I111 I111 13,129 E 101 11 1111 I1 II1 LIBRARY Michigan State University A Hm This is to certify that the thesis entitled MARRIED WORKING MOTHERS' EVALUATIONS OF AND RESPONSES TO SHARING AND NON-SHARING COMMERCIAL PORTRAYALS 0F MULTIPLE-ROLE WOMEN: AN EMPIRICAL STUDY presented by Sherri J. Crispin has been accepted towards fulfillment of the requirements for M.A. degeein Advertising Major professor Datew 0-7639 MSUi: an Affirmati... A .- "1 ' A" ‘ J Institution MSU LIBRARIES .__—_ RETURNING MATERIALS: ace 1n 00 remove this checkout from your record. be charged if BEEF-is rop to FINES will returned after the date stamped below. 51;; 3 r;- ’85-! 5S >. I_ m ‘.,. 3 M‘kv-J ("w 2 1m ”7&5 'Jwyfim \ If, H". rm I“ 1'3 “3-5-3 ”3; 361 W .- ,.‘. 3 ‘99‘2': 1&9 :7 I389 555398 1131323012 MARRIED WORKING MOTHERS’ EVALUATIONS OF AND RESPONSES TO SHARING AND NON-SHARING COMMERCIAL PORTRAYALS 0F MULTIPLE-ROLE WOMEN: AN EMPIRICAL STUDY By Sherri J. Crispin A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Advertising 1986 .___—___—_._L I he? ,‘ ,f, . 9/". 3, X— ' second type, "sharing, ABSTRACT MARRIED WORKING MOTHERS’ EVALUATIONS OF AND RESPONSES TO SHARING AND NON-SHARING COMMERCIAL PORTRAYALS 0F MULTIPLE—ROLE WOMEN: AN EMPIRICAL STUDY By Sherri J. Crispin This is a study of married working mothers’ evaluations of and responses to commercials that feature married working mothers. Two types of appeals are examined. The first type, "non—sharing," features a working mother character who performs multiple roles ( wife, mother, worker) without assistance from her spouse and without apparent stress. The " features a working mother character who performs multiple roles but shares a child care and/or km— '''''' domestic task with her husband. A quota sample of seventy—seven working married mothers was divided into two experimental groups distinguished by portrayal type. Analyses revealed subjects judged the "sharing" and "non-sharing" portrayals as significantly different along two attribute dimensions, task sharing and role conflict. Further analyses did not indicate that one portrayal type is more realistic or irritating than the other; portrayal appropriateness was a stronger predictor of Sherri J. Crispin effectiveness than group membership. Lastly, these data did not indicate that either portrayal type affects liking for or likeliness to purchase the advertised product. However after viewing "non-sharing" portrayals, subjects were less likely to believe that working mothers experience role conflict. This thesis is dedicated to ELEANOR JEANNE CRISPIN and LAWRENCE C. CRISPIN ACKNOWLEDGEMENTS I would first like to acknowledge several organizations which contributed to this project. I would like to acknowledge the seven companies whose commercials were the primary foci of this study.. I would like to acknowledge Michigan State University’s Office of the Assistant Provost, Office of Personnel and Employee Relations, the Clerical/Technical Union and the Supervisory Association for their assistance in the sampling process. I also thank the study’s participants, as well as the many volunteers who expressed interest in this study, for taking the time to make the study happen. I would also like to thank the Department of Advertising for supporting this endeavor. Dr. Bruce VandenBergh and Dr. Martin Block offered guidance throughout the study’s planning stage, and Dr. VandenBergh and Dr. Bonnie Reece devoted time in editing the final manuscript. I would also like to acknowledge several personal resources who helped me retain, or at least regain, my sanity throughout this process. These are people who assured me, in some way, that I could finish what I had started, that I could strive to do my best and know that it is good. I would like to thank James Dinkelacker for taking time to share his knowledge of research and his philosophies on life and learning. Jamie has challenged me to think, to observe, and to listen; to look below surfaces and to look carefully; to always be aware that what you see may not be what you get; and to never think that I have found "the answer." I would also like to thank Kim Downing for caring to offer advice and encouragement over the past two years, and I thank MIRA for the opportunity to view research from the inside and to see it done well. I would like to thank Shannon Brannam for her unique mixture of energy, ambition, and laughter, and for always finding the space in her life to react to mine. Just remember, Shannon, in ten years, Meryl and Marge will be history, but we’ll have just begun! I would also like to thank Susan R. Crispin for being my friend, for providing me with a sense of home away from home, and for sharing with me so many of the things she has learned along the way. I would like to thank Kerri Sue Wobma, a long-distance but unrivaled long-term support system. I feel very lucky to have a friend who I have known for so long and who shares much of my history. Thank you for the craziness of our "formative years" and for accepting my evolutions. iv I would like to name Jane Lovett, another long—distance sustenant, for sticking with me through the radical stage. Thank you for making me laugh—~a lot!-—and for showing me that differences can yield sameness. I would also like to thank Hollie Hollerbach for understanding and sharing my passion for writing and for the dramatic. Either one will probably get us into trouble . . . but it’s worth it! I would like to thank Lynda M. Waldecker for inspiring me to re-examine and reclaim the need for creativity in my life. Thank you for supporting my individuality and for never ceasing to amaze me with yours; for teaching me to appreciate gentleness; and for getting me to slow down. You showed me that the stars and the lake will still be there. I would also like to thank Pisi for affection and companionship. And lastly, I would like to thank Ferron for helping me make the Return and for showing me clarity of vision. If in my life I can do as well, I will be happy. TABLE OF CONTENTS Chapter I. INTRODUCTION . . Background . . . . . . . . . . . . Women and work The changing market of working women Advertising appeals to the working mother market . . . . Research Questions Summary . . . . II. LITERATURE REVIEW Introduction . . . Sociological Literature . . . . . . . . Advertising Literature . . . . . . . . Summary III. METHODOLOGY.. . . Introduction Hypotheses Commercial attribute hypotheses Commercial effectiveness hypotheses Sales effect hypotheses Women’s beliefs hypotheses . . . . . . . Questionnaire Development . Pre— test Commercial evaluations Post— test Experiment Design . . . . . . . . . . Experiment Preparation . . . . . . . . . . Stimuli Selection University Permissions Sampling Sample Selection . . Sampling Procedure . . . Instrumentation Summary vi Page TABLE OF CONTENTS (Continued) IV. ANALYSIS AND RESULTS . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . Sample Profile . . . . . . . . . . . . . . Preliminary Analyses . . . . . . . . . . . Data Transformations . C C S W S ommercial Attribute Hypotheses . . . . Scale modifications Testing the commercial attribute hypotheses . . . . . ommercial Effectiveness Hypotheses . . . Scale modifications Testing the commercial effectiveness hypotheses . . Examining the interaction effects . . . Treating the interaction . . . . . . ales Effect Hypotheses . . . . . . Product liking hypotheses Likeliness to purchase hypotheses omen’ s Beliefs Hypotheses . Scale modifications Testing the women’ s beliefs hypotheses ummary . . . . . . . . V. CONCLUSIONS APPENDIX APPENDIX APPENDIX APPENDIX ENDNOTES A: w O A Introduction . . . . . . . . . . . . . . Limitations Conclusions . . . . Recommendations Future research Recommendations for advertisers Summary Questionnaire Sample Letter One Sample Letter Two Sample Profile Breakdown ND BIBLIOGRAPHY 122 122 122 125 133 133 135 137 139 164 167 170 172 Table Number 3. 3. l 2 .10 .11 LIST OF TABLES Title Task Sharing Items . Role Conflict Items . . . . . 3 . . . . Perceived Similiarity Items Effectiveness Items Irritation Items . . . . . . . . . . Job Status Belief Items . . . . Task Sharing Belief Items Role Conflict Belief Items . . . . . . Role Expectation Items Experiment Design . . . . . . . . . Breakdown of Participant Volunteers by Mailings and Eligibility Attribute Factor Loading Comparisons T—tests of Task Sharing Attribute Item Task Sharing Attribute Means Table by Group and Commercial . T—tests of Role Conflict Attribute Item Role Conflict Attribute Means Table by Group and Commercial T—test of Perceived Similiarity Attribute Item . . . . Perceived Similiarity Attribute Means Table by Group and Commercial Effectiveness Factor Loading Comparisons . . . viii Page 40 42 44 47 48 52 54 56 57 62 68 78 84 84 85 85 86 86 88 4.9 4.10 4.11 4.12 4.23 LIST OF TABLES (Continued) T-test of Effectiveness Item T—test of Irritation Item . . . . . . Realistic Item Means by Group and Commercial Irritation Item Means Appropriateness Item Means by Group and Commercial Correlation Matrix for Realism, Irritation with Appropriateness Regression and Analysis of Covariance Results with Dependent Variable "Realism" . . . . . . . . Regression and Analysis of Covariance Results with Dependent Variable "Irritation" . T—tests of "Sharing" Group’s Product Liking Scores T-tests of "Non—Sharing" Group’s Product Liking Scores T-tests of "Sharing" Group’s Likelihood to Purchase Scores T—tests of "Non—Sharing" Group’s Likelihood to Purchase Scores Pre—test/Post-test Women’s Beliefs Factor Loading Comparisons Tatests of Job Status Belief Item T—tests of Task Sharing Belief Item T-tests of Role Conflict Belief Item T-tests of Role Expectation Item ix 91 91 92 92 94 95 97 99 102 103 105 106 109' 113 114 115 116 CHAPTER I INTRODUCTION This is an experimental study measuring married working mothers’ evaluations of and responses to commercial portrayals that feature married working mothers. The focus of this study is to examine subjects’ ratings of two types of commercial portrayals: those that feature working mothers who participate in task sharing and those that feature working mothers who do not participate in task sharing. This introductory chapter contains three sections which will familiarize the reader with the study’s topic and purpose. Those sections include "Background," "Research Questions,‘ and "Summary." The first section, "Background," discusses two issues that are centrally relevant to this study: 1) the entrance of women——including married women with children—~into the labor force, and 2) the effects of this change on the market of women consumers. The second section in this chapter, "Research Questions," will present questions addressed in this study regarding advertisers’ appeals to the working mother market. Lastly, the "Summary" will review the contents of this chapter and will provide an overview of this manuscript’s subsequent chapters. W Women and Work Change is an everpresent aspect of living. As individuals we change jobs, we change lifestyles, we change our minds, and we age. But individual behavior does not always occur in isolation. Within a social system, individual actions combine to form another source of change: social change. Social change occurs when individuals seek a more efficient management of social and individual resources, perhaps in response to a currently inefficient status quo. From this perspective, social change begins when individuals’ needs and/or abilities change, and when individuals pursue changes that will afford them a more satisfactory existence. One current example of social change concerns women’s entrance to the labor force. Increasingly, women have begun to pursue economic independence, self-sufficiency, and self— fulfillment aside from traditional domestic roles. As a result, more women are seeking paid employment outside the home. This is a change which the chairman of the National Commission for Manpower Policy calls not only a demographic trend, but "the single most outstanding phenomenon of our century."1 Government data reflect the magnitude of this change. From 1950 to 1983, the number of women in the work force more than doubled, rising from 18 million workers in 1950 to 50 million workers in 1984.2 Overall, 29: of the United States work force in 1950 was women while in 1984, 43.3% of civilian workers were 16 to 64 year old women. The most striking increase in women workers was in the 25 to 34 age bracket. In 1950, only 34% of these women worked; by 1983, this percentage had increased to 68.9%.3 However, the population of working women is not a "mass" market. This market can be separated into distinct groups with different needs and priorities; one such group is married working women with families. Historically, married women relied on their working husbands for economic sustenance. Yet presently, more married women are entering the work force for both economic and personal reasons. Whatever those reasons, the increasing number of married working women is a fact. The number of married women working in 1983 was five times greater than in 1940. Most recent census data show that only 12% of American households are "traditionally—structured" with a working father and housewife mother. The same data state that 60% of all families are dual—career households with both the husband and wife working.“ One industry member supports these data with the comment that "the major influx of women into the work force comes from an unexpected source: the married women we had assumed were happily engaged in keeping house."5 Similiarly, the number of working women with children has increased. In the early 1970’s, government forcasters underestimated the current rise of women in the work force by assuming that women with children under six years old would not go to work. However, mothers have entered the work force to a greater degree than any other group.6 In contrast to the forcasters’ expectations, Department of Labor statistics show that in 1984, 60.5% of all women with children under 18 years old were in the labor force. This number had increased to 62.3% by 1985, for a total of 20 million mothers in the work force.7 Similiarly, 52.1% of all women with children under six years old were working for pay in 1984.8 The Changing Market of Working Women Labor force trends suggest that many married women with children are working outside the home to bring home a second paycheck and increase their family’s standard of living. Government data state that the median annual income for husband/wife, dual—income families with children is 28% greater when the mother worked than when she did not.9 One marketer commented on this difference by observing that the working mother "doesn’t feel that she has to economize as much as other working women. . .she can spend without holding a family conference first." 