life experiences but it is the situational factors and context that are of the most importance. For both men and women this period does not seem to mark a crisis but more a time of self-reflection(Robertson, 1978: Neugarten, 1976: Rosenberg, 1976). STRESS Stress is defined in many and various ways. Researchers have defined stress differently depending on their profession or discipline and on the nature of their research. The reader must have a comprehensive under- standing of the definitions of stress in order to fully grasp the direction of this research. What follows is a general discussion of the conceptualizations of various researchers and theorists on the topic of stress. Lazarus and his associates have been leaders in the conceptualization of stress. Monat and Lazarus(l977) discuss the importance of recognizing that stress has various forms and it affects individuals in different realms of their lives. Physical stress is that which 32 causes a disturbance in the tissue system of the human body. Social stress involves the social unit or system becoming disrupted or disturbed. And finally, psychologi- cal stress consists of cognitive factors that lead to a stimuli or event being evaluated as threatening(Monat and Lazarus, 1977, p. 2). Threat is an important component in the understanding of psychological stress: it is the evaluation that a situation is harmful. Lazarus(1966) states that before any damage befalls an individual, there is the perception that harm is forthcoming: it is anticipated. This anticipated harm is dependent on the psychological structure of the individuals, their perceptions, learnings, judgements, memories, and thoughts. Each individual has a belief system about transactions with the environment. Coyne and Lazrus(l981) state that stress is a result of transactions between persons and environments. One transaction alone will not result in the perception that a situation is stressful: it is a continual interaction with the environment, which acts as feedback, that results in stress. Because of this constant interaction with the environment, stress cannot be seen as a linear process of causality but a system of feedback loops. One event does not cause stress but information from past experiences returned to the system and transformed results in an event or situation perceived as stressful. Boss(l984) includes this reciprocal and ongoing influence in her model when she diagrams the A, B, and C factors' 33 influences on each other and on the stress outcome as well as the stress outcome's influence on the A, B, and C factors and the influence of the other contexts and responses. Coyne and Lazarus(1981) discuss different types of situational appraising, that is the appraising of potential harm in certain situations. A primary appraisal is made to determine the significance of the environmental encounter on the well-being of the individual. It is deemed irrelevant if there appears to be no significance for one's life and the encounter is then ignored. A second type of appraisal is the benign-positive in which the stress is viewed as beneficial and desirable: it poses no threat or harm. The final appraisal is one of stress: something has happened to cause harm or loss and something warns the individual of potential harm or threat(p. 151). This third type of appraisal is the situation that is of concern to this study and is the only appraisal of concern as the participants are asked the degree to which they are disturbed by an event. Lazarus and Cohen(l977) also define the word stressor or stressor events. These are particular events which "tax or exceed the resources of the system"(p. 109). This definition is used extensively by other researchers and is part of the foundation for this study. The resources on which an individual draws in times of difficulty are extensively used, possibly to the point of depletion. 34 Cooper and Marshall (1977) take the work of Lazarus and his associates and build upon it. They adopt the definition that a stressor is a demand which taxes the system and its resources and that to be stressful an event must be cognitively appraised as such. Again, this is critically important to this study since the res— pondents must perceive the stressor in order to identify it on the research instrument. Cooper and Marshall state that internal and external factors equally influence the stress that is perceived, the reaction to the stress, the coping actions, and ill effects. The following diagram is what Cooper and Marshall call the Person/ Environment Fit Paradigm(p. 6). \ / .\i \\ \ Characteristics . Coping Stress Reaction of Long-term 111 Effects Individual \ Environmental Pressures Figure 3 The Person/Environment Fit Paradigm 35 A balance exists between the aspects of the person and the pressures of the environment. The external pressures can at times be measured with a good degree of success. However, the evaluation that an individual has of an event, or the load of the event, is subjective and very difficult to measure. An event will vary in the degree of stress it causes or the load it creates for each individual(Cooper and Marshall, 1977, p. 3). The following diagram displays three individual experiences of the same event but with different degrees of stress: the loads vary. Load Load Load Stress W Stress \/ Stress Figure 4 The Varying Loads of Stress 36 Hans Selye is also a leader in stress research. In his classic discussion(l956) he defines stress as a state manifested by a syndrome(p. 55). He labels the changes stress induces as specific and nonspecific. Nonspecific changes affect all parts of a system and are generally caused by many agent. Specific changes are caused by one or a few agents and result in alterations in one or only a few parts of the system. The changes are not necessarily harmful or damaging. Distress is harmful: eustress, although labeled as stress, has a positive impact. At no time is there an absence of stress: there is instead, a continuum of the stress experience(p. 33) as diagramed below. Stress i /\ \ / - extremely extremely + unpleasant pleasant Figure 5 A Continuum of the Stress Experience 37 There are periods when stress is at a low point, however, stress is present at high levels when a change is ex- tremely pleasant as well as when an event is extremely unpleasant. The study here deals primarily with the stress that is unpleasant: the respondents are asked to identify how disturbing an event was, which has a primarily negative connotation and suggests to the participant that he or she respond only to those event perceived as negative. The distinction between positive and negative is an important one to be made. Antonovsky(1980) states that the human body has homeostasis maintaining and restoring mechanisms; homeostasis is desirable and constantly sought. Daily, humans experience slight and continual changes as a result of environmental interactions. These disturb homeostasis: a demand is made on the individual which cannot be met at that time. These changes and demands are stressors that disturb the person. The individual responses, according to Antonovsky, will always be a search for homeostasis. This requires energy that the system does not have: additional energy must be called upon. Even the daily responsibilities of life can upset this balance and tax the individual's resources. Although these daily res- ponsibilities are not major events, they do produce stress. This then adds a new dimension to the growing definition of stress presented here. Another area of stress research that composes much of the literature is physical manifestations of stress. Rabkin and Struening(l976) are concerned with the physical 38 damage that results from internalized stress. The greater the number of stressful events and the greater the degree of intensity, the more likely the development of physical illnesses. "Mounting stress precedes sudden cardiac death, myocardial infarctions, accidents, athletic injuries, tuberculosis, leukemia, multiple sclerosis, diabetes, and many minor medical complaints"(p. 1014). This is but one list of the physical manifestations: many more have been associated with stress. Because these physical illnesses linked to stress are not the focus of this study, no additional time will be spent elaborating on them. Many studies have dealt with these physical manifestations and are important for more complete understanding of the stress experience. They are pre- sented here to make the reader aware of the potential harm of stress. These cannot be ignored because they are the results of the stress investigated in this study: they identify one reason why this study is important. Depression and other psychological disturbances are also manifestations of stress. In Ilfeld's(l977) study of 2299 Chicago adults between the ages of eighteen and sixty-five, it was discovered that the more stressors present in an individual's life, the higher the rate and intensity of depression. Depression was highest for situations involving marriage and parenting followed by situations involving jobs, finances, and neighborhoods. Within adulthood, stress perception and response seem to. 39 vary with age. Although major changes and shifts occur at every age and certain stressors will be repeated over the life cycle, there are some differences depending on age(Hefferin, 1981). Young adults rated events as more stressful than older individuals in approximately half of the categories in a study conducted by Horowitz, Schafer, Hiroto, Wilner, and Levin(l977). For young adults, the stressors experienced most often are nutri— tional deficits, rest and sleep problems, sexual diffi- culties, moral-religious conflicts, self-image, divorce, and bereavement(Hefferin, 1981). Although these are the experiences of persons at young ages, they could influence the stressors of those entering middle age. These stressors can neither be predicted nor are they universally experienced. What is known is that what is done in perceiving and responding to stressors in early adulthood shapes the later adulthood experience: earlier decisions and responses affect later decisions and res- ponses. This is important then to consider in this study of the middle years of life. Another point to con- sider is that although there is influence from previous experiences, there are differences with each age. A stimuli is cognitively appraised somewhat differently with each age and the sources of stress vary(Hefferin, 1981). For example, individuals in the middle years will have heightened stress due to the awareness that life and health are finite: these are different stressors than MO younger persons experience and alter the cognitive appraisal. Therefore, in the middle years unique stressors arise and how these are perceived is affected by past experience. The terms stress and stressor are confounded in the literature related to stress. They are often used inter- changeably and need clear definitions and appropriate usage in research. Although it is difficult to keep these terms conceptually independent, this must be done for precise understanding. The stressor as defined in this study is the life event which precedes the experienced stress: the event is perceived as having potential harm. The stress is the loss of desired homeostasis which disrupts the social unit and follows the perceived stressor. A final issue of concern in the literature on stress is that of research methodology. ~Because of the scoPe of the topic, methodology needs careful consideration. A first con- cern when considering methodology is the entity which is being measured and the definition given this. Again, the terms stress and stressor must be completely conceptualized if they are included in the research. This being accomplish- ed, a consideration then is the choice of instrument which is also dependent on definitions. Many researchers choose a life events questionnaire to identify stressors and the stress that results from the perception of these. These life events questionnaires have been critiCized for a number of reasons. First, the questionnaires do not provide information about the individual's adaptation or maladaptation to the stressor. It is not possible to determine if the respondent has adjusted to the changes that resulted from the stressor or not 41 adjusted(Horowitz, et a1, 1977). Questions arise about the validity and reliability of the questionnaire(Mullan, 1983: Rabkin and Struening, 1976). Rabkin and Struening(l976) question the ability of the instrument to completely and adequately assess the stressors and the stress experienced. It must be remembered that stress is broad in scope and an instrument cannot be expected to measure every dimension. In this particular study the domain is perceived stress in family life events. The researcher will not investigate other realms and the instrument utilized focuses on this particular concern. Other researchers would need to care- fully examine the goals of their studies in order to ap- propriately determine the instrument to be used and the sample selected. Mullan(l983) raises additional questions about the reliability and validity of life events questionnaires. A major concern involves the recall of the respondents. Do individuals accurately recall events that have taken place and the date of these? Mullan also is concerned that recall may be selective and respondents only remember and report certain items. Jenkins(et al, 1979) as reported by Mullan, has found that events involving children, death, relocation, and legal issues are most stable in recall while events involving occupation, finances, social-personal experiences, health, and relatives are least stable. Mullan also views the words and phrases used in the questionnaires as vague. Multiple interpretations may be made by the respondent and and this affects validity and reliability. In order to 42 more universally define words and phrases and to aid in recall, Mullan suggests combining the life events questionnaire with an additional research method such as an interview. Measurement is also a concern due to the various ways life events questionnaires are scored(Rabkin and Struening, 1976). Some researchers simply count the number of events identified as stressful: others weight the items. Rabkin and Struening(l976) believe that these various scoring pro— cedures make comparisons across studies very difficult. They are also concerned with confounding variables in stress studies. For example, socio-economic status and ethnicity may affect the results. Certain buffers can act as mediators between the individual and the stress. Many do not and cannot control for this. The Boss model of family stress is an ap- propriate approach because it takes into account these con- textual factors. Mullan(l983) is concerned also with opera- tional confounding in the questionnaire of symptoms and events: many of the events are actually symptoms and symp- toms should not be included in a life events questionnaire. Difficulty also arises in determining if the event has pre- ceeded the symptoms, if the symptoms lead to the event, or both. Also, does co-occurrence influence the overall stress situation and does the undesirability of the events influence reported stress? Mullan states that longitudinal data are necessary to adequately deal with these concerns. Sample selection is another concern of Rabkin and Struening(l976). Some researchers select a particular group such as cardiac patients or psychological out-patients at certain clinics. Collins and Frankenhauser(l978), for 1+3 example, used engineering students. The results of these studies cannot be generalized to other populations because of characteristics unique to these groups. These above con- cerns are appropriate concerns that require attention of both researchers and critics of research. They are necessary con- siderations and are included here because of their importance. Many studies exist that focus on stress and its impact on in- dividuals. It is a difficult aspect of life to study and much care must be taken when conducting research and reviewing reports and studies in order to control, as much as possible, for these problems. SEX DIFFERENCES AND STRESS As already discussed, stress can be experienced in different realms of life, physical, psychological, emotional, and social. Research on the topic of sex differences in each of these realms is relatively limited particularly for general p0pulations and also for healthy adults both physically and psychologically. More is known about the stress experiences of the psychologically impaired than about the well adjusted. Makosky(1982) has noted the lack of research on the topic of sex differences and stressful life events. This points to the appropriateness of this study. Actual physical differences do exist between the sexes in their response to stress. Morse(l982) contends that male and female brain chemistry may be different. Actual perceptions of situations may vary due to these biochemical brain struct- ures. Collins and Frankenhauser(l978) studied male and female engineering students and their responses to stress. Using 44 Stroup's Color Word Test, they found that not only did the females perform better on the task and significantly improve over three trials but the females' heart rates were higher than the males. Males had higher levels of adrenal and corti- sol excretion during the test which was stress producing. Martinson and Anderson(1979) worked in this same area of the male and female physical response to stress. They found that females release less epinephrine under stress situations. Men release more and tend to respond aggressively to stress which may be a result of an increase in the circulating an- drogens. They note that women are more prone to phobias, reactive depression, neuroticism, introversion, and anxiety (p. 91). Women tend to be more inhibited and have a more reactive behavioral inhibition system due to the physical integration of the activity of the hippocampus, frontal cor- tex, and medial septal areas of the brain which could inter- act with the socialization process. Martinson and Anderson (1979) state that more men die from health problems related to stress such as ulcers,coronary disease, hypertension, and colitis. For women, "levels of circulating estrogens can be a factor in the lesser incidence of vascular and hypertensive diseases(p. 92). Women also have a higher density of lopo- proteins which appear to remove cholesterol from the vascular system. These physical differences are included here in order for the reader to understand that many of the social and emo- tional responses to stress are influenced by physical changes and in order to fully understand sex differences and social- psychological stress, physical sex differences need to be identified. 