10 For advertisers, these data suggest that the market of married working mothers is an attractive and‘ potentially lucrative one. Many advertisers have already acknowledged the increased buying power the working mother market represents. Industries which are profitting from higher sales to working women include air travel services, car manufacturers, liquor manufacturers, luggage companies, and credit card firms. One writer acknowledges that "businesses everywhere are shifting their sales pitches to appeal to women who have moved out of the kitchen and into the office. Instead of scouring pads and perfume, firms are selling the modern women insurance policies, houses and time- saving goods of every worth as her role as bread-winner expands."11 Working women as a group doubtlessly represent a growing and dynamic market that has drawn advertisers’ attention. Women’s advances in the labor force and increasing economic power have necessitated that profit-oriented advertisers acknowledge them. Consequently, advertisers’ primary decision—making has not involved choosing whether or not to target this group. Instead, the greatest dilemma for advertisers has centered on hgw to appeal to working women to best gain their business. This dilemma has been especially pertinent for advertisers addressing the market of working mothers, since these women’s activities and commitments are diverse. Advertisers who have chosen to target the market of married working mothers want to tailor successful messages that will attract this group’s attention and that will, hopefully, persuade the viewers to respond positively to the message. But many of these advertisers are questioning what specific strategies will achieve these results most effectively. Will a traditional, family-oriented message that appeals to the woman’s role as wife and mother be effective? Or would a contemporary, job—oriented message appealing to the woman as a working individual be more useful? Or should the advertiser use a combination of both portrayal strategies? Advertising Appeals to the Working Mother Market Given the recent emergence of the working mother market, advertisers have only begun to understand which appeals can most successfully reach these women. However, this lack of certainty has not precluded advertisers from directing advertising messages to the working mother market. Enough advertisers have targeted working mothers that trends in the types of appeals to this market have become visible. Trends are particularly apparent in television advertising, since the medium allows for portraying a character in a variety of roles more easily than do the print media. Thus, the following section will discuss two main topics which have developed and which concern commercial portrayals of working mothers: 1) controversy over types of multiple-role portrayals and 2) portraying the "modern" woman . The Multiple—role Portrayal Controversy Predominantly, advertisers who have chosen to target working mothers with television commercials have used "slice of life" appeals that portray a character in a variety of roles: she is a wife, she is a mother, and she works outside the home. Some industry members believe that this portrayal strategy effectively selects working mothers as the primary audience by portraying the multiple-role aspect of a working mother’s life. However, other industry members and consumers have disagreed, stating that the multiple~role portrayal strategy is not effective and, from a consumer stand—point, is controversial. The major criticism against multiple—role portrayals of working mothers stems from what has been termed the "superwoman syndrome." Industry members assert that many multiple-role portrayals overlook the strain a working mother experiences when attempting to fulfill several demanding roles. Similiarly, consumer critics have described multiple—role portrayals as exaggerated and unrealistic and state that "working mothers are too often portrayed in ads as saints on earth with standards no one can live up to."12 One senior account executive, referring to multiple-role portrayals, affirms that "this is not the way to successfully portray women, because we ggnlt do it all. We are not super— amazons--we are just people."13 Sociologists support this criticism, stating the following: "The woman who combines a professional career with family obligations is under strain not only because she participates in two ‘activity systems’ where the allocation of time and resources is a problem, but also because of the conflict always present——her commitment to her career versus her responsibilities to her family."H However, one industry executive acknowledges that realism is elusive when working mother ads strive to empathize with the consumer: "In the process of creating 'realistic’ portrayals, we instead produce dreadful little dramas we mistakenly call 'slice-of—life’."15 In sum, the most common criticism of multiple—role portrayals of working mothers is that the ads show a working mother "doing it all," that the character manages her family responsibilities and her work/career responsibilities without any stress or conflict. Critics reason that this situation does not accurately portray working mothers’ experiences and, thus, state that the ads are unrealistic and exaggerated. However, other advertisers have utilized the multiple—role portrayal concept but have developed a strategy which, thusfar, has not been as controversial. Since "doing it all" portrayals have received the "superwoman" criticism, these advertisers have chosen to revise the strategy-—to not portray working mothers doing it all. As a result, such advertisers have produced advertisements that portray a working mother in multiple roles, but show her sharing child care or domestic tasks with her husband. These advertisers have incorporated a situation sociologists call "task sharingz" spouses in a dual—career family, a family in which the husband and wife work, who share domestic and child care responsibilities. By portraying the working wife assisted by her spouse in a domestic and/or childcaring task, these portrayals *employ a major social change in traditional role structures, a shift which is largely an effect of married mothers having jobs outside the home. While these portrayals have addressed the concept of task sharing, so have they attended to complaints that the original portrayals showed a "superwoman" who could perform her responsibilities without visible stress or strain. Some of these more recent portrayals show a working mother character who actually comments on the role conflict she and her husband experience. More specifically, however, the presence of task—sharing activity in these portrayals reflects society’s shifting role structure and acknowledges role conflict as an issue currently being addressed by dual- career couples. 10 The Potential for Stereotyping Despite the controversy over appeals to the married working mother, consumer and industry critics state that advertising’s portrayals of women, in general, have improved over recent years. Previously, critics protested that print and television advertising portrayed women in primarily domestic roles. In contrast, one writer recently stated, "It seems that advertisers are finally getting the message to scrap the old stereotypes of women. And it’s about time--the ad industry has innumerable real-life role models to portray when it wants to sell to real—life women."16 However, other individuals caution against stereotyping the "modern" woman: "while recognizing the importance of working women as a market, we do not want to invent a new set of working women stereotypes to match those of housewives."17 However, some industry observers have identified a portrayal pattern which they describe as comparable to the domestic stereotype. These individuals feel that appeals to working mothers, and working women as a whole, primarily portray the characters as female executives. These individuals state that the portrayals of working women misrepresent women’s actual status in the labor force, since 1984 data show that 77.5% of women workers do not hold professional or managerial positions.18 Similiarly, industry members propose that portrayals like these are not optimally effective since they may fail to reach the majority of women who are not professionals or managers. One marketer comments that a 11 "female executive in a tweed suit will turn off a lot of women who are not executives and cannot relate to that."19 Another industry member states that portrayals of working women and working mothers have "substituted one cliche for another. . . showing a female executive in a conference room wowing a board of directors seems too forced and contrived, and women sensed advertising’s clumsy attempts to stereotype them, to ‘buy’ their favor."20 Statements such as these indicate to advertisers which aspects of past portrayals have been controversial and, perhaps, detrimental to message effectiveness. However, the basic question——which message strategies reach the working mother audience most effectively——is still unclearly answered. Research Questions As stated in the previous section, advertisers have developed some common strategies for appealing to the target of married working mothers. Because these women participate in multiple roles-—wife, mother, employee——many advertisers have chosen to portray a working mother character participating in all roles. However, such portrayals have drawn criticism from both advertising industry members and consumers. In contrast, other advertisers have thusfar avoided such criticism by portraying working mothers participating in multiple roles, but also participating in task sharing with their spouses. Research examining the comparable effectiveness of these two portrayal strategies is infrequent in the literature. Thusfar, no published study has explored whether married working mothers view these two portrayals--those that feature task sharing and those that do not-—as different. Although advertisers are using a revised multiple—role portrayal strategy, the literature does not reveal if this difference is apparent to the audience. Likewise, a frequent criticism of "superwoman" portrayals has been that the character shows no stress or strain in accomplishing her multiple responsi- bilities. This study intends to examine whether the two portrayal types differ in the amount of role sharing and the amount of stress the working mother character experiences. The third question posed by this study examines whether working mothers judge one of the portrayal types as being more similiar to their own experiences with task sharing. This question addresses the accuracy of the portrayals, since critics stated that initial portrayals did not accurately represent the working mother’s multiple—role experiences. Moreover, published studies have not examined the effectiveness of portrayal types distinguished by the degree of task sharing. If the two portrayal types differ in terms of task sharing, this study questions whether one portrayal type is more effective than the other. The relevant aspects 13 of effectiveness include how memorable, I realistic, informative, appealing and convincing the portrayal is. Similiarly, this study questions whether one portrayal type is more irritating or upsetting to the audience. This study further questions the comparative effectiveness of the two portrayal types in terms of product attitudes and future behavior. More specifically, can viewing a group of portrayals that do not feature task- sharing affect the viewer’s attitude toward the product or the viewer’s intent to purchase the product? Can viewing a group of portrayals that do feature task—sharing achieve these effects? Lastly, this study presents questions in response to critics who state that portrayals of working mothers misrepresent the status of working women in the labor force. One question aSks, how might exposure to portrayals, that either do or do not feature task sharing, affect working mothers’ beliefs about working mothers in general? More specifically, can exposure to either portrayal type affect beliefs about working women’s job status? Can exposure to either portrayal type affect beliefs about the amount of task sharing other working mothers experience? Can exposure to either portrayal type affect beliefs about the amount of strain or role conflict other working mothers experience? Finally, this study questions whether exposure 14 to either portrayal type influences working mothers’ expectations about how much task sharing should be expected between working mothers and fathers. Summary This chapter has provided an introduction to advertising portrayals of working mothers. This section first provided a discussion of labor trends which show increasing numbers of women entering the labor force, a group which includes married women and mothers. This trend has affected advertisers’ treatment of the female audience, since working women-—particularly working mothers in dual-income families-— represent a market with attractively increasing buying power. Next, this chapter discussed the appeals advertisers have implemented in addressing the working mother target market. The majority of advertisers targeting this market have chosen "slice of life" portrayals that feature a working mother character. Two main types of portrayals can be recognized within this category of working mother appeals. Some advertisers have portrayed a working mother character who singularly performs her multiple roles without any stress or conflict between roles. These portrayals have attracted a considerable amount of criticism from consumers and industry members, who have evaluated the portrayals as unrealistic and exaggerated. In contrast, other advertising portrayals have emerged in which the working mother character is portrayed performing multiple roles, but also is portrayed sharing responsibility for a domestic and/or childcaring task with her spouse. These latter portrayals that featured task sharing have not received criticism as the previous portrayals have. However, no research has been conducted to determine whether working mothers actually perceive the portrayal types as different along the task—sharing dimension. Thus this experimental study’s purpose is to examine whether working mothers do perceive different amounts of task sharing in the two portrayal types. In addition, this study intends to explore whether one portrayal strategy is more effective or irritating than the other. This study also questions whether either portrayal type can affect working mothers’ attitudes toward and intent to purchase the product. And finally, the two portrayal types’ ability to affect working mothers’ beliefs along the following dimensions is explored: working mothers’ job status, degree of task sharing and role conflict, and expectations for task sharing behavior in dual— earner marriages . This manuscript contains four additional chapters following this introduction. The next chapter, "Literature Review," will discuss sociological studies which have explored the topic of married working mothers, and advertising studies which have examined the content and 16 effectiveness of role portrayal variations. Next, "Methodology" will describe the research hypotheses as well as the experiment’s development, design, sampling procedures and execution. The following chapter, "Analyses and Results," will discuss the sample profile, data treatments, and the hypothesis test results. Lastly, the "Conclusion" chapter will discuss the limitations involved in this study and will provide conclusions and recommendations based on the hypothesis