45 Women have been found by some researchers to experience stress uniquely. Morse and Furst(l982) studied women and stress extensively. Their findings were that women exper- ienced more anxiety then men and this could be brought on by external events more so than internal events. That is, happenings outside of the women, such as many of the life events investigated in this study, tend to influence the anxiety experienced. This is complemented by Dohrenwend (1973) who found the women experienced more stress when the events were out of their control. Although studies in the past have shown women to have an internal locus of control, Cellini and Katerowski(l972) have challenged these findings. In their study they found the sample of women to have moved from a position of internal locus of control to an external one. Females were highly external and, over a fifteen year period, there appeared to be an overall change toward a more external locus of control. Perceived control and anxiety in women appear to be related to external happenings. Morse and Furst(l982) support this in their report that an external locus of control is related to increased levels of stress. These findings are taken into consideration in the present study and are included because many of the life events are external to the individual. Martinson and Anderson(1979) considered the emotional and psychological experience of women under stress. They report that women are more prone to phobias, reactive depression, neuroticism, introversion, and susceptibility to anxiety all as a result of stress. Morse and Furst(l982) found women to have exaggerated emotional responses to social stressors 46 such as marriage, moving and divorce. Marriage seemed to stress women due to the changes that it brought, unprepared- ness for the new jobs, lack of financial control, sexual stress, and drudgery of being a housewife(p. 154). Although marriage reduces stress for men, it produces stress for women(Kinzer, 1979). Morse and Furst(l982) discuss other factors that women in families label as stress producing or that influence the stress experienced. The empty nest, al- though not universally viewed negatively, does appear to in- fluence the stress and depression experienced by some women. Menopause is no longer necessarily thought to be negatively experienced: however, physical changes do occur and if they result in a major change, such as a hysterectomy, stress and depression can result. Anxiety, stress, and depression also result for some women when their husbands experience illness, a relatively common experience in middle age. The housewife role can become boring: boredom and monotony are stressors particularly if there is no relaxation or private time(Morse and Furst, 1982). Housewives tend to experience more psychological stress symptoms than employed housewives and men. Studies exist which have resulted in few differences being found between the sexes and the stress experience. Klassen, Roth, and Hornstra(l974) studied the perception of life events as gains or losses and compared men and women. Although there were slight age differences, they found vir- tually no significant sex differences. No overall sex difference was found in the study of noncrisis, nonpsychotic :7 outpatients(Webb, Snodgrass, and Thagard, 1978). The Holmes Rahe Life Change Unit Scale was administered to these subjects with the results showing that men and women experienced the same number of life changing units. Both men and women ranked the same event, change in residence, as the highest. Finally, Dohrenwend(l973) found no difference in men and women in the number of stressful events that had taken place in the last year. Studies exist which report the opposite: that is, differ- ences do exist in the stress experience of men and women. At the Texas Research Institute of Mental Health(Weinman, et al, 1978) seventy men and women inpatients were given a self- weighted life events questionnaire. Men and women showed no difference in the actual number of events they identified. After a count of events experienced by men and a count of those experienced by women, it was determined that no differ- eance existed. However, there was a difference in the degree of stress resulting from these events. Women perceived events as being more stressful. 0n the items of the scale that related to minor flaws and faults to which most people would admit, men had higher reports. However, on the scale that identified major flaws and failures, men scored lower. The researchers also found that men were more eager to be soc- ially accepted and misrepresented or concealed information. The study was conducted with a sample consisting of inpatients and this does present a problem with generalizing to other populations. In a study of the masculine,feminine, and androgynous 48 personality, some sex differences were found related to the stress experienced(Hatzenbuehler and Joe, 1981). For females, the score on the masculine/feminine scale did not predict the level of stress. Androgynous men and women also did not per- ceive more stress in their lives. However, males who scored high in masculinity on the scale reported low amounts of stress. This may not be the absence of stress but a result of a lack of disclosure. The "masculine" men may be less likely to admit to stress. Bradley(l980) studied sex differences and the reporting and rating of life events for healthy and diabetic subjects ranging from age sixteen to eighty-one. One of her basic findings was that women reported that life events were more upsetting and required more adjustment. Women were more likely than men to verbalize about the stress they were ex- periencing. This study, although dealing with a comparison of healthy and diabetic individuals, provides relevant in- formation and support for the study being conducted here. One final study, reported in two different sources, was conducted by Lowenthal(et a1, 1975) and Chiriboga and Dean (1978). Chiriboga and Dean extended the original study longi- tudinally and provided additional information. These re- searchers were concerned with stress and sex differences for persons of various ages. Four age groups were used, seniors in high school, newlyweds, middle aged persons, and older pre—retired and retired individuals. Both reports found that the two younger groups(high school seniors and newlyweds) reported more stress than the two older groups except on 1:9 family related events where the middle-aged group reported more. Men and women in the middle years showed no differ- ence in the areas of being overwhelmed or challenged in re- gard to family and marital events. Women, overall, exper- ienced more interpersonal stressors in such areas as marriage, family and non-family relationships. Men reported more stress over legal events. The most stressful events and issues for men, in order, are education, work, residential changes, marriage, military, and family. For women, the most stress- ful areas are, in order, education, family, health, dating/ marital relationships, and work(theirs and their husbands). Lowenthal and her associates(l975) reported at length the differences between the age groups: their findings are re- ported here since a similar group of middle aged persons is used in the present study. For newlymarried, educated men, stress levels were high: they were more involved in the socio-political world. Young women showed increased stress as socio-economic status increased but did not show any difference as education rose. Older women with higher educa- tional levels showed less stress. More stress was exper- ienced by young men with strong masculine self-images. The more social roles the older men had, the more stress was ex- perienced. Young men with the potential for complex life styles and older men with complex life styles had higher stress levels. Less complex life styles for the older women reduced stress. These women were generally less intelligent, showed a flatness to life and were family-centered. Stress was overwhelming for young men and older women while for 50 older men, large amounts of stress were seen as challenging. Lightly stressed men and older women felt lucky while younger women with little stress were preoccupied with it. Older women had fewer resources than the other groups to combat stress. Middle aged men reported concern about financial resources for retirement:they were anxious and less hopeful about the future. The younger groups reported more positive stress while the middle aged group reported more negative stress. Overall, the younger groups reported more stress. Work was the most stressful area for men while the most stressful area for women was family relationships. Death was reported equally stressful among all groups. The older two groups had much stress focused on others: this was responsible for four fifths of all stress identified. The middle aged women were twice as likely than the older women to be stressed about children and one third of middle aged women reported stress related to their spouse. For middle aged men, lack of job promotion or salary increase was a prevelant stressor. Although research is relatively scarce for the topic of this study, the above research does give a direction and insight into sex differences in perceived stress. Some studies have found there to be few differences and more similarities in the stress experience of men and women. However, it is the view of the researcher that there exists adequate evidence supporting differences. 51 COPING AND SOCIAL SUPPORT Coping is defined differently by researchers and a variety of definitions exist. For this research, coping is taking direct action or employing palliative modes to master the conditions of harm, threat, or challenge when regular resources are not available(Monat and Lazarus, 1977, p. 8). Cognition is involved in coping. Lazarus, Averill, and Opton, as referred to by Belle(l982), deve- loped a cognitive mediation model which involved three processes: 1)the recognition and interpretation of a threat inherent in a situation, 2)the perception and appraisal of possible coping strategies, and 3)a recon- sideration of the initial appraisal of the situation(p. 182). The approach to coping must first begin with cognition or perception of the coping situation. This is true for the stressful event also. The event must be first perceived as stressful before a change or stress response takes place. The same is true for coping: the perception must come first before a response takes place. The coping resource of concern for this study is social support. Cobb(l976, p. 300) defines this as "in- formation leading the subject to believe that he is cared for and loved, esteemed and a member of a network of mutual obligations". Aid can be expected and affect received(Kahn and Antonucci, 1980). This describes an interaction with people that surround the individual. This study is not concerned with effectiveness of social support as a coping mechanism so much as male and female 52 differences in the use of social support. However, the effectiveness of c0ping though social support is important to know. Some degree of effectiveness must exist for the individuals to choose to employ it. If it were not effective, individuals would not choose it in order to cope with stress. Social support can be gained from large numbers of people who surround an individual. The person closest to the individual is important for support but many others can give aid. When social support becomes weak or low, stress and strain increase(Abbey, Abramis, and Caplan, 1981). This means that even in neutral circumstances, if social support is removed, strain and stress will increase. In situations when stress is present and social support is weak, it can actually lead to the situation being more stressful. Chan(l977) states that how one responds to stress is dependent on the perception the individual has of the availability of social support. Chan found that social supports helped individuals with social losses whatever they might be. Cobb(l976) also draws this conclusion in her review of life cycle stressors. She found that social support can act as a protection against the crises of life such as a low birth weight child, death, depression,'arthritis, tuberculosis, and alcoholism. Social support can also reduce the amount of medication required, accelerate recovery and facilitate compliance with medical regimes"(p. 310). 53 Warheit(l979) looked at stressful events and depression and found these to be positively correlated. But for individuals who had family and friends around less severe depression was experienced: the scores on a test of depression were higher for those with stressful events in their lives and no nearby friends. The presence of a spouse correlated with lower depression levels. This spousal support was found significant in a study by Kessler and Essex(l982). Married persons in their study were less depressed due to less exposure to stressful events and more resilient to the emotional damage of stress. The most powerful resource in coping with stress was a confiding, intimate relationship which in most cases is the spouse. A limited number of studies have been conducted that consider sex differences for coping in general. Ilfeld (1980) conducted a study of 2299 Chicago adults and their c0ping styles for combating stress from social roles. Three styles were found: 1)taking direct action, 2)rationa- lization/avoidance of the stressor, and 3)acceptance of the stressful situation without attempting alteration. Usually the respondents did not solely use one style but a variety of responses. These responses mediated between the life events and emotional distress. The sex of the respondent did not predict the coping responses but there was a tendency for women to seek outside help, and use rationalization/avoidance of the situation. That is, women used social support somewhat more than men. 54 Parasuraman(182) investigated coping with job stress. As in other research, it was found that not all coping is functional and individuals can respond in adaptive or maladaptive ways. Adaptive coping involves dealing directly with the stressful situation by seeking and implementing solutions. Maladative coping deals with stress by employing emotional, defensive, or self- protecting approaches. Parasuraman studied managers in a large Midwest utility company and found that the higher the degree of stress, the more these managers used mal- adaptive coping responses. However, female managers tended to use adaptive behaviors more than men which disconfirms the stereotype that women managers are emotional and unable to c0pe with the pressures of the job. Pearlin and Schooler(l978) have studied the coping process extensively. Their findings conflict with Parasuraman's results. Pearlin and Schooler state that there are two types of coping resources: 1)social resources which are interpersonal networks and 2)psycho- logical resources which are personality characteristics. In addition to these resources there are three types of responses: 1)responses to modify the situation directly which involves altering or eliminating the source of stress, 2)responses to control the meaning of the problem which involves positive comparisons and selective ignoring, and 3)management of stress which is accomodating the stress without being overwhelmed by it. The researchers found that men have more psychological attributes or 55 employ responses that inhibit the stressful outcomes: they are more direct in acting on the stress. Women were found to use more selective ignoring which ultimately leads to a worsening of the situation. In addition to the above cited research, a small number of studies were found linking social support to the sex of the individual. Phillips(l981) studied the networking characteristics of men and women. Men reported more happiness as their social network increased. Male happiness could also be predicted to a lesser degree by the number of areas in which social activities took place, such as the community, work, and church. The actual number of social activities best predicted the happiness of women. Kin involvement was higher for the women than nonkin involvement. When men lose support received from the work place, depression and psychosomatic symptoms increase(Holahan and Moos, 1981). These symptoms did not arise for men when family support was decreased. Depression was related to both decreases in family support and decreases in work support for the employed women while for the unemployed women, only loss of family support was related to depression and psychosomatic symptoms. For men, work support is particularly important(LaRocco, et al, 1980: House, 1981) and family support is more important to women(Brown, et a1, 1975: Paykel, et a1, 1980: Roy, 1978). In sum, coping is a part of the stress process. There are a variety of ways to cope with a stressful situation: 56 the one of concern for this study is social support. Social support has been found to lessen the effects of stress. The marriage relationship appears to be a signi- ficant support. Generally, studies on coping have found male and female differences. Although Parasuraman's study on managers found women employed adaptive behaviors to 00pe with stress, Pearlin and Schooler found women used selective ignoring, which is maladaptive,and men faced the situation directly. Ilfeld found women also to use avoidance and rationalization of the situation. Women in his study employed more social support for coping by seeking outside help. Women find the support of family as more important while men need the support from their work place. SUMMARY The four areas of the literature review are middle age, stress, sex differences and stress, and coping and social support. In the Boss Model(figure 2) the middle years are part of the deve10pmental component of the external context. Middle age involves normative life cycle events experienced by most persons. Differences appear to exist in events which are experienced. Stress is the core of the Boss model. Stress involves the inter- action between individuals and the environment. Bos .M‘ ~L. _.“ .1 _identifies this in her model by including internal and I”..