tests. CHAPTER 2 LITERATURE REVIEW Introduction The following chapter presents a review of related studies in two main sections: sociological literature examining dual—earner couples, role conflict and task sharing, and advertising literature that explores the content and effectiveness of women’s role portrayals in advertising. Sociological Literature Over the past decade, sociologists have recognized women’s increasing participation in the labor force and have chosen to study this trend’s effects on working mothers and their families. Some studies have focused specifically on working mothers’ experiences with multiple roles, while others have examined the allocation of domestic and child care tasks from a family perspective. One researcher chose to examine the impact of women’s multiple-role commitments on level of life satisfaction. Faver (1982) distinguished between women’s objective status and subjective value orientation. Objective status was 17 defined as the woman’s roles, including marital, parental and employment status. Subjective value orientation involved not only how women "enact the roles of wife, mother, and paid worker, but also their subjective experience of, and attraction to, these roles."21 Faver states that other researchers have acknowledged the effects of role structures on individuals’ mental health: they suggest that life satisfaction depends on whether indiv19231§;_objective status is congruent with their subjective expectations (Campbell, W\ \‘ 1975; Kanter, 1977a.) “\F" Faver secondarily analyzed 1977 data collected on a non- probability, cross—sectional sample of 1,120 women ages 22-64. The sample was split into three cohorts categorized according to age, marital status, and age of youngest child. Two attitudinal variables were measured~—value orientation and life satisfactiOn. The researcher conducted one-way analyses of variance to test for life satisfaction score differences between age, marital status and employment status groups. Faver found among the results that for married women with high career values, paid employment makes a significant contribution to life satisfaction. In addition, married mothers of pre- school aged children were the least satisfied life—cycle group in the youngest cohort. These women were especially dissatisfied if they had high career or low family values. 19 The researcher also observed that a sharp discrepancy between the level of career values and the rate of full-time employment among young mothers suggests that withdrawal from the labor force during early motherhood reflects constraint rather than choice (Faver, 1981.) Faver states, "this suggests that high career orientations are incompatible with the demands of early motherhood as it is structured in our society."22 Overall, Faver concluded that women’s life satisfaction is not dependent on role configuration alone, but that women’s well—being is a function of the fit between values and- roles. Thus structual changes that facilitate women’s career participation and provide societal support for childrearing might reduce dissatisfaction among young mothers with high career values: "our findings suggest that a broadening of career and family alternatives is needed in order to increase the probability of high satisfaction among women."23 Another researcher, Lieber (1980) examined working mothers’ experiences with multiple roles, but chose to explore role conflict and task sharing issues. Lieber conducted a small—scale qualitative research project "to assess the feelings, attitudes, and coping skills" of a nonrandom sample of thirty women. The women’s ages ranged from twenty—six to fifty-three, with the largest number of women in their thirties. Of these women, who were 20 predominantly professionals, most stated that their husbands were extremely helpful in sharing home and child care responsibilities; many husbands shopped and cooked on a regular basis. Almost the entire group stated that they experienced role conflict. Only two of the women expressed little conflict in managing their roles, and these women worked part—time. The women stated that they wanted to be able to spend more time with their children and husbands, but also find personal time for themselves and time for necessities‘“) such as housekeeping. One woman said that "both home andl// work want much of my time. I’ve tried to strike a balance which is comfortable to me, but it has meant some curtailing/// of ambition and job opportunities."24 The majority of the women saw their mother role as primary in their lives. Lieber came to several conclusions based on these interviews. First, a mother cannot satisfactorily handle multiple commitments without household help. Second, the husband’s attitude and supportiveness is an important factor in the wife’s ability to manage these roles. Lieber also concluded that for women attempting to establish careers, family demands are particularly difficult to meet. Nonetheless, these women stated that their lifestyles were worthwhile and rewarding despite the difficulties and demands. 21 However, Lieber’s conclusions are limited in that they are biased toward professional women and are thus not representative of working mothers as a population. However, the study does reveal a general consensus that women who are invested in their careers feel conflict from multiple roles and find that conflict is alleviated with help from their spouses. One research team chose to examine multiple roles from a family perspective. Maret and Finlay (1984) examined dual- career couples’ management of domestic and child-caring roles. These researchers investigated the hypothesis that women in dual—earner families maintain full responsibility for domestic tasks involving the care of the home and children. This research had three main goals: 1) to describe the extent to which women in dual—earner families bear sole responsibility for domestic tasks such as child care, meal preparation, and housekeeping; 2) to estimate change in such responsibility; 3) to indentify major correlates of observed variability in the extent of women’s domestic responsibilities. As Maret and Finlay state, "insofar as women’s roles are now linked to the occupational setting, one would expect some attenuation of role specialization within the conjugal setting."25 Other studies found that employed wives spend less time doing housework than do non-working wives and that working mothers’ husbands 22 and children share domestic tasks more than unemployed mothers’ husbands and children do (Blood and Wolfe, 1960.) Maret and Finlay conducted secondary analyses on data from the 1967—1977 National Longitudinal Surveys of Work Experience (NLS data). These data were gathered from a national probability sample of women interviewed eight times. The respondents’ ages were 30 to 44 at the time of the first survey in 1967. The researchers extracted a sub—sample of 1223 women who were married with dependent age children living in the home; another sub-sample was used in which the women were required to be employed at the time of the survey in 1977, resulting in a sample size of 622. The results indicated that the married mothers in sample do bear considerable responsibility but not responsibility for domestic tasks and that all women not alike in their levels of responsibility. Women in earner families were found to have higher levels of sharing than women not working outside the Approximately two—thirds of the working mothers were final the sole were dual— taski) 2 homen/ solely responsible for grocery shopping, cooking, and washing clothes. Less than half the women were solely responsible for child care, washing dishes and house-cleaning. The data also demonstrated a trend toward more familial task sharing. For every task category, fewer women solely responsible for the task in 1976 than in were 1974. 23 However, this difference was significant for only three task categories--grocery shopping, child care, and clothes— washing. Twelve correlates of home responsibility were included in the analysis, five of which were significant: race, resident, husband’s income, labor-force attachment, and wife’s wages. The data showed that "lower levels of home responsibility are found among black rather than white women, among those with greater labor-force attachment, among women with higher earned incomes, among residents of SMSA’s, and among those whose husbands had low rather than high incomes. . .husband’s attitude toward women’s roles was also found to be related with respondents’ level of the home responsibility.26 The researchers concluded that home responsibilities are variable across women. They also concluded that there was a decrease in the amount of domestic responsibility for married women between 1974 and 1976 and that working married women had lower levels of home responsibility than non- working wives. In light of these conclusions, the“.r researchers suggest that as married women’s labor force! participation increases and "as men and women approximate equality in the work place, they will move toward more egalitarian sharing of domestic responsibilities."27 Another team of researchers Holahan and Gilbert (1979), investigated role onflict in dual-earner families by 24 exploring the relative amounts of "interrole conflict" experienced by working wives and husbands. They hypothesized that women in dual-earner families would be found to experience more role conflict than their spouses. This hypothesis was based on previous studies which had found that the increase in dual—earner marriage has been accompanied by a greater expansion of sanctioned roles for the wife than for the husband because women are adding a role when they seek paid employment outside the home (Bernard, 1975; Rapoport & Rapoport, 1969.) The researchers specifically hypothesized that the highest amounts of stress would be experienced by married women with children. A last purpose for the study was to identify correlates of role conflict for women in dual-earner marriages. The study involved mailing questionnaires to a sample of dual-earner couples, 28 of which completed and returned the questionnaires. Three variables were of interest as potential correlates. High career commitment, high spouse support, and profeminist attitudes toward the role of women were predicted to be associated with women’s lower interrole conflict. Role conflict was measured using six scales representing conflict between four major life roles—- professional, spouse, parent, self. Each scale measured a potential conflict between the four roles: professional vs. spouse, professional vs. parent, professional vs. self, spouse vs. self, spouse vs. parent, and parent vs. self. 25 The study’s analyses showed that no significant gender differences were found for any of the role conflict scales that did not involve the parent role. Of those three scales that did include the parent role, only one significant effect emerged: a parenthood main effect was found for spouse vs. self, with the parent group reporting greater conflict than the nonparent group. Still no significant main effect was found for gender. Analyses of potential correlates revealed that lower conflict for both genders was related to higher self-esteem, more profeminist attitudes, higher satisfaction with major life roles, fewer working hours, higher spouse support, and higher career commitment. For non-parents, the following variables were found to be indicators of high role conflict: fewer working hours, low career commitment, and low aspiration level. The only similiar correlate of high role conflict for parents was more working hours. Other correlates of high role conflict for parents included traditional attitudes, unfavorable attitudes of spouse, lack of spouse support, and negative feelings from spouse about one’s career commitment. The researchers summarized that for dual-earner couples without children, spouse support was not a correlate while personal variables related to career pursuits were. For dual-earner parents, spouse support variables were correlates while personal, career—related variables were not. 26 Holahan and Gilbert concluded that the variables related to high role conflict in the parent group indicate that "the addition of the parent role appears to complicate the life situation of the couple, making dual-career pursuit more difficult to maintain."28 The researchers contrast these findings to the situation of couples without children, who can "act relatively independently without the constraints of children."29 Advertising Literature Many researchers have studied the content of women’s role portrayals in advertising. The majority of those studies concluded that women in print media and television were predominantly portrayed in domestic roles (Courtney and Lockeretz, 1971; Courtney and Whipple, 1974; Venkatesan and Losco, 1974.) However, only recently have researchers begun to examine the role of paid work for female characters in advertising portrayals. Studies examining media portrayals of working women and working mothers have taken two main forms: content analyses, which have quantified depictions of women in various roles; and preference studies, which have gathered viewers’ evaluative reactions to particular print or broadcast portrayals. 