- ‘u‘- w“-‘mum ,1 m ‘ 4 "Ara-h. external contexts and responses as impacting upon the 57 stress process as well as the process impacting upon the contexts and responses. The appraisal of the stress resulting from an event is the C factor in the Boss model. Sex differences are evident in the physical, psychological, and emotional aspects of an individual. Women tend to report more stress in their lives. Coping and social support represent the B factor of the Boss model. Social support acts as a resource to diminish stress. The coping resources(B factor) are influenced by the perception of the event(C factor) because the resources must first be perceived before they can be utilized. It appears that a tendency exists for women to seek more outside help or social support. CHAPTER III METHODOLOGY RESEARCH DESIGN This study has an ex post facto design. Sample survey research is the methodology employed. The inde- pendent variable is the sex of the respondent which deter- mines, by self—selection, the group in which the respond- ent is placed. This variable will not be manipulated and there is no direct intervention by varying either the dependent or the independent variables. The respondents answered a mail questionnaire regarding their perception of particular life events. The events were first‘ identified as having occurred or not occurred and then the degree of disturbance experienced was labeled. The design then takes on this schematic shape X Y Y Y 1 1 2 3 X2 Y1 Y2 Y3 where X1 is women and X2 is men(neither of these is a control group) and Y1 is the number of family life events identified, Y2 is the degree of disturbance experienced, and Y3 is the degree to which social support is used by men and women. The X's represent the independent variable of sex of the respondent. The Y's are the 58 59 dependent variables. This then diagrams the two groups, men and women, and the outcomes of the three dependent variables. This study does not ignore the confounding variables which make these two groups different such as education or sex role socialization but recognizes that they exist. These variables make for differences and this study hypothesizes that these differences then influence the perception of stress. The male and female groups then are comparison groups as opposed to control and experimental groups. Although there are limitations to this type of design including the inability to mani- pulate the independent variable, the lack of power to randomize group membership and the risk of improper interpretations(Kerlinger, 1973), this design lends itself particularly well to this study. Figure 6 depicts the design for analysis. THE RESEARCH INSTRUMENT The instrument used in this study is part of the questionnaire used for the Agricultural Experiment Station's regional research project, NC 164. The instrument was developed by the committee members of the project. It is thirty-two pages in length and includes a variety of components which represent the interests of the research committee. Approximately one hour is required to complete the questionnaire. It was subjected twice to pretests during its construction to minimize 6O Dependent Variables Number and Nature of Family Life Independent Events Variable Intervening Variable Sex Employment Status of Respondent ' Perceived Degree of Stress in Life Events Utilization of Social Support Day-to-day, Routine Relationships and Aspects of Life Figure 6 Design for Analysis 61 measurement difficulties. The entire instrument is not used in this study: only a portion is used and only data gathered in the State of Michigan are included. Appendix A contains sections of the questionnaire used. Family Life Events Scale The Family Life Events Scale was developed by the research committee. It is based on a scale developed by McCubbin, Patterson, and Wilson(l981) entitled FILE(Family Inventory of Life Events). A copy of File is located in Appendix B. The scale involves forty-eight potentially stressful family life events or changes which are both expected and unexpected. The respondents were asked to identify whether each of the events had happened in the last three years and in the past year. For the study at hand, only the three year response was considered as it included more time and also included the last year. The participant also identified the degree to which the event was disturbing. The scale ranged from one, "Not disturbing", to five, "Extremely disturbing". The scoring of the scale is in two parts. The first part involves the "event has taken place in the last three years" section and is a measure of the number of potentially stressful events. This section has a "yes" or "no" response. The researcher counted the total number of "yes" responses for both men and women over the forty- eight items. The "yes" responses were also counted for each particular category of events, defined by the nature 62 of the events. [These were counted for both men and women. Finally, each of the individual events had counts run on them for men and women. The second method of scoring involves the degree of disturbance and is a measure of the degree of stress experienced. The degree was measured on a one through five scale. A mean score was calculated for each respondent over all of the events which the respondent had identified as occurring. Each participant then had a mean disturbance or stress score. Also, within each of the categories of events grouped by nature of the events, a mean was calculated for each individual. Finally, for each single event a mean score was calculated for men and a mean score calculated for women. Many of the events were not experienced by in- dividuals and therefore they responded "no" as to whether the event had happened in the last three years. They, then, also left the "degree to which the event was disturbing" scale blank. These events were simply dropped from the calculations and did not enter into the scoring procedures. McCubbin, Olso, and associates(l983, p. 246) have tested FILE for reliability and report the Alpha Reliability and Test/Retest Reliability correlations by category on the scale(see Table 1). A concern has risen over one item on the scale. The first event listed is death of a member. It was the intention of the committee to obtain information about the death of a member in the immediate family. Items 63 Table 1 Alpha Reliability and Test/Retest Correlations for FILE Alpha Test/Retest Reliability Reliability FILE(stress) .81 _ .80 Strain .72 .73 Marital .16 .68 Pregnancy .24 .84 Finances .60 .64 Work .55 .80 Illness .56 .66 Losses .34 .71 Transitions .52 .72 Legal .66 .83 "2 "aa", and "ee" on the questionnaire ask for information involving the death of an extended family member. The researcher believes that some respondents may have in- dicated the death of the same family member in more than one item on the instrument. The instrument should be corrected for this problem in future use. Day-to-day, Routine Relationships and Aspects of Life Rosalie H. Norem of Iowa State University developed_ the Dayhto-day, Routine Relationship and Aspects of Life scale particularly for this project. It involves eighteen aspects or relationships routinely experienced in the respondent's daily life. The rating scale applied to each item ranges from one to five with one stating that the event or relationship had a "very negative" impact on the respondent's life to five stating a "very good" impact. The scale was scored by each event or relationship separately. A mean score was calculated for both men 64 and women on each item. These mean scores of impact were then compared between men and women. No total score over all of the items was calculated as each was unique in itself: no one number could describe the entire scale. The items were too diverse to combine into one measurement. Missing data were not included in these calculations. If the respondent did not identify the impact of the item, the case was dropped from the calculations. At the time of this writing there is as yet no information regarding the reliability and validity this scale. It is a new scale which will need further testing over time. The scale was pretested twice with the rest of the instrument but precise figures are unavailable. The reader should take this into considera- tion when examining the results using this scale. Social Support Utilization Scale The Social Support Utilization Scale was deve10ped by the research committee. It is a shortened version of a scale developed by McCubbin, Larsen, and Olson(l982) entitled, F-Copes. A c0py of F-COpes is in Appendix B. The scale consists of seventeen items which identify supportive individuals that the respondent may contact when stressed. The scale ranges from one to five with one being "never" and five being "more than once a week". The scoring of the scale was done in two parts. First, a mean was calculated for each social support item 65 for both men and women. A second mean for both men and women was calculated including all of the items on the scale. This yielded a mean social support usage score. For the calculation of this total social support score, missing data were not included. If the respondent left an item blank, no score was given to that particular event and it was not included in the calculation. The case, however, was not thrown out if an item were left blank. In many cases, the researcher believed that the respondent was indicating that they did not use that particular social resource when the item was blank. In the calcula— tion of male and female means for each item, the case was dropped from the calculation if a participant did not respond to a particular item. McCubbin, Larsen, and Olson(l982, p. 260-261) have calculated Alpha Reliability and Test/Retest Reliability correlations(see Table 2) for F-COpes. Table 2 Alpha Reliability and Test/Retest Correlations for F-Copes Alpha Test/Retest Reliability Reliability F-Copes .86 .81 Spiritual .81 .95 Social Support .84 .78 Reframing .82 .61 Mobilizing Family .71 .78 Passive Appraisal .64 .67 66 An alteration was made in the original scale when calculating the total mean social support score. Two items were completely eliminated from the calculation. Item "h", "facing problems 'head on' and trying to get solutions right away" and item "i", "watching television" were not considered to be social supports and were not consistent with the other events. These seem to be identifying other kinds of resources and were not appropriate to be included in a total social support score. A mean was calculated for both men and women on each of these items separately, however. THE RESEARCH SAMPLE The population selected for this study was identified by the research team of project NC-164 entitled "Stress, COping, and Adaptation in the Middle Years of the Family", supported by the Agricultural Experiment Stations of the North Central Region of the United States and the Cooperative Research Service of the United States Department of Agriculture. For the analysis reported here the population selected was defined as families in the middle years of life in the State of Michigan*. For the initial sampling frame, a middle age family was defined by the research team as a two-parent family where the mother was between the ages of thirty-five and fifty-five, with at least one adolescent child in the *The same procedures for sample selection were used in all nine states involved in the NC-164 project. 67 home. Because of the magnitude of the task and the ex- pense involved, a sample was not drawn from the entire State of Michigan but representative areas were selected. Four rural counties and one Standard Metropolitan Statistical Area(SMSA) were selected to represent the population of the state for the sample which was randomly drawn. Appendix C explains the procedures used to select the rural counties and Appendix D explains the selection process for the SMSA. These areas were selected essentially on the basis of education and income levels. Alpena, Charlevoix, Dickinson, and Hillsdale were the rural counties selected and the SMSA was Flint. The researchers involved in the regional project had special interest in rural families in keeping with the goals of the Agricultural Experiment Station. Therefore, rural families are over-represented and the sample is not pr0portionally representative of the rural and urban p0pu1ation in the State of Michigan. In the research at hand, rural and urban differences were not the focus and were not considered. The Donnelly Corporation was contacted by the research committee for help in selecting a sample. This corporation collects and distributes lists of names for businesses and other organizations across the nation. A list of names and addresses was provided for the research committee which was to meet the established criteria for the sample in the SMSA and rural counties. From this 68 list, a random sample was drawn. Details of the procedures used are in Appendix C. It was decided that 900 households would be selected. The research team set as its goal, 100 husband and wife response sets from three types of households, Urban House- holds, Rural Farm Households, and Rural Nonfarm Households. Three hundred families were randomly selected from Flint and 600 families from the rural counties. It was impossible to know before the questionnaires were completed how many farmers would be represented. It was hoped that the response would show a similar count of farmers and rural nonfarmers. The number of families selected from each rural county was dependent on the population of the county and the number of farms therein with equal weights given to these. Appendix C explains this procedure in greater detail. The following numbers of families were selected for each rural county, Alpena, 138: Charlevoix, 84: Dickinson, 48: Hillsdale, 330. As the surveys were returned, it became evident that the list that the Donnelly Corporation had provided was different from that which was requested. Some of the families had no teenage children in the home and many of the wives were older than the requested limit. A decision was made to revise the requirements for inclusion in the sample. The definition was expanded to include mothers who were between the ages of thirty-five and sixty-five. The family must have had at least one child but this child 69 did not need to be a teenager nor be living at home. (Much of that which is reported in this section was originally compiled by Smith, 1984) Response Rate A crucial component of survey research is the response rate as it reflects the representativeness of the sample to the population. Nine hundredfcouples, or 1,800 individuals were mailed questionnaires. Out of the 900 couples mailed questionnaires, 784 couples were potential participants. This number was arrived at as follows: 900 Sample couples mailed questionnaires -28 Sets undeliverable -34 Researchers informed that couples did not fit criteria -57 Couples completed questionnaires but were too old 784 Potential couples Of the 784 potential couples, 232 couples returned the questionnaire completed. Added to this are 42 individual cases in which one of the spouses did not complete the questionnaire. Eleven men and 31 women made up these forty-two cases. The spouse either returned no questionnaire or returned it partially completed. This then is a response rate of thirty-four percent. It is believed that many of the potential respondents may have self-selected themselves out when they realized that they did not fit the original criteria. After the questionnaires were mailed, it was dis- covered that the list(a "B" and "C" list) supplied by 70 the Donnelly Corporation did not meet the original criteria for the age of the wife and the presence of a teenager in the home. After the questionnaires were returned, the Donnelly Corporation provided an "A List", the highest grade list compiled, for the researchers. This list included the names of persons who did meet the original criteria. Approximately seventeen percent of the sample selected for the project was on the "A List". Of the "A List" sample couples, 42.4% completed the questionnaire. Since this "A List" is the most accurate portion of the sample list, the 42.4% rate may be the best estimate of the actual couple response rate. A response rate of approximately one in three is relatively low. There are a number of factors that im- pacted upon this rate. The Donnelly Corporation's original list did not meet the sample selection criteria. Had an "A List" been prepared initially, a higher response rate would have been likely. Perhaps a large influence on the response rate is the length and complexity of the questionnaire. It was determined by the pre-test that an hour or more was needed to complete the thirty-two page instrument. The NC-164 committee members were aware of the potential risk of low return rates but be- cause of conceptual and methological issues that were involved, they decided that the risk was worth taking. 71 Characteristics of the Sample The sample selected for the NC-164 project has certain demographic characteristics. Knowing these characteristics will allow the reader to judge to what extent the results of this study can be generalized to middle-aged Michigan residents. It must be remembered that this sample does not have the same proportion of rural and urban participants as found in the population of Michigan. The intent of the researchers on the project was to over-sample rural residents. In fact, 71.6% were from the rural counties and 28.4% were from Flint: 9.6% of the population were farmers and nearly 100 percent of these were from Hillsdale County. These proportions do not represent the actual proportions in the State of Michigan. The age range for men was thirty-one to seventy- mwm ,_.._, -w.‘ three with a mean age of 48.1. For the women, the age range was thirty-five to sixty-five with a mean age of 45.1. These means are appropriate for this study. They faIIwin the middle of the age range identified for the age of the wives. Originally, the researcher planned to control the age factor in order to remove those outlying cases deemed as "too old" or "too young". This would cluster the ages around the above cited means. A number of reasons kept the researcher from limiting the ages of the respondents who could participate. The size of the sample for both men and women at the extremes of the age 72 range was small. Approximately four percent of the men in the study were older than sixty-five and less than one percent were younger than thirty-five(see Appendix D for actual sample ages). It was determined that these would not significantly affect the results of this study. By eliminating a husband or a wife from the sample, the results could have been significantly influenced. This study has as one of its purposes comparison of men and women on a number of events that have taken place. The researcher believes that when both husband and wife are included in the calculations, this controls the context of the event to a degree. The husbands and the wives generally encounter events together and should be similar in recording those. Any differences that then arise are more likely a result of male and female differ- ences and not because the events were or were not taking place for separate independent individuals. The researcher does realize that some of the events would not take place for both persons. That is, the event listed only identifies one of the spouses as its focus. An example of this would be "death of husband's or wive's parent". This could be true for one spouse and not the other. These events, however, are few and do not have a major impact. It is true that an entire couple could have been eliminated from the study because one of the spouses was too old or young but the response rate was already low and the researcher wanted to retain as many participants as possible. 73 Another problem with limiting the age was determining a new age range. As previously discussed, researchers vary on the chronological age range for the middle years. Since there is no agreement among researchers in the field, the researcher for this study would be making an arbitrary range. As cited in the literature review, some researchers believe that it is not chronological age that determines the middle years but the experiencing of certain life events, changes, or transitions. In this case, those who did not experience the family life events on the questionnaire assumed to be mid-life experiences, dropped out of the statistical calculations. That is to say, if they did not experience mid-life transitions, they did not report them and therefore did not contribute to the calculations. At the time of this decision, the data were examined. It was found that events assumed to be mid-life experiences were happening through- out the age range depending on each individual family. For example, age when children were born and how many children they had played a part in this. Therefore, these mid-life events take place over a range of time depending upon the family, and to eliminate a family due to age would significantly alter the results of this study. For these reasons, the entire sample was used. The average number of years married was twenty-four years. The average number of years lived in their communities respectively WEE; 25.2 years for the men and 23.7 years for the women. Sixty-one percent of the families had two or three children, twenty-seven percent 74 had three or four children, five percent had a single child and six to seven percent had six children or more. The average number of children living at home was 1.7%. See Appendix D for the exact numbers of children per family. The\employment statuses of the men and women are found in Tables~3Mandm4j Fifty-five percent of the women were employed either full—time or part-time while forty- three percent were not employed for pay outside of the home. Twenty-four percent of the men were not employed while seventy-two percent were employed full-time(A summary of the occupational classifications can be found in Appendix D). II The average education level for women was 12.9 years. The years of school completed ranged from five to nineteen years for women. The men's educational range was six to twenty-seven years with an average of 13.1 years of school completed. For the urban sample, ninety-one percent of the women and eighty-one percent of the men completed four years of high school or more. For the rural sample, eighty-eight percent of the women and eighty- five percent of the men reported completing four years of high school or more. The racial identification of the rural sample was as follows: ninety-seven percent of the wives and 95.1 percent of the husbands were white, 0.6 percent of the women and 1.2 percent of the men black, and 1.2 percent of the wives and 3.7 percent of the husbands classified 75 Table 3 Women: Employment Status N = 263 Status Percent Employed full-time 32 Employed part-time 23 Unemployed 3 Homemaker 33 Retired 5 Disabled 2 Other 1 Missing Data 1 Table 4 Men: Employment Status N = 243 Status Percent Employed full-time 72 Employed part-time 1 Unemployed 8 Homemaker O Retired 12 Disabled 3 Other 1 Missing Data 2 76 themselves as other. In the urban sample, 95.4 percent of women and 92.3 percent of men were white, 4.6 percent of the women and 6.2 percent of the men black, and none of the women and 1.5 percent of the men classified them- selves as other. Minorities are underrepresented, particularly in the urban sample. The large portion of rural participants account for some of the overall underrepresentation of minorities because rural communities in Michigan generally have smaller minority populations. However, this still does not account entirely for the underrepresentation. The religious preference of the respondents is as follows: sixty-seven percent of the women and sixty— three percent of the men Protestant, twenty-three per- cent of the women and thirty percent of the men Catholic. Combined, these two groups account for more than eighty- five percent of the sample. Table 5 displays the mean and median income levels of the families in this study. Both the husbands' and the wives' reports of family income are includedhere as there are differences in the responses. Women generally reported lower incomes than men. Urban incomes were also higher than rural incomes. Table 5 Family Income for 1982 Mean Income Median Income Men Women Men Women Total Sample 35,087 32,384 30.047 2999911 Urban #29770 39,769 389500 34,333 Rural 32,154 29,857 28,000 27,013 TECHNIQUES OF DATA COLLECTION The method of data collection for the NC-164 project was survey methodology and incorporated Dillmanls(l978) survey methodology. In February 1983, packets were mailed to the sample. Within the packets were two questionnaires, one identified as husband and one as wife, two postage paid return enve10pes,which were to encourage independent work on the questionnaire by husband and wife, and a letter of introduction explaining the research project and requesting the family's cooperation. A follow up postcard was mailed one week after the packet was mailed to encourage the family's participation. A second set of questionnaires was mailed one month after the initial mailing to those families who had not yet res- ponded. During the summer and fall of 1983 and winter of 1984, coding took place. Details of the coding are in Appendix E- The data were then entered into computer files at the Michigan State University Computer Center 78 in the spring of 1984. Data were thoroughly cleaned during the summer of 1984 by the research supervisor. An SPSS system file was created and frequencies on all variables were run. DATA ANALYSIS The independent variable of concern in this study is the sex of the respondent which is a nominal variable. There are two types of dependent variables. The first is an ordinal variable which involves the "yes" and "no" identification that an event has happened. The variables of "degree of stress" and "social support usage" are at an ordinal level, measured on five point scales where the answers of "extremely" or "weekly" imply a higher ranking than "quite" or "monthly". Because one cannot be sure that the difference between each of these categories is equal, the data cannot be considered as interval data. However, research literature supports the validity of treating ordinal data as interval data(Baker et al., 1967: Bohrnstedt and Carter, 1971: Labovitz, 1970, 1971) for the purpose of statistical analysis. Labovitz(l970, p. 515) states that the advantages of considering ordinal data as interval are: "1)the use of more powerful, sensitive, better deve10ped and interpretable statistics with known sampling error, 2)the retention of more knowledge about the characteristics of the data, and 3)greater versatility in statistical manipulation". "The absence of interval data should not prevent one from 79 proceeding with model testing or parameter estimation" (Bohrnstedt and Carter, 1971, p. 90). T-tests were the primary statistic used in testing the hypotheses of this study. According to Baker, Hardyck, and Petrinovich (1967, p. 20), "The percentage of t's reaching the theoretical five percent and one percent levels of signi- ficance is not seriously affected by the use of non-equal interval measurements." A final statement from these researchers is, "The present findings indicate that strong statistics such as the t-test are more than adequate to cope with weak measurements and with some minor reserva- tions, probabilities estimated from the t distribution are little affected by the kind of measurement scale used"(p. 20). Two-way ANOVA was also used to analyze the data. This statistic tests the variances of the means of the dependent variable, perceived degree of stress, for the independent variables. Each independent variable, employ- ment status and sex of the respondent in this study, is tested for significance and an interaction effect is also examined. This interaction effect involves considering the means of the dependent variable when both independent variables come together to form sampling groups. Two-way ANOVA was selected instead of multiple regression because multiple regression is best used with interval level independent variables. The sex of the respondent and the employment status are categorical or nominal variables. The tests are very similar except for the level on which the independent variable is measured. 80 Each hypothesis will now be discussed and the appropriate statistic(s) identified to test each. H1 - The variables in the hypothesis are sex of the respondent(discrete, nominal) and the number of stressful life events. Three different statistics were used to obtain comprehensive understanding of these data. A contingency table was used to examine each individual life event(whether or not it has happened in the past three years) and the sex of the respondent. A Chi- square test was run on this table. The entire set of responses, as a whole, was analyzed by use of a t-test. H2 - H8 - The events are categorized under different areas of life such as "School", "Work", and "Health". Each of these categories was analyzed to determine the sex differences for each category. For this, a t-test was utilized. H - H16 — The independent variable is the sex of the 9 respondent and the dependent variable is the degree of stress experienced for an event. Each event was considered independently, then by category or nature of the event and finally as an entire set. T-tests were run on these in order to determine the degree of significance between .males and females. - H - The independent variable is sex of the H17 34 respondent and the dependent variable is the impact of each day—to-day, routine relationship and aspect of life. The day-to-day impact includes eighteen routinely ex- perienced relationships and aspects of life. The impact 81 of each of these is measured on an ordinal scale with the range from "very negative" to "very good". Each aspect and relationship was considered independently: data were considered as interval. A t-test was run on these in order to determine the degree of significance between the male and female groups. H35 - This hypothesis has two independent, nominal variables, sex of the respondent and employment status with the dependent variable of degree of stress experienced. A two-way ANOVA test Was the statistic used for this hypothesis. H36 - The dependent variable is social support. The scale measuring social support usage is a ordinal scale. Each source of support was analyzed individually as was the entire set of data. The t-test was utilized in both situations to test for significant group differences. T-tests were used for testing many of the hypotheses in this study. A z-test could legitimately be used since the sample size is greater than thirty. The t-test should be used when considering ordinal data as interval data because the t-test is a more conservative estimate: it uses a more conservative sampling distribution. The t-test is used to determine if there is a significant difference between two groups. The groups in this sample are considered to be independent of one another. When considering ordinal data as interval data, Baker, Hardyck, and Petrinovich(1967) suggest that group sample sizes be equal for t-tests. However, in this study group size is 82 rarely equal but because of the large number oerespondents in each group, the t-test is appropriate by retaining its conservative nature. Although this will influence the results slightly, these influences will not be significant in terms of the SPSS procedures and results: SPSS recog- nizes that in the social sciences, groups are not always equal and the program accounts for some discrepancies in group size. ANOVA(Analysis of Variance) was used to test certain hypotheses. Two—way ANOVA was utilized since two independent variables are being considered. The fixed effects model was used for the study. It is used when the researcher wished to make inferences to only those categories considered and not to other categories that are not examined in the particular analysis or study. The SPSS programs assumes this model in its ANOVA pro- cedures. Two-way ANOVA is designed to be used when the frequencies within each cell are the same. SPSS recognizes that in the social sciences this is not always possible and it corrects for this by providing an option(the Hierarchical Approach) on the ANOVA program. This option was utilized for this study. ANOVA considers the effects of each of the independent variables on the dependent variable and also considers how the two variables interact to influence the dependent variable. The purpose of ANOVA is to consider the means of the dependent variable when it is placed in a particular group of the independent variables and to analyze the variance between these means. 83 The ultimate goal is then to determine by examining the variance if the groups are significantly different. SPSS(Statistical Package for the Social Sciences) was the software utilized for the computer analysis. The data for this study were entered into a computer file at the Michigan State University Computer Center. The researcher had a tape made from that of the NC-164 project which included only those variables of importance to this study. It should be noted that nine states are involved in the NC-164 project, Indiana, Iowa, Kansas, Kentucky, Louisiana, Michigan, Minnesota, Missouri, and Nebraska. For the study at hand, data collected in Michigan are the only data used. CHAPTER IV RESULTS This chapter contains a report of the findings for this study. Each research question is stated, followed by the hypotheses developed from that question and the results of the testing of each hypothesis. The chapter includes results of the statistical tests selected to examine the hypothesis. In testing many of the hypotheses the total number of respondents, 548, could not be in- cluded. In some procedures where events are involved, certain cases were dr0pped because the respondents did not report the necessary information to be included in the calculations. The n for each statistical procedure is reported in the appropriate table. The previously discussed statistics provided the needed information to accept or reject the null hypotheses. However, in certain situations, more information was desired by the researcher and additional t-tests and Chi-squares were utilized. These statistics provided supplemental and significant information regarding the hypotheses being test. These additional statistics are identified for each related hypothesis. 84 85 RESEARCH QUESTION ONE Do men and women differ in perception of the number and nature of family life events experienced? The following hypotheses were generated for this research question. H : Men and women will perceive that an equal number- 01 of potentially stressful family life events has taken place. H1: Women will perceive that more potentially stressful life events have taken place than men. For this hypothesis a t-test was the statistic used. A mean for life events was calculated for men(n=189) and women(n=210) and a t-test was run on these means. The result was a t value of -2.07 which is significant at the .05 level of statistical significance. The null hypothesis, H is therefore rejected. Women reported a larger number 01’ of potentially stressful events than men. The t-test is a negative number because the mean of group one(men) is smaller than group two(women). Mean two is subtracted from mean one in the calculating of a t-test: in this situation the result is a negative number which is maintained through the completion of the calculations. This is mentioned because in most of the t-tests which follow, the results are negative indicating the means for women were larger. H : Men and women will perceive that an equal 02 number of events that are internal to the family has taken place. 86 Women will perceive that more events internal to the family have taken place than men. Men and women will perceive that an equal number of school events has taken place. Men and women will perceive that an equal number of work events has taken place. Men and women will perceive that an equal number of events involving relatives and close friends has taken place. Women will perceive that more events involving relatives and close friends have taken place than men. - Men and women will perceive that an equal number of health events has taken place. Women will perceive that more health events have taken place than men. Men and women will perceive that an equal number of household finance events has taken place. Men and women will perceive that an equal number of law events has taken place. A t-test was the statistical tool used to test these hypotheses. The independent variable is the sex of the respondent and the dependent variables are the counts of the "yes" responses within each category by nature. Table 6 reports the t values for the seven hypotheses, the significance levels, and the number of cases for men and women. Hypothesis six involving family health events is rejected at the .028 significance level with a t value Of “2020. ‘ Women reported more health related events than men. Hypotheses two through five and seven and eight were not rejected. However, hypothesis two invglyingliwu- o (”/7 events internal to family approached the .05 level of significance with a t value of -1.89 indicating that / / 87 Table 6 T-tests of Differences Between Men and Women and the Number of Events Identified by the Nature of the Events Number //J “ ficanc:>ign1 Nature of Event* of Cases {T Value) Level Events Internal Men 207 to the Family (2) Women 222 -1.89 .059 School Events (3) Men 239 Women 251 .16 .872 Work Events (4) Men 227 Women 243 -1.00 .320 Relatives and Men 236 Close Friends (5) Women 247 -1.54 .123 Family Health Men 233 Events (6) Women 247 -2.20 .028 Household Finance Men 233 Events (7) Women 246 -.34 .733 Law Events (8) Men 241 Women 252 .81 .420 * Number after each events refers to the number of each hypotheses. women reported more internal family events than men. 7*. i In additional analysis, a chi-square was run on the ,. sex of the respondent and the response of "yes" and "no" to each event in the life events scale. The finding of the largest magnitude was for the event, "household chores pile-up". The chi—square was 14.61 which has a significance level of .0001. Women reported this event more than men. "Relative dies" was the only other event that was statisti- cally significant with a chi-square value of 4.38: however, 88 other values approached significant levels. These events were "periodic absence of member", "member has major conflict with boss/other at work", "death of brother or sister", "member experiences menOpause", and "husband's or wife's parents or siblings require financial assistance". Appendix F contains the chi-squares for each event and the level of significance. RESEARCH QUESTION TWO Are there sex differences in the degree to which family life events are stressful?‘ The following hypotheses were generated for this research question. H09: There is no difference in the perceived degree of stress resulting from family life events H9: Women will perceive a higher degree of stress resulting from family life events than men. A mean stress score, based on the degree of disturb- ance reported for all experienced life events, was cal- culated for men(n=232) and women(n=254). The higher the mean score, the higher the level of stress. A t-test was then run on this score for men and women. The result was a t value of -1.67 with a signifcance level of .097. Women’repgrted a higher degree of stress than '1 men. The null hypothesis, however, is retained: the -1.67 t value was not significant at the .05 level. 89 The following seven hypotheses involve various categories of life events, based on the nature of the events. H010: Life events that are internal to the family will be perceived as equally stressful for men and women. H10: Life events that are internal to the family will be perceived as more stressful for women than for men. H011: There is no sex difference in the perceived degree of stress resulting from school events. H012: There is no sex difference in the perceived degree of stress resulting from work events. H013: Life events involving relatives and close friends will be equally stressful for men and women. H13: Life events involving relatives and close friends will be more stressful for women than for men. H01“: There is no sex difference in the perceived degree of stress resulting from life events involving health. H14: Women will perceive a higher degree of stress resulting from life events involving health than men. H : There is no sex difference in the perceived degree of stress resulting from life events involving household finances. 015 H016: Life events involving the law will be equally stressful for men and women. Mean stress scores for each category were calculated for men and women by adding the reported degree of dis- turbance for all of the events within each category, and dividing by the total number of events identified as taking place. A t-test was used to test the hypotheses. Table 90 7 reports the results of these tests including the ,» nature of the events, the t values, and the levels of.