27 Currently, no major studies analyzing the content of television commercials’ portrayals of women have focused on the topics of role conflict or task sharing exclusively. Although not an advertising study, one study examined the content of working mother portrayals in women’s magazines. This researcher (Henry, 1984) states that previous studies of magazine content indicate "married women’s roles as paid workers have received little recognition."3° Henry cites several journalism studies of women’s magazine fiction. These studies concluded that married women were most often portrayed as not having jobs or were portrayed with an emphasis on family concerns rather than job concerns (Franzwa, 1974; Loughlin, 1983; Lugenbeel, 1975.) Other studies found that women’s employment outside the home is a more common topic in these magazines’ non-fiction articles. However, the researcher importantly acknowledges that magazines will most likely deal differently with the role of work in women’s lives, depending on the magazine’s audience. Henry’s study involved qualitative analysis of the portrayal of women’s work and family lives in seven magazines from 1975 to 1982. Analysis included only non—fiction articles whose subjects were employed, married women with children. The main goal of the study was to examine the "kind and amount of information given about the effort required for women to combine a family and employment."31 In particular, the study aimed to determine whether articles 28 advocated that working mothers be primarily responsible for family household work, an image the researcher described as congruent with the "Superwoman" image. Henry reported the content of each magazine separately but did state general conclusions. One major conclusion was that all seven magazines were found to contain articles that promoted women working a "double day," that is, working full- time outside the home while maintaining major responsibility for domestic tasks. One third of the articles "unambiguously supported the concept of women working a double day." The researcher’s last conclusion was that the seven magazines did not present "realistic alternatives to the double day." In 19 of the 64 articles, the double day was supported, while the others offered few alternative coping strategies: "even when particular solutions to the problems of the double day were presented, they frequently are pictured as being implemented with unquestioned ease and full satisfaction."32 These findings were consistent over the period of 1975-1982. Another body of literature examining media portrayals of working women’s roles is that of preference studies. These studies, indirectly, investigate the theory that one’s attitude toward the ad may mediate apects of advertising effectiveness, particularly brand preference or brand liking. Several studies have directly investigated this theory. One theory researchers have tested is that a simple positive 29 relationship exists between the advertisement and reactions to the brand (Shimp, 1981). This research asserts that a person’s reaction to the commercial will be generalized to the brand (MacKenzie and Lutz, 1982; Mitchell and Olson, 1981.) One writer summarizes the theory: "the more they like the ad, the more they like the brand."33 However, a competing theory argues that the relationship between ad liking and the ad’s effectiveness is J—shaped. Researchers exploring this theory suggest that neutral advertisements, those generating little or no affect, are least effective. Advertisements that are "affectively—valenced" (Gresham and Shimp, 1985), meaning they are either disliked or liked, will be more effective than neutral messages. Well-liked advertisements, however, are expected to most positively affect brand attitudes (Moore and Hutchinson, 1983.) These theories necessarily impact upon studies examining viewers’ preferences for and reactions to working mother portrayals in advertising. Some commercial portrayals of working mothers have elicited strong consumer complaints; thus it is possible that these complaints, which represent an aspect of dislike for the commercial, could be generalized to a dislike for the advertised product. However, as alternative theories propose, ads that generated negative consumer affect in the form of irritation may still be 30 effective. Nonetheless, both theories suggest that ads provoking positive affect will likely have the strongest relationship to positive brand attitudes. Several researchers have specifically examined the relationship between attitude toward working woman portrayals and preference for the advertised product. One such study by Wortzel and Frisbie (1974) explored whether women’s pro—feminist attitudes were related to those womens’ judgments of a product’s desirability when the product’s advertisement featured a working woman. The researchers however found that pro-feminist attitudes were not predictive of the product’s desirability ratings. A similiar study by Duker and Tucker (1977) attempted to predict women’s judgements of various role portrayals using feminism as the predictor variable. Again, feminism did not significantly predict women’s preference for role portrayals in print advertisements. Another study, focusing on preference for traditional versus professional commercial portrayals of women, selected gender as a possible predictor of preferences for women’s roles in seven "slice of life" portrayals (Kelly, Solomon, and Burke, 1977). This study concluded that male students preferred the professional portrayals, while female students liked both professional and traditional portrayals. 31 Another researcher questioned whether women’s attitudes toward their multiple roles--wife, mother, worker——would predict their preferences for print advertisements’ role portrayals (Swartz, 1983.) The study utilized a convenience sample of 88 women between the ages of 18 and 65; thirty- three of the participants worked outside the home. The study used a "portfolio approach" in which the subject was given two portfolios--one containing pictures of products and one containing pictures of women in different roles, one of which pictured a working woman. The subjects were asked to match the product with the role portrayal that maximized that product’s desirability. Subjects were also asked to complete a measure of their perceived sex role. In analysis, both the advertising role depictions and the subjects’ perceived sex roles were collapsed into two categories: modern and traditional. Chi—square analyses were then performed to test the hypothesis that women prefer role portrayals that match their own perceived role status. However, there were no significant chi sqaure values supporting this hypothesis. Further analyses revealed no dominant preferences within product categories except one: traditional role portrayals were significantly preferred for food products. Another study chose to examine both consumers’ and industry members’ preferences for television commercial role portrayals. Whipple and Courtney (1980) chose to measure the 32 effectiveness of six commercials which were separated into two categories: traditional and progressive. The traditional portrayals included commercials that featured a female character dish-washing, a female character preparing a meal for her husband and children, and two women competing in a floor mopping contest. The progressive portrayals included commercials that featured a professional woman with her male co-workers, a husband and wife cooperating in meal preparation, and a husband and wife competing in a floor mopping contest. A representative, random consumer sample of 261 eligible women was drawn. Of this total, 144 interviews were completed. The practitioner sample included 68 consumer product manufacturers and advertising agency members whose titles ranged from advertising directors and brand managers to account executives and creative directors. The study involved two measures: unaided playback of each commercial’s major selling points and ratings of the commercial along a set of 14 evaluative items. These items described the commercial’s effectiveness or irritation value. Analysis revealed that all commercials received high levels of message playback, regardless of the role depiction. In terms of effectiveness, the only significant difference between the two role portrayal types occurred for the meal preparation depiction. Both consumers and practitioners rated the progressive commercial more favorably than the 33 traditional commercial. For the irritation measure, the traditional dish-washing depiction was judged by consumers and practitioners as more irritating than any of the other portrayals. Both the meal preparation and floor mopping commercials were judged as low in irritation, while both consumers and practitioners judged the progressive portrayals to be less insulting to women and less irritating. Whipple and Courtney concluded that progressive sex—role portrayals were rated equally or preferred to the traditional commercial portrayals. The researchers stated that the study demonstrated that progressive role depictions can be created effectively and realistically without creating consumer irritation and also suggested that "advertising should now be exploring more modern approaches to the portrayal of men and women as they demonstrate products and their benefits to the television audience."34 Summary This chapter has detailed literature related to the examination of working mother commercial portrayals. First presented were sociological studies examining the role of work in married women’s personal and family lives. One researcher examined the relationship between women’s roles and their life satisfaction and found that married mothers of pre—school age children were least satisfied of all the 34 study’s cohort groups. Faver concluded that these women’s withdrawal from the labor force reflects contraining role responsibilities. The researcher suggested changes in society’s role structure, such as societal support for childrearing, to improve mothers’ life satisfaction. Another researcher, Lieber, mirrored these conclusions. Lieber conducted interviews with working mothers who stated they experienced conflict between their multiple commitments. However, these women stated that assistance anstupporthrom their spouse was a factor in reducing such conflict. I ’Other researchers such as Maret and Finlay, and Holahan and Gilbert examined family dynamics in response to the mother working outside the home. Maret and Finlay concluded that married working mothers do bear considerable responsibility for domestic tasks, but the extent of responsibility varies across women. The data, however, revealed a trend toward increased domestic task sharing between working mothers and their husbands. Holahan and Gilbert investigated interrole conflict for dual—earner couples with children. For these couples, low role conflict was related to fewer working hours, non-traditional attitudes, favorable attitudes of spouse, spouse support, and positive feelings from spouse about one’s job commitment. These studies affirm that women do experience conflict between multiple roles--marriage, motherhood, employment. These data also suggest that the structure of role I: I 35 responsibilities is changing toward a more equal distribution of domestic and childcare responsibilities between working mothers and fathers. Also suggested is that this trend toward role diversification has the potential to attenuate working mothers’ degree of role conflict. The review of advertising literature focused on research that has examined content of and preference for particular portrayals of working women. First, this section discussed content studies which have found that portrayals of working women and working mothers have not been prevalent in women’s magazine articles. Articles that did discuss women’s work outside the home generally accepted the "double day" concept: women who work outside the home who are also primarily responsible for domestic and child care responsibilites. The next section examined consumers’ preference for particular portrayals of working women and working mothers. Underlying these studies are theories that have related ad liking to brand liking. Data have been found that support two theories: first, as liking for an ad increases, liking of the brand increases; second, an ad which is either disliked or liked will be more effective than an ad which elicits little or no affective response. These theories are helpful when examining the effectiveness of working mother portrayals, some of which have been criticized as unrealistic and exaggerated. It is possible that these commercials, despite their irritation 36 value, may still be effective. However, both affect theories suggest that ads which elicit the most positive affect will likely be most effective. Studies examining consumers’ preferences for role portrayals of working women have produced varied results. Several studies have attempted to predict consumers’ preferences for working woman portrayals from variables such as gender, pro—feminist attitudes, and respondent’s own roles; however, none of these variables successfully predicted commercial role portrayal preferences. The last preference study discussed examined business practitioners’ and female consumers’ recall and evaluation of traditional and progessive portrayals. In both samples for all commercials, the progressive portrayals were equal or preferred to the traditional portrayals. CHAPTER III METHODOLOGY Introduction The following chapter presents this study’s research hypotheses and methodology. The experiment involved a quota sample of married working mothers which was divided into two groups. One group viewed four videotaped commercials featuring married working mother characters who participate in task sharing with their spouses; the second group viewed four videotaped commercials featuring married working mothers who do not participate in task sharing. Eleven hypotheses are examined in this study. The first three hypotheses—Lthe commercial attribute hypotheses-— examine subjects’ ratings of the two portrayal types in regard to task sharing, role conflict, and degree of similiarity between the subject and commercial character. The next hypotheses——the commercial effectiveness hypotheses-- examine subjects’ judgements of the portrayals’ effectiveness. Lastly, the sales effect and the womens’ beliefs hypotheses are presented. These hypotheses investigate the potential relationship between exposure to "sharing" or "non—sharing" portrayals, and subjects’ attitudes toward the product and beliefs about women’s roles. 37 38 Hypotheses Commercial Attribute Hypotheses The following hypotheses are designed to measure three aspects of the test commercials’ attributes. Those aspects include the following: the amount of task sharing behavior conveyed in the commercial, the degree of role conflict experienced by the working mother character in the commercial, and the perceived similiarity between the subject and the working mother character in the commercial. Each of these hypotheses is presented and discussed individually in the following sections. Hypothesis I This first hypothesis examines the comparative amount of task sharing conveyed by the "sharing" and "non—sharing" portrayals. As previously stated, sharing portrayals are operationalized by the researcher as portrayals that feature a working mother sharing a household and/or child-caring activity with her spouse. Non-sharing portrayals are defined 1 as portrayals that feature a working mother who is shown performing household and/or childcare tasks singularly and/J thus is not shown sharing any task with her spouse. Although sociological literature proposes that task sharing is variable across individuals, there is no previous literature that applies this theory to advertising portrayals 39 of working mothers. Likewise, no previous advertising studies have used task sharing as a variable to distinguish between portrayal types. Thus the proposed difference Abetween "sharing" and "non-sharing" portrayals is a distinction which currently has no empirical support. Hence the first of the "commercial attribute" hypotheses is intended to examine subjects' evaluations of the amount of task sharing conveyed by the two portrayal types. Moreover, this hypothesis is offered to determine whether the distinction between "sharing" and "non-sharing" portrayals can be empirically supported. If this hypothesis is supported, then the difference between the two portrayal types will provide the empirical basis for additional comparisons between the two experimental groups. The research and null hypotheses are as follow: H-I: Subjects will report a greater amount of task sharing in the commercials categorized as "sharing" portrayals than in "non—sharing" portrayals. HO—I: Subjects will report an equal or lesser amount or task sharing in the commercials categorized as "sharing" portrayals than in the "non-sharing" portrayals. In this hypothesis, task sharing is operationalized with six items which explore task sharing dimensions that were extracted from sociological literature and through personal interviews with working mothers. These six items are listed in Table 3.1. 