*\¥ significance. Hypothesis ten which involved events internal to the family had a t value of -1.65, significance level of .10, and approaches significance: however, the null hypothesis was retained. Hypotheses eleven, school events, and twelve, work events, are also retained with t values of .37 and -.99 respectively. The t value for hypotheses thirteen, involving relatives and close friends, is -2.00 at a .047 level of significance. This hypothesis was rejected: women reported greater stress than men for events that included relatives and close friends. Hypothesis fourteen, health events, and hypothesis fifteen, household finance events, were both retained with t values of -.41 and -.23 respectively. Hypothesis sixteen, law events, was rejected at the .019 level of significance with a t value of -2.43. Women reported greater stress than men for events related to law. T-tests were also run on the degree to which each individual event was stressful. Appendix F includes the results for these events. Seven events had significant t values at the .05 level or less. These were: 1)death of a member(t=-2.10, p=.038), 2)family pet dies(t=-3.37, p=.001), 3)member demands new privileges, exemption from family rules, choice of friends, dates, etc.(t=-2.53, p= .01), 4)household chores pile-up(t=-2.02, p=.045), on Degree 91 Table 7 Num (1” Hf .... T-test of Differences Between en and Women f Stress by Each Natuquof the Eve ts ber Level of Nature of the Event of Cases T Value Significance Events Internal Men 156 -1.65 .10 to the Family (10) Women 175 School Events (11) Men 46 .37 .71 Women 40 Work Events (12) Men 148 -.99 .32 Women 156 Events Involving Relatives and Men 143 -2.00 .047 Close Friends (13) Women 158 Health Events (14) Men 106 -.41 .68 Women 134 Household Men 141 -.23 .816 Finance Events(15) Women 152 Law Events (16) Men 25 -2.43 .019 Women 31 92 5)death of close friend or confidant(t=-2.93, p=.OO4), 6)member experiences serious emotional problems(t=-2.04, p=.045), and 7)family member involved with courts, robbed or assaulted, arrested for crime or minor misdemeanor. jailed or involved in lawsuit(t=-2.43, p=.02). For all of these events, women reported more stress than men. RESEARCH QUESTION THREE Will the sex of the respondent influence the impact which day—to-day, routine relationships and aspects of life have? The following hypotheses were generated for this research question. H017: There is no sex difference in the impact that children make daily on the respondent. H17: The daily impact that health makes on the respondent will be greater for women than for men. H018' There is no sex difference in the impact that parents make daily on the respondent. H019. There is no sex difference in the impact that the spouse makes daily on the respondent. H020. There is no sex difference in the impact that the ex-spouse makes on the respondent. H021: There is no sex difference in the impact that in-laws make on the respondent. H022: There is no sex difference in the impact that brothers/sisters make on the respondent. H023: There is no sex difference in the impact that friends make on the respondent. H024 There is no sex difference in the impact that neighbors make on the respondent. 93 H : There is no sex difference in the impact that 025 work makes on the respondent. H026‘ There is no sex difference in the impact that leisure makes on the respondent. H027: There is no sex difference in the impact that transportation makes on the respondent. H028‘ There is no sex difference in the impact that health makes on the respondent. 28: The daily impact that health makes on the respondent will be greater for women than for men. H029. There is no sex difference in the impact that meals make on the respondent. H030: There is no sex difference in the impact that household chores make on the respondent. H031: There is no sex difference in the impact that finances make on the respondent. H032: There is no sex difference in the impact that pets make on the respondents. H : There is no sex difference in the impact that 033 errands make on the respondent. There is no sex difference in the impact that time use makes on the respondent. The t-test was used to test these hypotheses. Table 8 reports; the results of the t-tests. Day-to-day, routine relationships and aspects of life which have t values with significance levels of .05 or less are: 1)in-1aws(t=2.20, p=.03), 2)work(t=-2.15, p=.03), 3)meals(t=3.63, p=.OOO). Men reported more poSitive impact related to inelaws and meals while women reported more positive impact from work. Although these are the only relationships and aspects of life which are significant at the .05 level, others Table 8 94 T-tests of Differences Between Men and Women in the Impact of Day-to—day, Routine Relationships and Aspects of Life Relationship, Number Level of Aspect of Life of Cases T Value Significance Children(17) Men 233 Women 251 -1.32 .19 Parents(18) Men 202 Women 214 -.50 .62 Spouse(l9) Men 238 Women 252 1.45 .15 Ex-spouse(20) Men 32 Women 31 1.70 .09 In—laws(21) Men 216 Women 221 2.20 .03 Brothers/sisters(22) Men 220 Women 238 -1.40 .16 Friends(23) Men 218 Women 237 1.73 .10 Neighbors(24) Men 236 Women 252 -1.11 .27 Work(25) Men 224 Women 206 -2.15 .03 Leisure(26) Men 235 Women 249 -.70 .48 Transportation(27) Men 233 Women 246 -1.31 .19 Health(28) Men 237 Women 249 —.54 .59 Meals(29) Men 237 Women 251 3.63 .000 Household chores(30) Men 237 Women 251 1.80 .07 Finances(31) Men 237 Women 251 .54 .59 Pets(32) Men 210 Women 203 -1.47 .14 Errands(33) Men 231 Women 241 1.48 .14 Times Use(34) Men 241 Women 232 .36 .72 95 approach this significance level: ex-spouse, friends, and household chores. Men report a more positive impact from all of these relationships and aspect of life. RESEARCH QUESTION FOUR Will the employment status of the respondent influence the perception and the degree to which an event is stress- ful and will this be related to the sex of the respondent? The following hypothesis was generated for this research question. H035: Employment status and the sex of the respondent are not related to the total perceived degree of stress. Two-way ANOVA was the statistical tool selected to test hypothesis thirty-five. There are three F values of the Twé-way ANOVA. The two main effect variables, sex of the respondent and employment status, have P values of 3.59 with a level of significance equal to .059 and 2.75 with a significance level of .012 respectively. The two- way interaction has an F value of .385 with a significance level of .819. Employment status does influence perceived degree of stress significantly and the influence of sex of the respondent approaches significance. Hypothesis thirty—five was retained because of the interaction P value. The researcher was not satisfied with the amount of information produced by this statistic. Although it does provide a statistical basis for acceptance or 96 rejection of hypothesis thirty-five it does not examine the differences in the perceived degree of stress for men and women of the same occupational status. Therefore, the researcher recoded the variables, sex of the respond- ent and each employment status, into one variable. This resulted in a nominal independent variable and an interval level dependent variable, the total degree of stress. Therefore, for each employment status, a men's and a women's group was formed and a t-test run for these groups. The results of these t-tests are reported in Table 9. Table 9 T—tests of Differences Between Employment Status of Men and Women and Total Degree of Perceived Stress Number Level Employment Status of Cases T Values Significance Full-time Men 160 -1.52 .13 Women 77 Part-time Men 3 .99 .369 Women 42 Unemployed Men 19 -.58 .575 Women 6 Retired Men 19 -1.15 .264 Women 11 Disabled Men 9 -.4O .70 Women 5 97 For some of the t values the number of cases is very small. T-tests are designed to compensate for small n values. No men reported being full-time homemakers and no women and only one man reported being in school so t-tests were not run for these. No significant t values were found. Additional t-test were run individually for men and women to compare perceived stress in different employment statuses. For example, full-time employed men and part- time employed men were compared. Again, not all categories had a sufficent number of cases and t-tests were not run on these. The results for men and women are reported in Table 10. The only significant difference between the groups was for unemployed and retired men with a t value of 2.44 and a significance level of .02. Unemployed men were more stressed than retired men. Although not the focus of the study, the employment status for men and women combined was examined since the two-way ANOVA found that employment status was signifi— cantly related to the total degree of perceived stress. The means for each employment status in order of most stressed to least stressed were: disabled 3.26, unemployed 3.16, part-time 3.13, full-time 2.89, home- maker 2.81, and retired 2.63. The differences in these means accounts for the significant outcome on the two-way ANOVA. 98 Table 10 T-tests of Differences Between Employment Status and the Total Degree of Perceived Stress of Men and Women Number Level of Employment Status of Cases T Value Significance Men Full-time vs 160 part-time 3 -1.49 .232 Full-time vs 160 Retired 19 1.83 .082 Unemployed vs 19 Retired 19 2.44 .02 Full-time vs 160 Unemployed 19 -1.42 .168 Women Full-time vs 77 part—time 42 .75 .457 Full-time vs 77 retired 11 .27 .789 Part—time vs 42 homemaker 83 .35 .725 Full—time vs 77 homemaker 83 1.22 .225 Part-time vs 42 retired 11 -.O2 .986 99 RESEARCH QUESTION FIVE Are the social supports that are selected and utilized during times of stress different for the sexes? The following hypothesis was generated for this research question. 6‘ Men and women utilize social support equally H03 in times of stress. H36: Women will utilize more social support in times of stress than men. T-tests were used to test this hypothesis where the independent variable is the sex of the respondent and the total utilization of social support score is the dependent variable. This score was a mean of all the items on the social support scale except for, "facing problems 'head on' and trying to get a solution right away" and "watching television". These items did not appear to the researcher to reflect social support usage. The higher the mean social support utilization score the higher the frequency of social support usage. The results of the t-test revealed a t value of -4.53 at a significance level of .OOO(a significance level of .000 results when the level is smaller than .1 percent). The number of women in this computation was 211 and the number of men was 209. Hypothesis thirty-six is then rejected. Women utilized social support more extensively than men. ”'Each item on the social support utilization scale was also tested for significance using t-tests. The 100 results are reported in Table 11. All of the t values are negative indicating that women had larger means then men. Seven of the seventeen items are significant at the .05 level. The significant items were: 1)sharing difficulties with relatives(t=-4.67, p=.000), 2)seeking information and advice from persons in other families who have faced the same or similar problems(t=—4.16, p= .000), 3)seeking advice from relatives(t=-3.19, p=.000), 4)attending church services(t=-2.43, p=.015), 5)sharing concerns with close friends(t=-5.50, p=.000), 6)doing things with relatives(t=-1.98, p=.048), and 7)asking relatives how they feel about the problems we face(t=-3.87, p=.OOO). Women utilized social support more frequently than men in all instances. A two-way ANOVA was used to test the differences be- tween employment status and sex of the respondent on the mean social support utilization score. The main effect of sex of the respondent had an F value of 6.961 with a significance level of .009. The main effect of employ- ment status had an F value of 1.155 with a significance level of .330. The two-way interaction F value was .540 with a significance level of .706. Because no significance was found between the interaction of the independent variables, no further testing as described for hypothesis thirty-five was conducted. The researcher was interested in knowing the level of social support usage when the total degree of stress was high. The researcher selected those respondents who had 101 Table 11 T-tests of Differences Between Men and Women in Social Support Utilization Number Level of Item of Cases T Value Significance Sharing difficulties Men 234 with relatives Women 252 -4.76 .000 Seeking information and advice from persons in other families who have faced the same or Men 234 similar problems Women 252 -4.16 .000 Seeking advice from Men 232 a relative Women 253 -3.91 .000 Asking Neighbors for Men 232 assistance and favors Women 254 -1.88 .06 Seeking assistance from community agencies and programs designed to help families in Men 232 our situation Women 254 -1.58 .114 Accepting gifts and favors from Men 235 neighbors Women 251 -1.69 .092 Seeking information and advice from Men 232 family doctor Women 251 -.41 .682 Facing problems "head on" and trying to get solutions Men 224 right away Women 238 -1.82 .068 Watching Men 227 television Women 242 -.O6 .949 Attending church Men 230 services Women 246 -2.43 .015 Sharing concerns with Men 231 close friends Women 249 —5.50 .000 102 Table 11 (cont'd.) Doing things with Men 232 relatives Women 250 —1.98 .048 Seeking professional counseling and help for family Men 231 difficulties Women 249 -.77 .443 Participating in Men 232 church activities Women 232 -1.56 .119 Asking relatives how they feel about Men 232 problems we face Women 245 -3.87 .000 Seeking advice from Men 231 a minister Women 244 -1.16 .248 Sharing problems Men 234 with neighbors Women 249 -.77 .444 a mean degree of stress score greater than three(moderately disturbing) and then ran a t-test between men and women and the mean social support usage score. The t value for this test was -4.50 with a significance level of .000. The number of cases was 139 for men and 152 for women. A second t-test was run on a group who had a mean degree of stress score of four or more(quite to extremely disturbing). The t value was -2.65 with a significance level of .009. The number of men was sixty three and the number of women was fifty. For both degrees of stress, women used social support more than men. CHAPTER V DISCUSSION The discussion that follows will focus on the research questions and the findings from Chapter 4. A section of this chapter is devoted to the possible influence that sex role socialization has on the findings. Also included is a discussion of the relationships between the Boss model of family stress and the Human Ecosystems model. DISCUSSION OF RESEARCH QUESTION ONE Research question one involved male and female differences in the perception of the number and nature of family life events experienced. In the total count of events women reported significantly more events than men. This was a predicted difference and an important finding in this study. When the events were considered by nature of the category, men and women are similar in their reported number of events. In all the categories men and women showed no significant difference except in the Family Health Events category. Women reported more Family Health Events than men. Women may identify more health events because they have been informed of these from other family members. Members may be less likely to share these events with men. Women may con- sider more events as health problems than men. For example, women identify menopause as a health event more than men at 103 104 a level approaching significance. Women may identify and define menopause as a health event because it is something that happens to them. Men may not define it as a health event because they do not experience it, are not aware of it or do not perceive it as a problem. The signs of all the t-test for the categories, except for law events and school events, are negative indicating that women had higher means for number of reported events. The predicted difference between men and women reporting events internal to the family was not found. However, the results did approach significance. The pre- dicted difference between men and women reporting events in- volving relatives and close friends was not found. Although the actual t-tests are not significant, in all categories but one, the analysis does show a trend toward women identi- fying more events as occurring within each category. For'most of the single events, the number of the men and women reporting that particular event was similar. The events, "household chores pile-up" and "relative dies" had significantly more women reporting their occurrence. Women may report that the event, "household chores pile-up" more frequently because they are more often responsible for these chores. They are aware of what chores have gone unattended and may experience pressure from spouse, children, or others to complete these tasks. The women may experience guilt and embarrassment as a result. A possible explanation of women's greater report that "relative dies" may be related to the past findings that women are more involved with the family and kin(Lowenthal, et a1, 1975: Chiriboga 105 and Dean, 1978: Phillips, 1981). Women may identify more individuals of the extended family as relatives than men. They may also be in contact with more kin members and there- fore be notified of deaths. The actual number of persons identifying each event as taking place in the last three years is approximately twenty to forty-five men and women for each event. Four events stand out as having many more persons identifying them. The event "household chores pile-up", was reported by 117 women and 70 men. "Member accepts time consuming, unpaid assignment in voluntary association" was reported by 73 men and 75 women. The event, "relative dies" was reported by 84 men and 112 women. And finally, 72 men and 72 women identified "family forced to dip heavily into family savings". These four events were experienced most frequently by both men and women. DISSCUSSION OF RESEARCH QUESTION TWO Research question two focuses on the perceived degree of stress experienced by men and women. When an overall stress score was calculated from the perceived degree of stress identified by men and women for all of the items, no significant difference between men and women was found. Although women reported more events, they perceived a similar amount of stress as ment However, the results gained from testing the hypothesis regarding the total degree of stress approached significance. The t value was -1.67, significant at the .097 level, and for many researchers is within the acceptable range for rejecting 106 the null hypothesis. Subsequent research carried out to validate this study could find a greater difference. Past research has found significant differences for men and women in the degree of stress experienced(Morse and Furst, 1982: Bradley, 1980: Martinson and Anderson, 1979: Weinman, et a1, 1978). In examining the perceived degree of stress by nature of the event, two findings were significant. Events in- volving Relatives and Close Friends and Law Events were identified as most stressful for women than for men. It was expected that these events involving family relationships would be more stressful for women as Lowenthal(et a1, 1975) and Chiriboga and Dean(l978) found since women are involved with family and friendships to a greater degree. However, Chiriboga and Dean also found men to be stressed more over legal issues and the opposite of this was found in the present study: that is, women were stressed more with Legal Events. A possible explanation for this difference could be the fear and anxiety experienced by women if they have been assualted or robbed. Another possible explanation would be that women experience shame, guilt, and embarrassment if a member was jailed or involved in the courts particularly if these involved their husbands or their children. The category, Events Internal to the Family, approached significance for the women which was expected. All but one of the categories had negative t values which indicates a trend toward women being more stressed. The category which involved School Events was the only category for which men reported more stress. 107 Individually, seven events were significantly perceived as more stressful for women than for men. These were 1)death of a member, 2)fami1y pet dies, 3)member demanding new privileges,exemptions from family rules, choice of friends, dates, etc., 4)household chores pile-up, 5)death of a close friend or confidant, 6)member experiences severe emotional problems, and 7)family member involved in courts, robbed or assaulted, arrested for crime or minor misdemeanor, jailed or involved in law-suit. All of these were perceived as more stressful for women than for men. Of the forty- eight events on the scale, only eight had positive t values. This adds support to the findings that women tend to per- ceive more stress than men. "Member demanding new privileges, exemption from family rules, choice of friends, dates, etc.," may be more stressful for women if children are making these demands because of closer ties between mother and children. The children may be showing their need to be independent which can be difficult for mothers. Women may deal with these demands more often than men and are therefore more stressed by these. Because friendships have been found to be more significant for women, death of a close friend or confidant may be particularly difficult. Certain events had means that were relatively high. A response of three on the questionnaire identified moderate stress: four identified the event as being quite stressful and five determined the event to be extremely stressful. Table 12 reports the means for the events that were greater 108 Table 12 Family Life Events Perceived as Moderately to Extremely Stressful Event Death of a member Marital separation occurs Pregnancy of unmarried member Member drops out of school before com- pleting training Major wage earner loses or quits job Member has major conflict with boss/others at work Relatives/in-laws become intrusive Death of husband's or wife's parents Married children "freeze out" parents Major wage earner ex- periences serious illness or accident Aged parent becomes seriously ill or disabled requiring direct care Death of brother/sister Death of close friend or confidant Member experiences serious emotional problems Child member experiences serious illness or accident Family takes major loss in stock market, bank failure, bad debts, etc Family members involved with courtS. robbed or assualted, arrested jailed, lawsuits Member experienced demotion Men 3.82 3.90 3-55 3.46 3.56 3.40 3.38 3.52 3.67 4.06 3-73 3-55 3-29 3.88 4.06 4.09 3.48 3.28 Women 4.23 4.00 3-55 3-37 3.95 3.50 3.92 3-79 3.54 437 4.19 3.71 109 than three and are therefore events that were more than moderately stressful. In most cases the women had higher means than the men. The eighteen items are those which were perceived to cause moderate to extreme stress. All other events had means lower than three. Three of the listed events are internal to the family, two are work events, one is a school event, five involve relatives and close friends, two involve household finance, one involves the law, and four involve health. The means for each category, in order from most stress- ful to least stressful are: Law Events Men 3.48 Women 4.19 Health Events Men 3.56 Women 3.61 Relatives and Close Friends Men 2.93 Women 3.19 Events Internal to the Family Men 2.71 Women 2.89 Household Finance Men 2.72 Women 2.75 School Events Men 2.59 Women 2.50 Work Events Men 2.32 Women 2.45 Total Stress Means Men 2.80 Women 2.93 None of the categories had means under two which indicates that at least some stress is being felt. Although this is an age in which work related stress is receiving much attention, this research found this area to be less dis- turbing than several other categories. When compared to family and health events, work events may have a lesser value and hence generate less stress. The family may be seen as more important and be given a higher priority hence demanding more emotional effort. Health events, because of their serious nature could be expected to be 110 at the beginning of the list. However, Law events were almost equally stressful and this is somewhat unexpected. The Law events involved a family member coming into con- flict with the law, a member being robbed or assualted, or a member involved in court. All of these can be events which are very intense for a relatively short time period. They also can cause shame and guilt for the family. Be- cause of this the Law events may be reported as being more stressful. Events involving relatives, friends, and members of the immediate family fell in the middle while Financial and School events were towards the end of the scale. The Household Finance category produced less stress than most other categories. This was somewhat unexpected as the State of Michigan at the time of this study was in economic straits due to problems in the auto industry as well as difficulties within farming. Even with the problematic economic con- ditions, the respondents still identified events involving immediate family members, relatives, friends, and health higher in the production of stress indicating perhaps that the sample has a higher degree of impact and influence than finances on the lives of the respondents. DISCUSSION OF RESEARCH QUESTION THREE Research question three deals with day-to—day, routine relationships and aspects of life. The differences be- tween men and women in the impact made by these relationships and aspects of life were investigated. The scale to measure the impact had five levels. A score of one 111 represented a "very negative" impact, two, "moderately nega- tive", three, "little effect", four, "good" and five, "very good". Three of the eighteen events showed significant dif- ferences. Men ranked meals as having a more positive impact on their lives. The higher impact is likely a result of men enjoying the consumption of the meal or the occasional pre- paration while women may find the preparation dull, demanding, and unpleasant. The "in-laws" relationship was also more postive for men than for women. Women may believe their abil- ities as a wife are continually evaluated by the in-laws possibly leading to stress, guilt, and diminished self- esteem. Women may deal more with the in-laws than men re- sulting in opportunities for negative encounters to take place. Men may not experience this same type of relation- ship. "Work" was more positive for women than for men. Work for women may be something they have freely choosen and also something which they can freely leave. They may experience increased self-esteem from work hence making the impact more positive. Men may feel trapped in their work because of their responsibility to support the family and therefore, be less positively impacted. The impacts of health and children were predicted to be different for men and women but differences were not found. All but one of the relationships and aspects of life,"ex-spouse", had means greater than three. This indicates that the aspects of life and the relationships that occur on a daily basis were re- ported to have "little effect" to a "very good" effect on the lives of the participants. Since the means are similar for men and women on all but three of the aspects and 112 relationships of daily life, the general conclusion is that, overall, these aspects and relationships have "little effect" or a "good" effect on the lives of men and women. Three aspects and relationships had means greater than four. The "children" relationship had a mean of 4.09 for men and 4.21 for women: women reported a slightly more positive impact of chilren. This is perhaps due to the larger involvement of the mother with the children. The "leisure" aspect resulted in means of 3.97 for men and 4.03 for women, with women reporting slightly more positive impact. None of these were significantly different for men and women. The "ex-spouse" relationship was the only relationship or aspect of life which had a mean lower than three: these were 2.73 for men and 2.16 for women which shows a slightly more negative impact on the women. This particular relationship could be anticipated as having the most negative impact. It is interesting, however, that the means are close to three signifying "little effect" from the relationship instead of a negative impact. This was the only scale which did not indicate women as being more negatively affected or involved. Men and women appear to be equally affected either neutrally or positively, by these aspects and relationships. The signs on this scale show nine aspects and relationships as positive(the means were higher for the men) and nine as negative(the means were higher for the women). However, on this scale means for men and women were very similar. 113 The differences between the sexes are minimal: day-to-day, routine relationships and aspects of life appear to impact upon men and women in a similar fashion. However, weaknesses in the scale are recognized. The items in the scale are very different from one another making a con- sistent appraisal across items difficult at the same time giving no sense of wholeness or unity to the scale. The respondent is only given one word from which to identify an impact. Individual interpretations of these could be diverse causing the reliability and validity of the scale to be questioned. The difficulties with the scale should be considered when examining the results. DISCUSSION OF RESEARCH QUESTION FOUR Research question four involves the total degree of stress identified by men and women and the employment status of each. The twowway ANOVA resulted in a signifi- cant finding for employment status as it impacts upon the total degree of perceived stress. The sex of the respond- ent did not prove to be a statistically significant in- fluence on the total degree of stress perceived, although the probability level approached significance(p=.059). This concurs with the results found in research question two. The question of concern here is the interaction effect of both employment status and sex of the respondent as they relate to the perceived degree of stress. The interaction of these two variables did not have a significant relationship 114 with the total degree of perceived stress. Men and women in various employment statuses were equal in the degree of stress that they perceived. Therefore, it appears that each status has its own unique set of potential stressors: no one status is void of these and both men and women experience stress within each status similarly. When men and women within each employment status were compared, no significant differences were found. The total amount of perceived stress was equal for men and women in each of the employment categories, full-time, part- time, unemployed, retired, and disabled. The means for men and women in these categories centered around the middle of the scale, three, which indicated a moderate degree of stress. The lowest means for one particular employment status were for retired men and women(men 2.38, women 2.87) and the highest means were for disabled men and women(men 3.19, women 3.35). It should be remembered that the two-way ANOVA resulted in a finding that employment status did relate to total degree of stress perceived. Until this point employment statuses have been separated by groups of men and women. When men and women were combined and only employment status considered, the following means for degree of stress resulted, in order of highest to lowest: disabled 3.26,unemployed 3.16, part-time 3.13, full-time 2.89, homemaker 2.81, and retired 2.63. Retired persons are least stressed while disabled persons are most stressed. Retired persons may be less stressed because they no longer have the pressures 115 of a job and time table. They are free to choose activities and may spend more time in leisure hence reducing potential stressors. It is likely that some degree of stress remains a part of daily experience but they spend more time relaxing which minimizes the stress. Disabled persons may be the most stressed because of what they cannot do. They may feel inadequate to do the tasks society has identified as the responsibility of middle aged adults, whether it is providing financial support, caring for household tasks, or caring for self. Guilt, shame, and embarrassment can result from the inability to carry out normal and expected adult functions and tasks. The full-time employed and homemakers are on the bottom half of the list which indicates that these employment statuses are slightly less stressful. The individuals in these statuses are perhaps satisfied with their positions to a greater degree. Although stress is inevitable, the factor of choosing to be in that status or desiring that status,may minimize the stress. Housewives may feel less societal pressure because of their status and full-time employed men and women may be contributing to a necessary degree to family finances hence providing security and reducing stress. When men were compared across different employment statuses, no two groups were significantly different ex- cept unemployed and retired men, although differences be- tween full-time and retired men approached significance. The result here was that retired men were slightly stressed 116 with a mean of 2.38 while unemployed men were moderately stressed with a mean of 3.05. The unemployed status can cause stress due to financial reasons. Much pressure can be on an individual who needs employment, but cannot find it, to support self and others. Unemployed as a status identifies persons who want to work but do not have the opportunity: they have not choosen to be unemployed. Feelings of self-worth may suffer in this status. For women, all employment statuses were equally stress producing. The means for these statuses were clustered around the middle of the scale, three, which identifies women of all employment statuses as moderately stressed. As discussed in Chapter 4, not all employment statuses could be involved in the calculations due to the small numbers of individuals in some statuses. DISCUSSION OF RESEARCH QUESTION FIVE Research question five involves the utilization of social support in times of stress. A total social support utilization score was calculated. Women were significantly higher in the degree of social support used. That is to say, women when stressed used more social supports than men. Women contacted other persons more than men for aid, assistance, information, and advice when stressed. This supports findings in previous research(Ilfeld, 1980: Paykel, et al, 1980: Roy, 1978: Brown, 1975). 117 When each item of the scale was examined individually, all had negative t values indicating that women used these specific social supports more than men. Seven individual social supports used significantly more frequently by women were: 1)sharing difficulties with relatives, 2)seeking information and advice from persons in other families who have faced the same or similar problems, 3)seeking advice from relatives, 4)attending church services, 5)sharing con- cerns with close friends, 6)doing things with relatives, and 7)asking relatives how they feel about problems we face. Four of the seven significant social supports involved relatives indicating the importance of relatives in pro- viding social support for women. The mean score for each of the social supports was between 1.5 and 3.0. This indicates that the social supports are used between slightly less than once or twice a year to monthly. No single social supports were used to a great extent. This may indicate that the respondents rely on their own resources and not other persons. One particular social support is used more than weekly. "Watching television" as a support had means of 4.35 for men and 4.36 for women. This parti- cular support actually involves minimal interpersonal interaction and was not included in the total social support score. However, television was used to the greatest degree. This possibly resulted because television is used by indi- viduals as a source of relaxation. It is socially 118 acceptable for both men and women and is readily available. This, however, cannot be considered a social support. When an individual watches television they may actually limit their involvement with others. The results of the two-way ANOVA involving the employ- ment status of men and women and their utilization of social supports concurs with what was reported above. Sex of the respondent is significantly related to social support utilization. Employment status of the individual was not related to the social support utilized. The interaction between employment status and sex of the respondent also proved to be insignificant. No further testing was done in relation to employment status and social support as was done for research question four because the two-way ANOVA proved there to be no signifi- cant relationship between the variables. Women, regardless of employment status, utilized more social support: no one employment status influenced social support utilization for men or women. When the respondents who perceived themselves as being moderately to extremely stressed were examined, it was found that women used more social supports to a highly significant degree(t=-4.50, p=.000). When res- pondents who were quite to extremely stressed were examined, it was also found that women used more social support when stress was high(t=-2.65, p=.009). However, men 119 and women who are quite to extremely stressed were more similar in utilization of social support than those who were moderately to extremely stressed. This indicated that when stress mounts to higher levels some men will utilize social supports and men and women may move toward equal utilization. When examining the means of men and women in the moderately to extremely stressed category(n=139, men: n=152, women) and the quite to extremely stressed category (n=63, men: n=50, women), some differences can be seen. For the moderately to extremely stressed category, the means were 24.90 for men and 30.63 for women. In the quite to extremely stressed category, the means were 17.75 for men and 23.80 for women. The means of the first category indicate a mean score by which social support is utilized once or twice a year while the means of the second category indicate social support is virtually never used. This indicates that when higher levels of stress are experienced, less social support is utilized. This might mean that when stress rises to higher levels, persons may tend to withdraw to some degree. On the other hand, it might also indicate that persons who use social supports to a limited extent, perceive greater stress because they do not use social support as a buffer. The findings relating to research question five are the most significant of this study and indicate a strong relationship between the sex of the respondent and social support utilization. Women reported higher utilization 120 on every item on the scale with many being at a signifi- cant level. It can be stated with confidence that women utilize social support to a greater extent than men. SEX ROLE SOCIALIZATION The results of this study show that women report more stressful events than men, women perceive more stress from particular