40 TABLE 3.1 Task Sharing Items 1. This working mother’s husband spends as much time taking care of their child/children as does his working wife. 2. This working mother’s husband does no dish-washing or clothes-washing. 3. This working mother’s husband does not spend as much time "taxi—ing" their child/children around as does his working wife. 4. This working mother’s husband does grocery-shopping for their family. 5. This working mother’s husband does not prepare family meals. 6. This working mother spends as much time as her husband attending their child/children’s activities. 7. The working mother in this ad spends more time than her husband helping their child/children with their homework. Hypothesis II The second hypothesis explores the topic of role conflict in commercial portrayals of working mothers. It is predicted that subjects will Judge the "sharing" characters “Jr ___.__...«..._i__i.___i__-____-g",_,__,.~—.———’—~'”""'"‘”‘“"‘" as experiencing more role conflicfvfihgn the.\flpgn:sharing" characters. This relationship is expected because consumer and industry critics have stated that the "non-sharing" portrayals present working mother characters who "do it all"——family and work duties-—with no apparent stress and 41 with no apparent need for domestic task sharing. Conversely, the "sharing" portrayals present working mother characters who perform family and work duties but participate in domestic task sharing with their spouses. By featuring tasks shared by the couple, these portrayals directly address the issue of role conflict and the stress that can be experienced by multiple-role couples. In contrast, the "non- sharing" portrayals do not suggest that their working mother characters experience any such role-related stress. The second research and null hypotheses are as follow: H—II: Subjects will rate the working mother characters in "sharing" portrayals as experiencing more role conflict than the working mother characters in "non-sharing" portrayals. HO—II: Subjects will rate the working mother characters in "sharing" portrayals as experiencing the same or less role conflict than the working mother characters in "non—sharing" portrayals. In this hypothesis, role conflict is operationalized with five items. These items explore the extent to which the working mother character is judged to experience conflict between her commitments to family and her commitments to work. These five items are listed in Table 3.2. 42 Table 3.2 Role Conflict Items 1. The working mother in this ad has trouble balancing her work responsibilities with her family responsibilities. 2. This working mother is able to spend the desired amount of time at her job while being able to spend enough time with her children. 3. The working mother in this ad does not find having a job outside the home stressful. 4. The working mother in this ad is dissatisfied with her role as mother. 5. The working mother in this ad is satisfied with her job. Hypothesis III This hypothesis proposes that subjects will report the degree of task sharing they experience in their personal lives as more similiar to the "sharing" portrayals than the "non—sharing" portrayals. This hypothesis is based on literature and personal interviews that indicate a shift toward less traditionally—structured roles. This shift, which is apparent in many dual-career family situations, especially those families with children, consists of a trend toward greater sharing of domestic and childcaring responsibilities between working mothers and their husbands. In other words, it is predicted that the subject will 43 perceive the portrayals in which the working mother is assisted by her spouse in a domestic and/or child—caring task as more similiar to her own situation than the portrayals in which task sharing does not occur. The third research and null hypotheses are as follow: H—III: Subjects will judge the "sharing" characters as more similiar to their personal degree of task sharing than the "non—sharing" characters. HO-III: Subjects will judge the "sharing" characters as equally or less similiar to their personal degree of task sharing than the "non-sharing" characters. Perceived similiarity, in this hypothesis, is operationalized with eight items which explore similiarities between the subject and the working mother commercial characters. The items focus on the same content areas as were mentioned in the "task sharing" and "role conflict" hypotheses. However, in this case the subject is asked to compare her own degree of task sharing and role conflict to the degree of task sharing she judges the commercial character to experience. These eight items are presented in Table 3.3. 44 Table 3.3 Perceived Similiarity Items My life is very similiar to that of the working woman featured in this ad. I would like my life to be similiar to that of the working mother in this ad. I experience more job—related stress than the working mother in this ad. The working mother in this ad is happier being a mother than I am. My spouse and I do not share child—care responsibilities as much as the working mother and her spouse in this ad do. My spouse helps out with dish—washing and clothes-washing as much as this working mother’s spouse does in the commercial. My spouse does not help prepare family meals as often as often as the husband in this ad does. This working mother’s husband grocery shops for their family more often than does my husband. 45 Qggmercial Effectiveness Hypotheses As previously stated, the following two hypotheses are contingent on the task sharing hypothesis, since it is assumed that empirical evidence has shown the two portrayal types as significantly different along the aforementioned task—sharing dimension. Given this assumption, the following hypotheses explore subjects’ evaluations of "sharing" and "non—sharing" portrayals. Hypothesis IV The first commercial effectiveness hypothesis proposes that the commercials in the "sharing" group will be judged as more effective than the "non—sharing" commercials. This hypothesis is based on the expectation that subjects will respond more positively to commercials in which task sharing is present than to those in which there is no task sharing. The research and null hypotheses are as follow: H-IV: Subjects will judge the "sharing" portrayals as higher in effectiveness than the "non-sharing" portrayals. HO—IV: Subjects will judge the "sharing" portrayals as equally or less effective than the "non—sharing" portrayals. The concept of "effectiveness" in this hypothesis is operationalized with eight items that represent positive responses to the commercial. This group of items was modeled 46 after a scaling scheme used in a previous study of women’s portrayals in advertising (Whipple and Courtney, 1980.) These items are listed on the following page in Table 3.4. Hypothesis V The second commercial effectiveness hypothesis examines the potential for irritation as an affective response to the commercial portrayals. Since "non—sharing" portrayals have received the majority of consumer and industry criticism, it is expected that subjects will react more negatively to the "non—sharing" portrayals than to the "sharing" portrayals. Thus, this hypothesis proposes that the "non—sharing" portrayals will be rated as more irritating than the "sharing" portrayals. The research and null hypotheses are as follow: H—V: Subjects will judge the "non—sharing" portrayals as higher in irritation than the "sharing" portrayals. HO-V: Subjects will judge the "non—sharing" portrayals as the same or lower in irritation than the "sharing" portrayals. In this case, irritation is operationalized with seven items that represent negative reactions to the commercial. These items, which were also modeled after the Whipple and Courtney effectiveness/irritation scheme, are listed in Table 3.5. 47 Table 3.4 Effectiveness Items 1. Portrays a realistic situation 2. Appeals to women like me 3. Something I will remember 4 Provides me with useful information 5. Convinces me about the product’s effectiveness 6 Makes me feel good about myself 7. Makes me feel good about my spouse 8. Makes me want to try the product Sales Effect Hypotheses This study’s first five hypotheses required the subject to view and evaluate each commercial individually. The next two hypotheses, however, are not concerned with the commercials’ individual ratings; instead, the following hypotheses intend to measure the cumulative effects of each experimental group’s four commercials. The two sales effect hypotheses-—product liking and likelihood to purchase--will compare a pre—test measure, completed before any commercial exposures, to an identical post—test measure which is completed after viewing and evaluating all four commercials. 48 Table 3.5 Irritation Items l. Irritating to working mothers. 2. Irritating to housewives 3. Irritating to working fathers 4 Irritating to someone like me 5. Gives me a bad feeling about myself 6 Upsetting to me 7. Gives me a bad feeling about my spouse Hypothesis VI The first sales effect hypothesis examines product liking. It is hypothesized that subjects’ liking of products advertised in the "sharing" experimental group will increase after viewing the four commercials. Conversely, it is hypothesized that subjects’ liking of products advertised in the "non—sharing" experimental group will decrease after viewing the four commercials. The research and null hypotheses are as follow: HO—VI: 49 Subjects’ liking ratings for the products featured in the "sharing" portrayals will increase after viewing the four "sharing" commercials; subjects ratings of products featured in the "non—sharing" portrayals will decrease after viewing the four "non-sharing" commercials. Subjects’ liking ratings for the products featured in the "sharing" portrayals will be the same or lower after viewing the four "sharing" commercials; subjects’ ratings of products featured in the "non—sharing" portrayals will be the same or higher after viewing the four "non—sharing" commercials. The product liking variable is operationalized as subjects’ numerical rating on a zero to ten scale of how much they like each of the four test products. Hypothesis VII The second and last sales effect hypothesis focuses measuring the subjects’ intent to purchase the product. research and null hypotheses are as follow: H-VII: HO-VII: Subject ratings of how likely they are to purchase the products featured in the "sharing" portrayals will be higher after viewing the four "sharing" commercials; subject likeliness ratings for products featured in the "non-sharing" portrayals will be lower after viewing the four "non—sharing" commercials. Subject ratings of how strongly they intend to purchase the products featured in the "sharing" portrayals will be the same or lower after viewing the four "sharing" commercials; subject ratings of products featured in the "non-sharing" portrayals will be the same or higher after viewing the four "non—sharing" commercials. 50 The intent to purchase variable is operationalized as the subject’s numerical rating along a scale of zero to ten of how likely they are to purchase the test product. Women’s Beliefs Hypotheses As with the sales effect hypotheses, the women’s beliefs hypotheses are concerned with the commercials’ cummulative effects. However, while the sales effect hypotheses measured the subjects’ pre-test/post—test attitudes toward the product, the following hypotheses concern the subject’s pre— test/post-test beliefs about working mothers. These beliefs are presented in four main categories: job status, task sharing, role conflict, and role expectations. Hypothesis VIII The first hypothesis in this section examines working mother’s beliefs about other working women’s job status, which involves the types of jobs that working women have. Content analyses have suggested that in the past, women have primarily been portrayed in television commercials and print advertising as domestics. Although that method of portrayal still exists, the number of commercial portrayals of women working outside the home is growing. Critics state, however, that these advertisements predominantly portray women in one particular role: the female executive. 51 As previously discussed, government data state that the majority of women do not hold managerial or professional positions. This hypothesis, then, is concerned with measuring the degree to which viewing commercials that portray women exclusively in one job type--executive or managerial-~can influence working mother’s beliefs about the types of jobs commonly held by working women. The research and null hypotheses are as follow: H—VIII: After viewing the four commercials, subjects in both experimental groups will agree more strongly that most working women have executive or managerial jobs. HO—VIII: After viewing the four commercials, subjects in both experimental groups will agree equally or less strongly that most working women have executive or managerial jobs. This hypothesis is operationalized with six items exploring beliefs about the types of jobs held by working women and working mothers and their average annual salaries. The actual items are presented in Table 3.6. '«r 4- 4f 52 Table 3.6 Job Status Belief Items 1. Most working women have executive or managerial jobs. 2. Most working mothers have executive or managerial jobs. 3. Few working women make more than $13,000 a year. 4. Few working mothers make more than $13,000 a year. 5. Most working women make more than $20,000 a year. 6 Most working mothers make more than $20,000 a year. Hypothesis IX The next hypothesis in this section explores subjects’ beliefs about working mothers and task sharing. This hypothesis also refers to the task sharing commercial attribute hypothesis, in that it is assumed that subjects have been found to judge "sharing" working mother characters and their spouses as participating in more task sharing than "non—sharing" characters. If this premise is empirically supported, then this hypothesis proposes that viewing "sharing" portrayals may increase the likelihood that subjects will agree that task sharing occurs between working mothers and their spouses. Likewise, it is proposed that viewing "non-sharing" 53 portrayals may increase the likelihood that subjects will not agree task sharing occurs between working mothers and their spouses. The research and null hypotheses are as follow: H-IX: After viewing the "sharing" portrayals, subjects will be more likely to agree that working mothers participate in task sharing with their spouses; after viewing the "non-sharing" portrayals, subjects will be less likely to believe that working mothers and their spouses participate in task sharing. HO-IX: After viewing the "sharing" portrayals, subjects will be equally or less likely to believe that working mothers participate in task sharing with their spouses; after viewing the "non—sharing" portrayals, subjects will be equally or more likely to believe that working mothers participate in task sharing with their spouses. In this hypothesis, task sharing is operationalized with seven items modeled after those stated for Hypothesis 1. These items are presented on the following page in Table 3.7. Hypothesis X The third hypothesis in this section examines subjects’ beliefs about working mothers and role conflict. Like hypothesis IX, this hypothesis is contingent on one of the commercial attribute hypotheses——hypothesis II. It is assumed that support was found for this hypothesis, in that "sharing" working mother characters were judged by subjects as experiencing more role conflict than "non-sharing" characters. 