events and categories of events, and that women use more social support in times of stress. The general trend is for women to perceive more stress in their lives than men and to actively use social supports when exper- iencing stress. A major question which arises around the issue of actual stress experienced and amount of stress reported on the instrument is, do individuals, and particularly men, report the actual amount of stress that they perceive in their lives or do they withohold at least part of this information? Are the male and female differences in the measured stress due to men and women actually perceiving differences or do they perceive stress similarly but do not report it? These questions cannot be answered through the use of data on the questionnaire. No information from the participants is available on whether there are actually differences in stress perceived or if the sex role sociali- zation of men and women has taught men to be strong and silent and women to freely express anxieties. The researcher believes, however, that sex role socialization 121 is a contributing factor to these differences. In middle age or older adulthood, men and women have been found to move away from traditional sex roles towards androgynous roles(Fisher and Narus, 1981). Adults appear to trans- cend sex roles and take on cross-sexed characteristics. Romer(181) concurs when stating that there is a shift toward egalitarian roles in middle age and Brim(l976) states that men and women move towards a unisex position in the middle years. However, this is not a complete transcendence and some basic roles still remain even if they have somewhat diminished over time. This study found women to report more stressful events in their lives than men. This is in keeping with research on sex role socialization. Balswick and Peek(l977) identi- fied men as being inexpressive either as the strong, silent, he-man individual or as the cool, resourceful, shrewd man. These results from personality factors which are deve10ped and encouraged and from cultural expectations. Hatzenbuehler and Joe(l981) studied men and women between the ages of seventeen and forty-four. They found that men reported lesser amounts of stress. The researchers deter- mined that it was not due to actual lower levels of stress but that the men were less likely to disclose and admit the stress. Richmond—Abbott(l983) concurs in her study where it was found that women were more likely to admit fear and anxiety. The results of this study show women to report more stressor events involving household chores and realtives. According to Albrecht, Bahr and Chadwick(l979) 122 housekeeping :roles, child care and concern roles, and kinship roles remain constant for women over time. The higher degree of stress experienced by women can also be linked to sex role socialization. Numerous studies exist which identify sex role differences in this area. Archer and Lloyd(l982) identified women as having more fear and anxiety. This is supported by other researchers in the field(Block, 1976: MacCoby, 1966: Terman and Tyler, 1954: Tyler, 1965). Men appear socialized to be less Open and invulnerable(David and Brannon, 1976) and unemotional and rational(Rosenber, 1973: D'Andrade, 1966) while women are socialized to be emotionally free (Rosenberg, 1973). These researchers believe that women and men are socialized into these roles. These then could have a significant impact on how men and women respond on the questionnaire. It is the researcher's belief that these sex roles could have influenced the results of perceived stress in this study. Women clearly reported the use of social supports more than men. This could also be influenced by sex role socialization. Kimmel(l980) states that men are not en- couraged to form and do not form close relationships with other men. Friendships for men are not intimate and do not involve self-disclosure(Lewis, 1978). Women are more socially involved and oriented and have affliative interests (Block, 1976: MacCoby, 1966: Terman and Tyler, 1954, Tyler, 1965). The results of the present study which find women more involved-in social support usage then could 123 be influenced by sex role socialization. As previously stated, there is no way to determine from the questionnaire what the true influencing factors are for the male and female differences in perceived stress and social support utilization. Sex role sociali- zation does provide at least one possible explanation of the differences. Women appear to be encouraged to be emotional and self-disclose and hence find it easier to report anxiety and stress. Men are generally not encouraged to do so and do not self-disclose to as great a degree. Women are encouraged to self-disclose with friends and family while men are not. There are a number of results within this study which show men and women are similar in perceived stress. This could be a result of the shift toward more cross-sexed characteristics in middle age. Differences between men and women are not seen in every event because of the androgynous role change for persons in the middle years. Although sex role socialization could influence some of the results from this study, other results in which no differences were found could be influenced by this shift toward egalitarian and similar roles. 124 THE THEORETICAL FIT BETWEEN THE BOSS MODEL OF FAMILY STRESS AND THE HUMAN ECOSYSTEMS APPROACH The Boss model was selected for the theoretical basis of this study because of the inclusion of a contextual approach which more completely explains family stress. The model, in many ways, is similar to the ecosystems model both in its concepts and its assumptions. In the eco- systems approach, the human environed unit(an individual or a family) is not viewed in isolation. It interacts with the environments which surround it, the human behavioral environment, the human constructed enviroment, and the natural environment. The human environed unit is impacted upon by these and in turn impacts upon the environments. To adequately understand the human environed unit, all of the influences of the environment must be examined. Figure 6 diagrams a general human ecosystems model. The natural environment encompasses the unit and the other environments and includes those things that man cannot create such as air, water, earth, time, climatic zone, and natural plant and animal life. However, these things can be altered by man. The human constructed environment involves those aspects that have been altered or developed by man such as language, norms, culture, religion, cities, buildings, clothing, and other artifacts. Finally, the human behavioral environment involves the presence and behavior of other people. All of these interact with each other and with the human environed unit. Feedback is 125 Climatic Zone Neighborhood Community Food, Enemy Clothing, Shelter 7 _ . I 3: Ir a (“u IF. fif‘fi. ffirlastu-a. .w "Ia'mIIy, IIIends Colleagues 1322-. Z," Iichoolrnatefi‘u-U .‘-'-I if?" ,‘f‘p h Riff 1“" ‘ III Other . Natural Resources Figure 6 The Human Ecosystem Model from Bubolz, Eicher, and Sontag, "The Human Ecosystem: A Model", Journal of Home Economics, Spring, 1979. 126 important to this interaction process such that input into the system results in some type of output which then goes back into the system. The Boss model approaches family stress in a similar manner. The stress process is the unit of concern. This unit cannot be seen in isolation: the context of the process impacts upon it. There are three contexts: 1)the external context involving cultural, religious, constitutional, developmental, economic, and historical aspects, 2)the internal context, embedded in the external context, includes sociological, psychological and philosophical aspects of the family, and 3)the response context involves the cognitive, emotional, and behavioral responses of the family. These contexts are to be examined in order that a more complete understanding of the stress experience can be gained. Feedback takes place within the stress process such that the degree of stress feeds back into the A(the stressor event), B(coping resources), and C(perception of the event) factors(see figure 2). There is not a perfect fit between the two models. However, the concepts in the Boss model can be viewed as more specific and at a different level of abstraction. The ideas in both models are similar. Both models emphasize the influence of factors beyond the unit or process and the feedback that takes place. Inclusion of the environments and the contexts indicate that the process is embedded within something larger than the 127 individual or other unit under study. The overall depth of the two models allows for the stress process to be understood in its complexity. The Boss model of Family Stress is ecological in nature and provides a relatively complete framework for understanding family stress. The sample and results of this study relate to the Boss model in a number of ways. It must first be noted that individuals, not families, are the focus of this study. The Boss model is a model primarily used in the study of family stress but is employed here for the understanding of individual stress resulting from family life events. It is therefore important to realize that the family is a context for the stress experience of men and women in this study. The sample was comprised of married couples and the life events involved family members, therefore, this study must be seem it the light of the family context. A study of stress emphasizing another context such as the work place, may yield different results. Although the individual is the unit of analysis, the family is the context in which the individual is immersed, and identifies the need to utilize the Boss contextual model for understanding the stress experience. The results of this study identify the need to con- sider the stress experience contextually. It was found that women report more health events as taking place than men. Health is an aspect of the external context, the constitutional context, which involves the physical and 128 biological health and strength of the individual. Health issues are not only specific incidents but also act as a context for the stress experienced, whatever its source. The piling up of household chores is related to the internal, sociological context. Women may report this event more often due to their functions as wife and mother within the structure of the family. Women perceive more stress resulting from events involving relatives and close friends. This may also be related to the sociological context but also could be related to other findings in this research that identified relatives as particularly important to women. Women turn to relatives for social support more than men and use relatives as a coping resource(B factor). Therefore, events involving relatives may be perceived as more stressful because of closer contact between women and relatives. Employment status is part of the external economic context. It was found that employment status does relate to perceived stress. Social support is used more by women than by men. This soical support is part of the coping resources employed(B factor) and influences the A and C factors as well as the degree of stress experienced. The findings identify how components of the Boss model work to influence the stress process. This study is not broad enough to include all aspects of the model but does begin to use the approach to study a research question. CHAPTER VI SUMMARY, CONCLUSIONS, AND IMPLICATIONS SUMMARY The purpose of the study was to examine male and female differences in perception of stress related to family life events and in the utilization of social support and in the influence of employment status on perceived degree of stress. The research questions generated for this study were: 1)Do men and women differ in perception of the number and nature of family life events eXperienced?, 2)Are there sex differences in the degree to which family life events are stressful?, 3)Will sex of the respondent influence the impact which day-to-day, routine relationships and aspects of life make upon the individual?, 4)Will the employment status of the respondent influence the percep- tion of the total degree of stress experienced and will this be related to the sex of the participant?, 5)Are the social supports that are selected and utilized during times of stress different for the sexes? The study involved Michigan residents in four rural counties and One Standard Metropolitan Statistical Area. The respondents included 232 couples in the middle years of life with the women between the ages of thirty-five and sixty—five. Forty-two additional individuals were also included. 129 130 Sample survey was the methodology employed. The research instrument was a mailed questionnaire developed by the research team of the NC-164 project entitled, "Stress, Coping, and Adaptation in the Middle Years of the Family" which is supported by the Agricultural Experiment Stations of the North Central Region of the United States and the Cooperative Research Service of the United States Department of Agriculture. T-test and two-way ANOVA were the primary statistics used for analysis. CONCLUSIONS Analysis of the data from the NC-l64 project related to family life events, day-to-day, routine relationships and aspects of life, and social supports suggest several conclusions applicable to the participants of this study. The conclusions are a result of the findings in Chapter 4 and the discussion of the findings in Chapter 5. These contribute to a more complete understanding of the relationships between sex of the respondent and the per- ceived degree of stress in the lives of the individuals as well as the social supports utilized. Based on the research findings, the following con- clusions have been reached. 1. Women perceive that more family life events have taken place than men. 2. Women perceive more health events as taking place than men. In all other categories, no male and female differences exist. 10. 131 The piling up of household chores is perceived as more of a stressor for women than for men. Also the event "relative dies" is more often reported by women than by men as a stressor. No other single event is identified as happening more often for one sex ox'the other. In total life stress across all events, men and women perceive an equal degree of stress in their lives. Neither sex perceives significantly more stress than the other in overall life stress. However, there is a trend toward women perceiving more stress from the life events. Women perceive more stress than men in events involving relatives and Close friends and law events. Men and men perceive stress similarly in other categories of events. Day-to-day, routine relationships and aspects of life impact upon men and women in equal intensity. These appear to affect men and women very little or in a positive manner. Only three relationships and aspects of life, in-laws, work, and meals are experienced differently for men and women. Men are impacted more positively by in-laws and meals and women are more positively impacted by work aspects. The employment status and the sex of the respondent do not interact to influence the total degree of perceived stress. Employment status alone does relate to the perceived degree of stress experienced for both men and women. Men and women in each employment status show no difference in the total degree of stress perceived in their lives. The same employment statuses are perceived equally stressful for men and women. Women of various employment statuses perceive stress equally. Retired men perceive less stress than unemployed men. All other combinations of employment statuses for men result in equal amounts of perceived stress. Women utilize social support much more frequently than men in times of stress. Women use the following social supports more than men: 1)sharing difficulties with relatives, 2)seeking information and advice from persons in other families who have faced the same or similar problems, 3)seekin advice from relatives, 4)attending church services, 5 sharing concerns with close friends, 132 6)doing things with relatives, and 7)asking relatives how they feel about the problems we face. Relatives are important in providing social support for women. 11. The employment status and sex of the respondent do not interact and influence social support utilization. 12. Middle-age, like all other periods in the life cycle, has potentially stressful events: many events are unique to the time period. Middle-aged men and women show both similarities and differences in the amount of stress perceived and in the events that are potentially stressful. These are the conclusions drawn from the findings in this study. In general, women label more family life events as occurring, are more stressed by these events, and utilize more social support: employment status impacts little on the total degree of stress perceived by men and women. Referring back to figure 6, the design for analysis, there is a significant relationship between the independent var- iable, sex of respondent, and the dependent variable, number and nature of family life events. There is not a significant relationship between sex of respondent and the dependent variables of perceived degree of stress in life events and routine relationships and aspects of life. However, a definite trend does exist for women to perceive more stress. The dependent variable, utilization of social support is signi- ficantly related to the sex of respondent. The intervening variable of employment status does influence perceived stress but does not interact with sex of respondent to influence perceived stress. of 133 LIMITATIONS The following are the acknowledged research limitations this study. 1. The mail questionnaire as a research instrument generally has a low response rate. For this study, the rate of return was at the one third level. This is relatively low when compared to the participation of respondents in other types of studies. 2. The data collected by the questionnaire are a self- report. It is impossible to determine if the responses given are the true attitudes of the respondents or if the respondents answered the questionnaire withegtq thewinfluence.or-aid~of”anotherwperson. 3. Each respondent's interpretation of "degree of dis- ‘turbance" is unique. The interpretation of the scale by the individuals is not uniform. This scale, as well as the others, is subject to individual inter- retation. 4. The life events scale does not include all possible family life events that could take place in the middle years. This scale should not be considered as a complete list. 5. The day-to-day, routine relationships and aspcets of life scale is highly subject to personal inter- pretation. The respondent isrsimply given one word with Which to respond with no further clarification. This should be considered when examining the results from this scale. 6. In order to truly understand the impact of the life events, it would need to be determined if an individual adapted or maladapted to the events. This is beyond the scope of this particular research project but is necessary for a more complete understanding of the stress experience. 7. It is impossible to determine if the stress is due to the undesirability of the event or the change that results from it. 8. Confounding variables enter this research that could potentially alter the results but which are not con- trolled: it is impossible to control for these con- ;founding variables. 13L: 9. This study does not identify situations which results in positive stress. The word stress has a negative connotation. The questionnaire asks "how disturbing was the event?" which also has a negative connotation. 10. Time will affect the ABC-X model, according to McCubbin(et a1, 1983), by altering available resources, the perception of the event, and the number or pile-up of potentially stressful events. The Boss model of Family Stress does not adequately include time: time is not a variable in this study. 11. Although the Boss model includes many contextual components for understanding the stress experience, this particular study could only deal with a limited number of variables thus complete under- standing from the Boss model perspective is not possible. 12. This study cannot be generalized to the population at large. It can be generalized to other similar geographic areas in the State of Michigan. Various rural and urban settings were chosen from whence to draw the participants. Similar types of urban and rural living environments are found throughout the State of Michigan maintaining many of the same qualities and characteristics. It is the researcher's belief that the results should not be limited to the designated communities alone but could be applied to other adult residents throughout the state. IMPLICATIONS The implications from this study fall into two categories, implications for future research and implications for professional work in the field of family studies. The first implication involves the Boss model of Family Stress. Although this study had as its goal to test this theoreti- cal approach, it was only able to work with a piece of the entire model. Much more work utilizing this model needs to be done in order to further test it. This model is a new one with a great potential but requires more extensive testing. Samples need to be drawn from a variety of 135 contextual areas such as the work place or from populations experiencing chronic health problems. This would provide information about the stress process within a variety of contexts. The theoretical testing of the stress process could also be expanded to include the Human Ecosystems model. This would include more variables than the Boss model and could prove to generate significant information re- garding the stress experience. This would broaden the understanding of stress to include more influencing factors. This study is one of the first to examine male and female differences in the perception of stress. Continued research needs to be carried out to validate the findings reported here. The middle-aged population and sample should be tested again as well as samples from different age populations. The research here is a first step: much more research needs to follow to more accurately and com- pletely understand sex differences in perceived stress. Also, the research on male and female differences in the utilization of social support is very weak. Much more testing could take place on various samples of various ages. The social support scale utilized here was relatively small and could be expanded to test more extensively the utilization of social support. The scales utilized in this study are relatively new. Instruments for testing all the areas of this study should be further developed and/or new instruments created. The 136 stress experience is a complex one and the instrumentation must be developed to accurately research the experience. Educational programs of various natures could utilize the results of this study. Courses which deal with sizess and/orthe middle years of life could incorporate this information. The information would be particularly useful to the student whoeis entering young adulthood and wants {3fo trad} to better understand his or her parents or to students A4,?” yx’W g 4%“, who are preparing for professions that will deal with persons of this age group or with stress in families. This information could be well incorporated into adult education programs. Organizations such as churches, business, and C00perative Extension Service, who educate and work with the public in self-awareness could utilize this information. Persons in the middle years of life would be especially interested in the findings. The specific findings of this study would need particular attention in educational programs. Health events were reported more by women. Women of various ages could benefit from instruction dealing with prevention, preparation, and perception of health events. Men also ‘ede'i could benefit from this information as their identification of existing heilth events may be inhibited. This same approach couIdEe Igéewtg deal with family events as a whole. Women report more events than men and educational programs can be developed which aid in awareness of events and realistic evaluation of their potential to create stress. For example, pile up of household chores is more of a concern for women. Educational programs can aid men in an awareness of this situation and can help wgmen, as ----s *Mwm‘ well as men, to learnfihowvtomcope with household tasks more ermi’mm -- -- - "I realistically. Law events were more st‘essful for women. Programs may want to provide women withIOpportunities to become familiar with the courts and standard law procedures to prepare f°?.IhiS type of event,/"‘ \X -- 1-1.1. ‘_ The total degree of perceived stress is similar for men and women with a trend toward women perceiving more stress in family events. Educational programs could focus on the similar experience of men and women in order that neither sex is excluded in instruction. Men and women experience a similar impact of day-to-day, routine aspects and relationships of life and would likely benefit from similar instruction. Men should not be excluded because of past notions that daily family life impacts more upon women. It was also found that sex was not a factor in interrelationship with employment status as it influences perceived stress. Employment status alone did relate to perceived stress. Therefore, men and women would both benefit from education directed at a particular employment status. For example, disabled men and women could both gain from a learning experience involving stress and the disabled. Emphasis should not be placed on the male and female differences but on differences between employment status such as retired persons vs the unemployed. The 138 needs would vary between the statuses. Social support utilization is very different for men and women. Women need to be made aware that they do rely more on others for support. This may ultimately influence decisions made by women. Men may_penefit from an awareness of the benefits of social support utilization and also would likely benefit from aid provided that would teach men how to develop and utilize social support. Practicing counselors may find this information useful when dealing with patients or clients. Since internalized stress is harmful to health, reduction and elimination of some stress could be helpful. Social support acts as are buffer in times of stress and helping persons develop social supports could positively affect the health of the individual. Men may be particularly in need of help in' light of previous research findings that men internalize stress more, and the findings of this study that men do not use social support as a buffer to a great extent. Men may need additional help dealing with stress. Social service agencies could also put this information into practical use in tailoring their services to meet the needs of their clients. For example, women report more family life events as taking place, more health events as taking place, and more stress from law events and events involving relatives and close friends. This information allows the clinician to prepare for situations which may arise in- volving these stressors and also provides cues to look for 139 when dealing with women. The clinician can develop group experiences based on this information and the needs of the clients. Clinicians must also be aware of similarities between men and women and aid both in dealing with the stress experienced in life, whether it be from family events, day-to—day, routine aspects and relationships of life or employment status. Groups formed to assist persons in dealing with stress would likely be more beneficial if both men and women were involved in the same group. Mutual awareness would likely be rewarding. Men and women do utilize social support in different ways. Therefore, services and programs can be helpful in promoting health through the use of social support in times of stress and also in aiding men and women in dealing with stress. Relatives are a frequently used source of social support for women. Women who are geographically mobile may lose this support and need help in dealing with stress. {Programs deve10ped t0’ meet the social support needs of women\would be highly beneficial. Businesses, churches, and schools may want to recognize these needs of women and develop programs to aid them particularly if they are geographically mobile. Reduction and management of stress are important to the physical well-being of individuals. The results of this\\ study could be used by educators and clinicians to help [I men and women reduce and manage stress in the middle yeare of life and in the development of social supports. APPENDICES APPENDIX A THE RESEARCH INSTRUMENT 1422 Demographic Information 1 He would like to have some background information about your family to help us in our study. Please fill in the following information about each member of your household, identifying each person by their relation- ship to you. First, think about yourself. Sex. M F Mo. & Yr. of Birth Yrs. of School Completed (circle one) Marital Status If Married, Mo. 5 Yr.of Marriage Next, think about each of your children, starting with the oldest child. He will be asking questions about each of your children later in this questionnaire. Please make sure your answers are from oldest to youngest in each instance. Birth Yrs. of Living at (If NO) Date 1 of Date School Home Reason for Left Support Q2 Sex Mo. Yr. Completed Yes or No Leaving Mo. Yr. You Provide a. Child 1 M b. Child 2 M c. Child 3 M d. Child 4 H M 'fl‘fl'fl'fl“ e. Child 5 f. Child 6 M F (add on if necessary) I:Finally, think about each other member of your household.g] Yrs. of S of Relationship Sex Birth Date School Marital Support 3 to You H or F Mo. Yr. Completed Status You Provide Religious Preference: 04 Catholic Protestant Jewish Other (please specify) 1A3 05 Which of the following best describes your racial or ethnic identification? Black Hhite (Caucasian) Chicano (Mexican-American) Oriental Native American (American Indian) Other (please specify) Q6 How many years have you lived in your present comunity? Q7a. Hhat is the size of the community in which you live? (circle one) 1. Less than 2,500 persons and outside an urbanized area 2. More than 2,500 persons but less than 50,000 persons 3. 50,000 or more persons b. (If you live in a comunity of less than 2,500 persons) Is your home on less than 1 acre of land or on a city or suburban lot? 1. Yes (skip to 08) 2. No c. (If no) Do you live on a farm? 1. Yes 2. Mo (skip to 08) d. (If you live on a farm of more than 1 acre) Did your farm produce $1,000.00 or more in sales of crops, livestock, or other farm products during the preceding year? 1. Yes 2. No 11m Family Life Events Questionnaire Please read each of the events listed below and mark whether it was ex- perienced by any family member in the last three Jears. If yes, please circle the number showing how disturbing it was and indicate whether it occurred in the last twelve months. How Disturbing was This Event? Has This Event 5 >- qa FAMILY LIFE EVENTS Happened to " .- Q Did It Your Family In E g =3 13‘: Occur In A. Internal to the The Last Three 5 .— c: S o- The Last Family Years? 2 a.” 2. o 6 12 Months? a. Death of a member Yes No 1 2 3 4 5 Yes No b. Marriage of a member Yes No 1 2 3 4 5 Yes No 1M5 How Disturbing Has This Event? Mas This Event :3 > Happened To ‘5 {‘3 :3 Did It Your Family In E”: 3‘: w E. Occur In The Last Three .— 2". ‘5’ t: {‘5 The Last Years? 2 a 2 8 z 12 Months? . Member moves out of home (for independence, for added schooling, for job, for marriage) Yes No 1 2 3 4 5 Yes No . Member moves back ' (unemployed, divorced, or separated, etc.) Yes No 1 2 3 4 5 Yes No Non-member (renters, boarders, etc.) moved into home Yes No 1 2 3 4 5 Yes No Marital separation occurs Yes No 1 2 3 4 5 Yes No Periodic absence of family member due to work demands Yes No 1 2 3 4 5 Yes No . Family pet dies Yes No 1 2 3 4 5 Yes No . Pregnancy of unmarried member Yes No 1 2 3 4 5 Yes No . Member demanding of new privileges, exemptions from family rules, choice of friends, dates, etc. Yes No 1 2 3 4 5 Yes No . Adult child has trouble achieving independence Yes No l 2 3 4 5 Yes No . Household chores pile up Yes No 1 2 3 4 5 Yes No . Family took a stressful vacation Yes No 1 2 3 4 5 Yes No This Question Continues On The Next Page B. Family, School and Hork . Member drops out of school before com- pleting training . Member returns to school after time away . Major wage earner loses or quits Job Major wage earner starts or returns to work Member given promotion . Member changes to new job or shifts career, . Major wage earner retires from work Member accepts time consuming, unpaid assignment in volun- tary association (scouting, church, or service agency) Outside activities draw adult members away from family Member's hours/ scheduling of work change Has This Event Happened To Your Family In The Last Three Years? Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes 146 No No No No No No No No No No How Disturbing Has This Event? MODERATELY QUITE EXTREMELY SLIGHTLY NOT Did It Occur In The Last 12 Months? Yes Yes Yes Yes Yes Yes Yes YES Yes Yes No No No No No No No No No No bb. CC. ff. . Member has major conflict with boss and/or others at work C. Family, Relatives and Close Friends Relatives/in-laws become intrusive (offer un- welcome advice, gifts) . Death of husband's or wife's parents . Death of brother or sister Death of close friend and confidant Married children “freeze out“ parents . Member breaks up with close friend or confidant Relative dies (not parent or sibling) D. Family and health Major wage earner experiences serious illness or accident Member experiences serious emotional problems 147 Has This Event Happened To Your Family In The Last Three Years? Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No No No No No No No No No No How Disturbing Has This Event? NOT 1 SLIGHTLY 2 23 >- E 5 Did It ('5 g 5 Occur In 8 '5 E The Last 3: c :5 12 Months? 3 4 5 Yes No 3 . 4 5 Yes No 3 4 5 Yes No 3 4 5 Yes No 3 4 5 Yes No 3 4 5 Yes No 3 4 5 Yes No 3 4 5 Yes No 3 4 5 Yes No 3 4 5 Yes No This Question Continues On The Next Page 1148 How Disturbing Has This Event? Mas This Event 5 >- Mappened To 5 i3 6 Did It Your Family In E 5 g 33 Occur In The Last Three .5 E ‘5’ g .‘E The Last Years? 2 6 E a 33 12 Months? hh. Child member experi- ences serious illness/accident Yes No I 2 3 4 5 Yes No 11. Aged parent(s) becomes seriously ill or disabled requiring direct care Yes No 1 2 3 4 5 Yes No 11. Member experiences menopause Yes No 1 2 3 4 5 Yes No kk. Aged parent committed to institution or placed in nursing home Yes No 1 2 3 4 5 Yes No E. Family, Household Finance and the Law 11. Husband's or wife's parents or siblings require financial assistance Yes No I 2 3 4 5 Yes No um. Cut in total family income Yes No 1 2 3 4 5 Yes No on. Expenses exceed total family income requiring going into debt Yes No I 2 3 4 5 Yes No 00. Family takes a major loss in stock market, bank failure, bad debts, etc. Yes No 1 2 3 4 5 Yes No pp. Family receives windfall funds (inheritance, lottery win, or other unanticipated gain) Yes No I 2 3 4 5 Yes No 1249 How Disturbing Has This Event? qq. tr. It. ll“. Mas This Event :3 >- Mappened To E if i'u" Did It Your Family In 3: § us 5 Occur In The Last Three t— ‘3 g g E‘- The Last Years? 52’ tn" 2 a 5 12 Months? Member starts receiving public assistance in the form of food stamps, rent subsidy or AFDC Yes No 1 2 3 4 5 Yes No Member takes out or refinances a loan to cover increased expenses Yes No I 2 3 4 5 Yes No Family member involved with courts; robbed or assaulted, arrested for crime or minor mis- demeanor, Jailed, or involved in lawsuit Yes No I 2 3 4 5 Yes No Family forced to dip heavily into family savings Yes No 1 2 3 4 5 Yes No Member taking on additional Jobs Yes No I 2 3 4 5 Yes No . Member experiencing demotion, job bumping, or retooling Yes No l 2 3 4 5 Yes No F. Other Events Not Covered Yes No 1 2 3 4 5 Yes No Yes No 1 2 3 4 5 Yes No 1f50 The following relationships and aspects of day to day routine are stress- ful for some persons. Please circle the number which best represents the impact each of these has on your life most of the time. A rating of 1 would mean a very negative effect on your life. a very good effect. A rating of 5 would mean Children Parents . Spouse Ex-spouse . In-laws Brothers/Sisters . Friends Neighbors Hork . Leisure . Transportation . Health . Meals Household Chores Finances . Pets . Errands . Time Use . Other (specify) VERY NEGATIVE l l pppwwwwflppflflfifl MODERATELY NEGATIVE NNNNNNNNNNNNNNNNNN LITTLE EFFECT 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 GOOD b“...bb.b..&..¢bbb VERY GOOD mmmmmmmmmmmmmmmmmm 151 To get an accurate financial picture of the families in our study, we need to know something about your family's work and income. First, we would like some more information about your work: QZl..Are you working for pay, either full-time or part-time? Yes, employed full-time (35+ hours/week) or with a job but not at work at present because of temporary illness, vacation, or strike. Yes, employed part-time (less than 35 hours/week) _____ynemployed, laid off, looking for work _Ful 1 -time homemaker _Reti red _____jn school _Di sabl ed Other (Please specify ) qzz Please give us some information about the type of work you do/did. (a) Hhat is/was your main occupation or Job title? (b) Hhat king of work do/did you do; that is, what are/were your main duties on the Job? (c) In what type of business or industry is/was this; that is, what product is/was made or what servicevis/was given? Q23. Next, think about ur ersonal income before taxes for l982. Be sure to include all sources of income that you receive personally; such as earned income, investments, social security, your own business, Job-related benefits, welfare benefits, rent and so on. If you farm or have your own business, we would like you to indicate your net fanm or net business income before taxes. 3 (nearest $1,000) Q24 Now, think about your total family income for 1982. This is total income before taxes for all members of'your family, including yourself and your children. Be sure to include all sources of income; such as earned income, investments, social security, your own business, joborelated benefits, wel— fare benefits, and so on. If your family farms or has its own business, indicate net farm or net business income before taxes. 8 (nearest $1,000) 1552 Q39a. Every county or parish in the state has a Cooperative Extension Service which provides information and programs, including 4-H, Home Economics or Family Living Programs, Expanded Nutrition Programs, Agriculture, Horticul ture, Natural Resources and comunity development programs. Have you used any program or assistance from the Cooperative Extension Service to help deal with a problem, change or stress? yes no - b. If yes, what was the change, problem, or stress? If used, how helpful was the assistance? (If not used, leave blank) :: 2 a; 0).” 0 2325...? mucus 3:28-48 zmz