54 Table 3.7 Task Sharing Belief Items 1. Working fathers spend as much time with their children as do working mothers. 2. Working women receive help from their spouse when it comes to household chores such as dish-washing or clothes-washing. 3. Working fathers help their working wives when it comes to fixing family meals. 4. Working fathers spend as much time "taxi-ing" their kids as do working mothers. 5. Working fathers do not spend as much time attending their kids’ activities as do working mothers. 6. Working fathers spend as much time helping their child or children with homework as does the working mother. 7. Working fathers do not grocery shop for their families as often as do working mothers. Thus, Hypothesis X asserts that viewing "sharing" portrayals may increase the likelihood that subjects will believe working mothers experience role conflict. In contrast, "non—sharing" portrayals may decrease the likelihood that subjects will believe working mothers experience role conflict. The research and null hypotheses are as follow: 55 H—X: After viewing the "sharing" portrayals, subjects will be more likely to agree that working mothers experience role conflict; after viewing the "non-sharing" portrayals, subjects will be less likely to agree that working mothers experience role conflict. HO-X: After viewing the "sharing" portrayals, subjects will be equally or less likely to agree that working mothers experience role conflict; after viewing the "non-sharing" portrayals, subjects will be equally or more likely to agree that working mother experience role conflict. Beliefs about role conflict is operationalized with three items modeled after the first three listed for Hypothesis II. The role conflict belief items are listed in Table 3.8. Hypothesis XI The final women’s belief hypothesis introduces the topic of role expectations. This hypothesis proposes that a subject’s knowledge of other working mothers’ task sharing experiences may influence her beliefs about how much task sharing is desirable or expected. For example, if a subject is exposed to commercial portrayals of working mothers who participate in task sharing with their spouses, then she may be more likely to expect that working mothers and their husbands participate in task sharing. In contrast, if a subject is exposed to commercial portrayals of working 56 Table 3.8 Role Conflict Belief Items 1. Working mothers easily balance their job responsi— bilities with their family responsibilities. 2. Working mothers are able to spend the desired amount of time at work and also spend enough time with their child/children. 3. Working mothers do not find having a job outside the home stressful. mothers who do not task share and do not express any stress due to role conflicts, then the subject may be less likely to expect task sharing to be an expected behavior. The research and null hypotheses are as follow: H-XI: HO-XI: After viewing the "sharing" portrayals, subjects will be more likely to expect task sharing to occur between working mothers and their spouses; after viewing the "non- sharing" portrayals, subjects will be less likely to expect task sharing to occur between working mothers and their spouses. After viewing the "sharing" portrayals, subjects will be equally or less likely to expect task sharing to occur between working mothers and their spouses; after viewing the "non-sharing" portrayals, subjects will be equally more likely to expect task sharing to occur between working mothers and their spouses. 57 In this last hypothesis, role expectation is operationalized by six items which are modeled after those used in Hypotheses I and IX. The role expectation items, however, are different in that they assess the types of sharing behavior the subjects states should be expected to take place between a working mother and her spouse. These items are listed in Table 3.9. Table 3.9 Role Expectation Items l. A working father should not be expected to spend as much time with his kids as does his working wife. 2. A working father should be expected to spend as much time helping his children with homework as does his working wife. 3. A working father should not be expected to spend as much time dish-washing and clothes—washing as does his working wife. 4. A working father should not be expected to be responsible for preparing family meals. 5. A working father should help perform household chores such as grocery shopping. 6. A working father should not be expected to spend as much time "taxiing" their child/children around as does his working wife. 58 Questionnaire Development This study’s questionnaire design mirrors the experimental design, which is discussed in the following section. The questionnaire contains a pre-test, four commercial evaluation measures, and a post—test. The majority of these measures use zero to ten scaling which are ordinal but are assumed to be interval for analytical purposes. Other measures are taken at the ratio level using raw numerical estimates, while several demographic questions are measured nominally. The following section will discuss each portion of the questionnaire and its purpose. A sample questionnaire is presented in Appendix A. Pre-test The first section of the pre-test is a product liking measure. This measure asks the subject to rate a list of 24 products from zero, "Do not like at all," to ten, "Like totally." The second section contains the same list of products, but asks the subject to rate them in terms of likelihood to purchase; in this section the scaling ranges from zero, "Not at all likely," to ten, "Totally likely." Both of the product lists are identical, containing the names of the test products to be featured in that experimental group’s commercial evaluation; these test 59 product names are imbedded among a list of non-test product names for a total of 24 products to be evaluated by the subject. The pre-test’s last section asks the subject to respond to several statements regarding her beliefs about women’s roles using zero, "Totally disagree," to ten, "Totally agree." Four aspects of women’s roles are examined: job status, role conflict, task sharing, and role expectations. The "job status" section includes six items examining the subject’s beliefs in two areas: types of jobs most working women/working mothers have and average salary. The "role conflict" measure explores the subject’s beliefs about working mothers’ ability to manage job and family responsibilities. The "task sharing" section explores beliefs about the degree to which husbands participate in household and child—caring activities. Finally, the "role expectations" section measures beliefs about the degree to which husbands should be expected to participate in task sharing. Commercial Evaluations The body of the questionnaire features the commercial evaluation measures which are repeated four times, one measure after each of the four test commercials. The 61 Finally, the commercial evaluation section concludes with a measure of "appropriateness." This single item asks the subject to evaluate how appropriate the commercial portrayal is for its product category. The scale range is from zero, "Not at all appropriate" to ten, "Totally appropriate." Post—test The first three portions of the post—test duplicate those featured in the pre-test: product liking, likeliness to purchase, and women’s beliefs. Two additional sections are included in the post—test for descriptive purposes: a personal role situation measure and a demographic section, which includes personal and family-related demographics and media usage information. Experiment Design The design chosen for this experiment is a pre-test/ post-test quasi-experimental design. The study is described as "quasi—experimental" for two reasons: subject selection is not random, and the study’s design does not include a control group. However, due to minimal previous research in this area, the primary focus of the study is exploratory. 62 Table 3.10 Experiment Design Pre- Commercial Post- issi Eyslsstigss test Group 1: X 0 X 0 X 0 X 0 X X Group 2: X 0 X 0 X 0 X 0 X X Participants gathered by quota sampling were assigned to one of two experimental groups. As Table 3.10 demonstrates, instrumentation for the two groups is identical. All subjects completed a pre—test measure, viewed the four videotaped commercials which they evaluated after each individual exposure, and finally completed the post-test and demographics section. The two experimental groups are distinguished by the type of portrayals viewed. Group one viewed portrayals which will be referred to as "sharing"; these commercials feature a working mother who is assisted by her spouse in either a domestic and/or a childcaring task. The second experimental group viewed portrayals that will be referred to as "non-sharing"; the working mother characters in these portrayals do not task share. 63 Experiment Preparation Stimuli Selection Professionally-produced commercials were selected as the experimental stimuli. Eight on-air, color, 60 second commercials were selected as stimuli based on their character content. Only ads that featured working mothers with a child or children were considered in the selection process. Four commercials were chosen based on the task sharing criterion for the "sharing" experimental group and four other commercials were chosen for the "non—sharing" group. Seven companies were contacted for permission to use the eight commercials (two of the commercials were different versions of the same product——one a "sharing" version and one a "non—sharing" version.) These organizations were contacted by phone or by mail. They were informed of the study’s purpose and a 3/4" copy of the specified commercial was requested for use in the study. All companies agreed to release a copy of the commercial except one. This last organization agreed to release the commercial after a formal contract had been signed by the researcher. This contract specified that the commercial would be viewed in an experimentally-controlled environment by sixty or more subjects; that no copies of the commercial would be made for any purpose; that the commercial would not be shown to any person not connected to the study nor used 64 for any purpose other than furthering the goals of this study; and the videotaped copy of the commercial and a copy of the research results would be forwarded to the company within a reasonable time following the completion of the study. UCHRIS & Union Permissions Several other groups needed to be contacted for permission to conduct this study. First, a description of sampling procedures, methodology, and consent procedures was submitted to the University Committee on Research Involving Human Subjects (UCHRIS). Permission to conduct the experiment was granted by the committee. However, UCHRIS required that the two employee unions—- the Clerical/Technical Union and the Supervisory Association——be contacted for permission to involve their employees in the study. Both unions agreed that their employees were free to participate under the conditions that participation was voluntary and that the employee understood release time or personal leave time must be requested to participate during normal business hours. 65 Sampling Sample Selection The subject sample of married working mothers was drawn using two criteria. First, the population of interest was defined as all female employees at Michigan State University in non—professional, non—academic positions. Women with non- professional, non-academic jobs were selected because they typify the majority of working women in the labor force in reagrd to occupation and annual salary. This population consisted of clerical/technical employees and supervisors whose positions included the following: clerks, secretaries, office assistants, office supervisors, cashiers, lab assistants, and lab technicians. The second criterion for the population of interest required these women to be married with one or more children under 16 years old living at home. Sampling Procedure The population of female C/T employees and supervisors was contacted with assistance from Michigan State University’s Office of Personnel and Employee Relations, and the Office of the Assistant Provost. A mailing list computer—generated by the Personnel office provided the names and campus addresses of all female clerical/technical employees and supervisors at MSU, a total of 1,979 women. 66 Participation Request Letters Two mailings were conducted to gather a sufficient number of participant volunteers. Both mailings consisted of three parts. First, the mailing included an introductory letter from the Assistant Provost announcing the opportunity to voluntarily participate in "a study of women’s attitudes toward advertising that features female characters." This letter also stated that the study had received approval from the University Committee on Research Involving Human Subjects (UCRIHS), and that only the names of interested participants would be released to the researcher. The second page of the mailing, written by the researcher, explained the study’s purpose, the population of interest, and times during which the experiment would be conducted. Recipients of the mailing were informed that all information would remain confidential, and that volunteers would need to obtain release time or personal leave time to participate during university business hours. Finally, a pre-addressed response sheet was provided for the recipient to indicate interest in participating. (Sample letter one is presented in Appendix B.) Response Rates As presented in Table 3.11, the first mailing, sent on June 27, 1986, achieved a 9.8% response rate. However, 65.5% of the 194 total participant volunteers did not qualify to 67 participate. The remaining 34.5% (67 volunteers) were scheduled for participation dates, and 60 actually participated. (See Table 3.11 for breakdown of participant volunteers.) On July 25, a follow—up mailing was sent to the same population to remind employees of the opportunity to participate in the study if they had not already volunteered to do so. (See Appendix C for sample letter two.) This mailing also included letters from the Assistant Provost and the researcher, and a response form. However, in an effort to enlarge the number of eligible participants, this mailing required the participant to have one or more children under 18 years old, rather than 16 years old, living at home. Table 3.11 again demonstrates that of the 1,785 employees who did not respond to the first mailing, 74 responded to the second, for a response rate of 4.2%. Of this total, 64.9% of these volunteers did not qualify to participate. The remaining 35.1% (26 volunteers) were assigned participation dates, and 20 actually participated. 0f the 80 eligible participants, three subjects turned in unacceptable questionnaires. These questionnaires were eliminated, for a total of 77 completed questionnaires. ’———-vvr~—» ~~ ~r 68 Table 3.11 Breakdown of Participant Volunteers by Mailings and Eligibility lst Mailing 2nd Mailing Total Ineligible Respondents: Single women 39 (20.0%) 9 (12.2%) 48 (17.9%) Single mothers 32 (16.5%) 9 (12.2%) 41 (15.3%) Mothers/children over 18 years old 21 (10.8%) 8 (10.8%) 29 (10.8%) Married/no children 29 (15.0%) 9 (12.2%) 38 (14.2%) Refused 3 ( 2.6%) 3 ( 4.1%) 6 ( 2.2%) Inaccess. by phone 3 ( 2.6%) 10 (13.5%) 13 ( 4.9%) Sub—totals 127 (65.5%) 48 (64.9%) 175 (65.3%) Eligible Respondents: No shows 7 ( 3.6%) 6 ( 8.1%) 13 ( 4.9%) Questionnaires w/ missing data 0 ( 0%) 3 ( 4.1%) 3 ( 1.1%) Completes 60 (30.9%) 17 (22.9%) 77 (28.7%)* Sub-totals 67 (34.5%) 26 (35.1%) 93 (34.7%) TOTALS 194 (72.4%) 74 (27.6%) 268 (100.0%) * Denotes total number of eligible participants completing acceptable questionnaires. 69 Instrumentation Experimental sessions were conducted between the dates of Tuesday, August 13, 1985 and Wednesday, September 4, 1985. On the average, two or three subjects participated in a session, with the largest session involving five subjects. All experimental sessions were held in the same room in the Communication Arts & Sciences building on MSU’s campus; the same video equipment was used for all sessions. The average session lasted 50 minutes to an hour. When participants arrived, they were asked to sign a participation roster. They were then assigned a subject number for coding and confidentiality purposes. Next, subjects were asked to read and sign a consent form which detailed the format of the session and stated that there were no physical, social, or economic risks involved. After the consent forms were collected, the questionnaires were distributed. Subjects were asked to begin the pre-test section of the questionnaire, but were reminded to not proceed past this section until they were told to do so. After the pre-test was completed, subjects were told that they would view and evaluate four videotaped commercials. The first commercial was then shown, and the subjects completed the evaluation; the second commercial was 70 then shown and evaluated by the subjects. This process continued until all four commercials had been viewed and evaluated. After the fourth commercial was viewed, subjects were told to complete the evaluation and to proceed through the remainder of the questionnaire until they had completed the last section, the demographics. Completed questionnaires were then collected, and the subjects were given a "thank you" sheet. This sheet thanked the subject for participating and requested that she not discuss the content of the experiment with others since the sessions would be continuing through the months of August and September. Summary This chapter has discussed the present study’s methodology in five sections. First discussed was the hypothesis section which explained the hypotheses in four main hypothesis sets. The first set of hypotheses were the commercial attribute hypotheses, which predict that "sharing" portrayals will be judged as containing a higher degree of task sharing, role conflict, and perceived similiarity. The first commercial attribute hypothesis, if supported, establishes the empirical premise for distinguishing between two portrayal types "sharing" and "non-sharing"--that is, subjects will have been found to rate the working mother 71 characters in "sharing" commercials as participating in more domestic and/or child care task sharing than the "non- sharing" commercials. The next set of hypotheses investigate the potential differences between "sharing" and "non-sharing" portrayals’ effectiveness. It is predicted that "sharing" portrayals, will be judged as more effective and less irritating than "non-sharing" portrayals. Next, the sales effect hypotheses suggest that after viewing "sharing" portrayals, subjects will report that they like the advertised products more and are more likely to purchase the product. Conversely, it is suggested that after viewing "non-sharing" portrayals, subjects will state that they like advertised products less and are less likely to purchase the products. The last of the hypotheses predict that after viewing either "sharing" pp‘"non—sharing" portrayals, subjects will be more likely to agree that most working women have executive or managerial level jobs. This set of hypotheses also proposes that after viewing "sharing" portrayals, subjects will be more likely to agree that working mothers experience role conflict and participate in task sharing with their husbands. After viewing "non-sharing" portrayals, subjects are expected to be less likely to agree that working mothers experience role conflict and that they task share with their husbands. Finally, it is predicted that after viewing their respective portrayals, subjects in the 72 "sharing" group will be more likely and subjects in the "non— sharing" group will be less likely to expect working fathers to task share with their working wives. Next, this chapter discussed the questionnaire design, experiment design and implementation. The questionnaire design reflects the experiment design, which includes a pre- test, four commercial exposures each followed by a commercial evaluation, concluded with a post—test and demographics section. The experiment design section also explained that the present study is exploratory and is quasi—experimental: subjects are not randomly selected and the design includes no control group. University committee and union permissions were granted in order to offer potential participants-- clerical/technical employees and supervisors--the opportunity to participate in a study of women’s attitudes toward advertisements that feature female characters. And lastly, the "Sampling" discussion stated that a total of 77 subjects were gathered from the defined population via quota sampling. CHAPTER IV ANALYSIS & RESULTS Introduction to Analyses This chapter will discuss hypothesis tests of the eleven research hypotheses. First, the sample profile will be discussed, followed by an explanation of the analyses and data treatments which preceded analysis. Lastly, the hypotheses will be discussed in four main content groups: commercial attribute hypotheses, commercial effectiveness hypotheses, sales effect hypotheses, and women’s beliefs hypotheses. The hypotheses and their statistical tests will be discussed individually in numerical order. Sample Profile Demographic data were collected and analyzed to descibe the working mother subject sample. These demographics covered two main topics, work/family~related descriptors and media usage descriptors. As discussed in the section on "Sampling," all subjects were employed, married women with one or more children under 18 years old living at home. (See Appendix D for sample profile breakdown). 73 74 Three of the demographic questions were open—ended and required coding schemes: area of residence, spouse’s occupation, and magazine readership. Before keypunching, all the open—ended responses were listed verbatim and their frequency of mention was recorded; then the responses were categorized according to content. "Area of residence" was collapsed into six categories based on geographic location: Lansing, East Lansing, Okemos/Haslett, Lansing Areas, East Lansing Areas, and Further Outskirts. "Spouse’s Occupation" contains eight categories based on skill level and job type. Those eight categories included the following: Professional, Managerial, Technical, Clerk, Skilled, Unskilled, Unemployed, and Other. Finally, magazines were categorized based on their editorial content, resulting in nine categories. Those categories included the following: Women’s, Home, Career, Parent/Child, Business/News, Black, Sports/Hobby, Reader’s Digest, and Other. Preliminary Analyses Preliminary analyses were conducted on the experimental groups to determine if the sub-samples were significantly different along any relevant descriptors. Statistical Package for the Social Sciences (SPSS) independent samples t- tests were conducted using group membership as the independent variable. Subjects in the "sharing" and "non— 75 sharing" groups were not found to be significantly different in regard to the following demographics: number of years married, number of people in household, ages of children, subject’s age, average hours worked a week by spouse, subject’s annual salary and total household income. Data Transformations Descriptive statistics conducted on all variables in the data set showed some of the variables to be positively skewed. This abnormality was predominantly found in the product liking and intent to purchase ratings, and the commercial evaluation sections, which included the commercial attribute and commercial effectiveness items. These variables’ distributions contained standard deviations that equalled or exceeded the sample mean, characteristic of a Poisson distribution.35 It is likely that this abnormality is an example of "edge effects," which can occur with bounded scales. Since the respondent was offered a response range of zero to ten, any response exceeding ten in intensity is necessarily marked by the respondent as a "ten." Thus, the upper bounds of the scale are found to represent their own values as well as the values exceeding that upper bound. This phenomenon can produce a clustering of reponses at the scale’s upper edge that positively skews the distribution. 76 Transformations were conducted on these variables in an attempt to minimize their skewness. Each of the scores was taken to the square root plus one and the descriptive statistics were re—run.36 This transformation procedure condensed the original scale, such that the new values ranged from 1.0 to 3.317, and successfully reduced the skew. Although the variables’ distributions no longer appeared Poisson, they were still positively skewed. Commercial Attribute Hypotheses The first analytical goal in this study was to examine the commercial attribute hypotheses. This goal was established because the expected difference between "sharing" and "non-sharing" portrayals is the basic premise for further analyses of between group differences. Thus the first step in analyzing the commercial attributes was to examine the commercial attribute items factor structures. To examine each commercial’s individual attribute factor structure, separate factor analyses were conducted on the attribute items for each of the eight test commercials. The individual factor outputs were analyzed in several stages. First, a "skew" test was conducted on the factor eigenvalues to determine which factor(s) contributed a sufficient proportion of explanation to error. The factor eigenvalues were graphed, and factors residing before the beginning 77 of the curve were considered acceptable. The remaining factors were considered to be in the "error zone" and were eliminated from the analysis. To interpret the factor outputs, variables were excluded from analysis under the following conditions: 1) if the variable did not load highly onto any factor; 2) if the variable loaded highly onto two or more factors; 3) if the variable loaded highly onto a factor in the "error zone" of the skew test. The remaining variables’ loadings were examined and the high—loading variables in each factor were noted,. using >.40 as a decision point. Next, these loading patterns were compared across all eight commercials to determine which attribute items consistently factored together. It was expected that items within the three scales——role conflict, task sharing, and perceived similiarity——would factor similiarly across commercials. This, however, did not occur. Table 4.1 illustrates the attribute items’ loading patterns for the eight commercials. The table presents these patterns graphically rather than numerically, which allows item clusters to be seen immediately. As the table demonstrates, the high~loading items are not consistent across all eight commercials. 78 Table 4.1 Attribute Factor Loading Comparisons ITEMS: 1 2 3 4 5 6 7 8 9 10 ll 12 l3 14 15 16 17 18 19 20 TASK SHARING FACTOR [Group 1 COMMl: COMM2: COMM3: CONDFR (Group 2 COMMl: COMM2: COMM3: COWBER ROLE CONFLICT FACTOR [Group 1 CGWHl: -l COMM2: COMM3: CGWMk [Group 2 CGHWI: COMM2: COMW3: COMM4: PERCEIVED SIM. FACTOR [Group 1 COMMl: COMM2: COMM3: CGWflk [Group 2 COMMl: COMM2: COMM3: COWBH; 6 9 11 6 9 11 p 6 9 6 7 8 9 10 11 12 —8 9 11 6 9 11 12 6 9 6 11 17 19 19 17 18 19 17 19 —13 17 19 17 19 17 19 79 These factor structures indicate that different items factored together or, different items were found to measure the "same thing," for different commercials. Some studies have avoided this problem by using test commercials that are specially—produced executions. These executions contain identical messages which are then altered so that the variable of interest is the only difference and all other aspects of the message remain the same and are thus controlled. In studies using this method, scale items were found to factor similiarly even across different commercials (Whipple and Courtney, 1980.) However, this study’s test commercials were not altered verions of one another; instead, the commercials were professionally produced commercials that were developed independently of this study. The commercials were required to be similiar in terms of the task sharing distinction but all other content in the commercials was uncontrolled, which may have contributed to the factorization differences. Scale Modifications Since factor analyses revealed differing factor structures across the eight commercials, constructing common scales such as those originally suggested was impossible. Thus an alternative to the originally—developed scales was sought that would allow use of the task sharing, role 80 conflict, and perceived similiarity measures, despite the commercials’ different factor structures. The researcher chose to examine the original scales and select one item from each that fit at least one of two following criteria: the selected item must be the item that most effectively represents that scale’s original intent, or, in the case of a scale that contains several conceptually diverse items, the selected item must be the item most relevant to the goals of the study. One item was then selected from each of the three original commercial attribute scales--task sharing, role conflict, and perceived similiarity--to represent each scale’s content. The following section will discuss the three items that were selected and will state the revised hypotheses when necessary. Hypothesis I The first hypothesis states that subjects will evaluate the "sharing" portrayals as containing more task sharing than the "non-sharing" portrayals. The following item was chosen to operationalize the task sharing dimension: This working mother’s husband spends as much time taking care of their child/children as does his working wife. Selecting this item alters the intent of the task sharing scale. Originally, the scale was constructed to measure various aspects of role sharing including child-care and household duties. However, since these items did not 81 factor consistently across commercials, the child care item was selected as most useful and relevant. This decision was based on literature and personal interviews which highlighted motherhood as a crucial factor in distinguishing working mothers’ responsibilities from other working women’s. The revised research and null hypotheses for the child— care task sharing item are as follow: H—I: Subjects will report a greater amount of child care task sharing in the "sharing" portrayals than in the "non sharing" portrayals. HO-I: Subjects will report an equal or lesser amount of child care task sharing in the "sharing" portrayals than in the "non—sharing" portrayals. Hypothesis II The next hypothesis proposes that subjects will view the working mother character in the "sharing" portrayals as experiencing more role conflict than the characters in the "non-sharing" portrayals. The following item was selected to operationalize role conflict: The working mother in this ad has trouble balancing her work responsibilities with her family responsibilities. Hypothesis III The last hypothesis for the commercial attribute section addresses the similiarity between working mother character and subject. It is expected that subjects will view 82 themselves as more similiar to the "sharing" character than the "non-sharing" character. The item selected to represent the perceived similiarity measure also refers to child care task sharing, as does the selected task sharing item. My spouse and I do not share child care responsibilities as much as the working mother and her spouse in this ad do. The revised research and null hypotheses for the perceived similiarity item are as follow: H—III: Subjects will rate the "sharing" portrayals as more similiar to their own degree of child care task sharing than the "non—sharing" portrayals. HO-III: Subjects will rate the "sharing" characters as equally or less similiar to themselves than the "non-sharing" characters. Testing the Commercial Attribute Hypotheses The SPSS independent samples t—test procedure was employed to analyze the three commercial attribute hypotheses. The results of these tests are presented in table form along with tables containing the item’s individual commercial mean scores. Task Sharing Hypothesis The first research hypothesis proposes that more child- care task sharing is conveyed by the "sharing" portrayals than by the "non—sharing" portrayals. As demonstrated by Table 4.2, this difference was found to be significant in the 83 predicted direction at greater than the .001 probability level. Table 4.3 lists the individual commercial means, which were consistently higher for the "sharing" group than for the "non-sharing" group. These data support the research hypothesis that more child care sharing is contained in the "sharing" portrayals than in the "non-sharing" portrayals. Likewise, the null hypothesis that "sharing" portrayals contain the same or less child care sharing than "non—sharing" portrayals can be rejected. Moreover, this evidence provides the basis for further between—group analyses, since the difference between the two groups’ levels of task sharing is empirically established. Role Conflict Hypothesis The second research hypothesis was also found to be significant in the predicted direction at greater than the .001 significance level. The results of this t—test are presented in Table 4.4 along with the individual commercial means in Table 4.5. These findings support the research hypothesis that the amount of role conflict experienced by working mother characters in "sharing" portrayals is judged as greater than for those characters in "non—sharing" portrayals; the null hypothesis that "non-sharing" characters experience the same or more role conflict than the "sharing" characters is not accepted. 84 Table 4.2 T—test of Task Sharing Attribute Item GROUP NUMBER MEAN STANDARD T-VALUE DEGREES l-TAIL OF CASES DEV OF FREEDOM PROB. GROUP 1 40 2.766 .379 GROUP 2 37 2.253 .498 5.12 75 .000 Table 4.3 Task Sharing Attribute Means Table by Group and Commercial Comm. #1 Comm. #2 Comm. #3 Comm. #4 GROUP MEANS Group 1: "Sharing" 2.748 2.716 2.792 2.807 2.766 Group 2: "Non—sharing" 1.877 2.601 2.545 1.990 2.253 85 Table 4.4 T-test of Role Conflict Attribute Item GROUP NUMBER MEAN STANDARD T-VALUE DEGREES l~TAIL OF CASES DEV OF FREEDOM PROB. GROUP 1 40 2.164 .398 GROUP 2 37 1.396 .338 9.10 75 .000 Table 4.5 Role Conflict Attribute Means Table by Group and Commercial Group 1: "Sharing" Group 2: "Non-sharing" Comm. #1 1.615 1.2 Comm. #2 2.081 1.344 Comm. #3 2.384 1.330 Comm. #4 2.611 1.636 GROUP MEANS 2.173 1.396 Perceived Similiarity Hypothesis The last commercial attribute hypothesis, which tests the perceived similiarity item, was not significant at the .05 probability level. The t—test results for this item are presented in Table 4.6. The perceived similiarity research hypothesis predicted that subjects would indicate that their own degree of child care task sharing was more 86 Table 4.6 T-test of Perceived Similiarity Attribute Item GROUP NUMBER MEAN STANDARD T-VALUE DEGREES l—TAIL OF CASES DEV 0F FREEDOM PROB. GROUP 1 40 1.784 .112 GROUP 2 37 1.645 .091 .95 75 .173 Table 4.7 Perceived Similiarity Attribute Means Table by Group and Commercial Group 1: "Sharing" Group 2: "Non-sharing" Comm. #1 2.771 2.938 Comm. #2 2.792 2.880 Comm. #3 2.698 2.946 Comm. #4 2.753 2.880 GROUP MEANS 2.754 2.911 similiar to that portrayed in the "sharing" portrayals than in the "non-sharing" portrayals. Although the individual commercial means presented in Table 4.7 were in the predicted direction, the difference between the two groups was not great enough to achieve signficance. Thus, the research hypothesis that subjects would judge the "sharing" characters as more similiar to their own degree of child care task sharing than the "non—sharing" characters 87 is not supported. In turn, neither can the null hypothesis, that "sharing" characters would be judged as equally or less similiar to the subjects’ own degree of child care sharing than the "sharing" portrayals, be rejected. Commercial Effectiveness Hypotheses As with the commercial attribute scales, the first analytical goal in treating the commercial effectiveness scales was to examine the scale items’ factor structures for the eight commercials. Factor analyses were conducted on the eight lcommercials’ effectiveness item scores and the outputs were analyzed using the same criteria as for the commercial attribute factor outputs. The items’ loading patterns for the eight commercials were diagrammed and compared. It was expected that the items within the two dimensions--effectiveness and irritation-— would factor similiarly across the eight commercials. This did not occur for the effectiveness items. Table 4.8 presents the items’ factorization patterns for the eight commercials. General factorization patterns can be seen across the effectiveness items that match the original scale. Table 4.8 shows that for commercials one and four in group one, the eight effectiveness items did all load together. These patterns, however, are not consistent enough to justify constructing common scales. 88 Table 4.8 Effectiveness Factor Loading Comparisons ITEMS:123456789101112131415 EFFECTIVENESS FACTOR [Group 1 COMWl: 1 CGMWZ: 1 3 g NNNN fi-b-b-b 0101mm qq oooooooo_ 3 CW: 1 3 0303030) <1 AD 010! 0000 g w H MNN WOO (D Q IRRITATION FACTOR [Group 1 CONMl: 9 10 11 12 13 COMVI2: 9 10 ll 12 13 14 15 CW3: 9 10 11 12 14 CONM4: 9 10 11 12 [Group 2 COMMl: 9 10 11 12 COWZ: 9 10 ll 12 14 COMM3: 9 10 11 12 14 COD/M4: 9 10 ll 12 13 l4 l5 89 Scale Modifications Since factor analyses revealed dissimiliar effectiveness factor structures for the eight commercials, one item was selected from the original scales. Because the effectiveness scale’s items are somewhat diverse, the selected item was chosen based on the criterion that it most effectively furthered the goals of the study. Thus, the item that was chosen to represent the effectiveness measure was selected based on its relevance to the study’s purpose. The following item was chosen to represent the effectiveness concept stated in H—IV:. Portrays a realistic situation. The revised research and null hypotheses reflecting this change are as follow: H-IV: Subjects will judge the "sharing" portrayals as more realistic than the "non—sharing" portrayals. HO-IV: Subjects will judge the "sharing" portrayals as equally or less realistic than the "non~sharing" portrayals. To be consistent with the decision to use only one of the original items to represent the effectiveness scale, the following item was selected to represent the irritation dimension: Irritating to someone like me. p: v...— 90 The research and null hypotheses remain as originally stated: H-V: Subjects will judge the "non-sharing" portrayals as more irritating than the "sharing" portrayals. HO—V: Subjects will judge the "non—sharing" portrayals as equally or less irritating than the "sharing" portrayals. Testing the Commercial Effectiveness Hypotheses The first commercial effectiveness hypothesis proposes that subjects will judge the "sharing" portrayals as more realistic than the "non—sharing" portrayals. As demonstrated by Table 4.9, this hypothesis was found to be significant at greater than the .05 probability level. Thus, the research hypothesis is supported, and the null hypothesis can be rejected. The second hypothesis to be tested in this section was the irritation hypothesis. Table 4.10 presents the t-test results that show the irritation item’s between—group difference was insignificant at the .05 probability level. In this case, the research hypothesis that subjects would judge the "non—sharing" portrayals as more irritating than the "sharing" portrayals is not supported. Likewise, the null hypothesis that subjects will rate the "sharing" portrayals as equally or more irritating than the "non- sharing" portrayals cannot be rejected. 91 Table 4.9 T-test of Effectiveness Item GROUP NUMBER MEAN STANDARD T-VALUE DEGREES l-TAIL OF CASES DEV OF FREEDOM PROB. GROUP 1 40 2.188 .463 GROUP 2 37 2.009 .468 1.68 75 .0480 Table 4.10 T—test of Irritation Item GROUP NUMBER MEAN STANDARD T-VALUE DEGREES 1-TAIL OF CASES DEV 0F FREEDOM PROB. GROUP 1 40 1.843 .595 GROUP 2 37 2.012 .669 -l.l7 75 .1225 Examining the Interaction Effects Although the realism item’s group means were found to be significantly different for the two experimental groups, both the realism and the irritation items contained individual commercial means that revealed an interaction effect between the criterion variable and group membership. This interaction effect is apparent in Tables 4.11 and 4.12. 92 Table 4.11 Realistic Item Means by Group and Commercial Group 1: "Sharing" Group 2: "Non—sharing" Comm. #1 1.947 2. Comm. #2 1.876 1.664 Comm. #3 2.189 1.513 Comm. #4 2.739 2.563 GROUP MEANS 2.188 2.009 Table 4.12 Irritation Item Means Group 1: "Sharing" Group 2: "Non-sharing" Comm. #1 2.025 1.926 Comm. #2 2.143 2.294 Comm. #3 1.764 2.216 Comm. #4 1.440 1.614 GROUP MEANS 1.843 2.012 93 In the case of the realism item, it was predicted that the "sharing" group’s commercial means would be higher while the "non-sharing" group’s means would be lower. However, Table 4.11 shows that the means for commercials one and two in the "sharing" group were unpredictedly low. Likewise, the means for commercials one and four in the "non—sharing" group were- unpredictedly high. Although the "non-sharing" group was predicted to have consistently lower realism scores than the sharing group, these four commercials’ scores deviated from the group predictions. The same pattern occurred for the irritation item, but in the opposite direction. As Table 4.12 demonstrates, the means for commercials one and two in the "sharing" group were high, although this group was predicted to have lower scores. Similiarly, the means for commercials one and four in the "non—sharing" group were low, even though this group was predicted to have higher scores. These four means demonstrate the interaction effect between "irritation" and the individual commercials. Treating the Interaction Effects As the "realism" and "irritation" mean scores were found to vary within their experimental groups, another independent variable-—other than group membership-—was sought to better explain the mean scores’ differences. Since previous studies found a relationship between portrayal appropriateness and 94 Table 4.13 Appropriateness Item Means by Group and Commercial Group 1: "Sharing" Group 2: "Non—sharing" Comm. #1 1.939 2.612 Comm. #2 2.150 2.083 Comm. #3 2.592 1.970 Comm. #4 3.035 2.571 GROUP MEANS 2.429 ‘ 2.309 subjects’ responses to commercial portrayals of women, the last measure in the commercial evaluation section-- appropriateness—~was introduced to the analysis. "Appropriateness" is operationalized as the subjects’ numerical rating of a portrayal’s appropriateness for its product category. The individual commercial means for the appropriateness variable, which are presented in Table 4.13, were examined and were found to follow a pattern matching those found in the "realism" and 'irritation" means tables. The pattern showed that, again, the "sharing" group’s means were low for commercials one and two, while the means for three and four were high. Likewise, the "appropriateness" commercial means in the "non—sharing" group were low for commercials two and three, but high for commercials one and four. 95 Further analyses were conducted using the the two commercial effectiveness items and the appropriateness measure. Correlational analysis was selected to examine the degree to which the subjects’ "realistic" and "irritating" ratings were related to the "appropriateness" variable. The correlation matrix involving these three variables is presented in Table 4.14. These data show that both the "realistic" and "irritating" items have significantly strong relationships with the appropriateness of the portrayal for the product category. More specifically, the matrix shows that "irritation" is moderately, inversely related to the "appropriateness" of the portrayal; conversely, "realism" is strongly, positively related to "appropriateness." Table 4.14 Correlation Matrix for Realism, Irritation with Appropriateness EFFECT 1 EFFECT 12 APPROP (Realism) (Irritat.) (Prod. Category) EFFECT 1 1.0000 -.5074 .6017 P=.001 P=.001 EFFECT 12 ‘l.0000 -.3355 P=.001 APPROP 1.0000 96 Next, SPSS’s NEW REGRESSION and ANCOVA procedures were used to compare the independent variables’ power as predictors of the two criterion variables-—"rea1istic" and "irritating." The regression and analysis of covariance procedures produced identical results; however, because the procedures provide different analytical perspectives, both outputs are presented in tables 4.15 and 4.16 for the reader’s reference. Testing the Realism Measure To accomplish the regression analysis for EFFECT1—-the "realistic" measure--three independent variables were specified: GROUP, which was the experimental group variable, dummy—coded; APPROP, which was the "appropriateness" measure; and a new variable, INTERl, which represented possible interaction between group membership and appropriateness. Table 4.15 summarizes the results of the regression analysis for EFFECTl, "realistic." "Appropriateness" was the first variable entered into the equation and contributed the largest explanation of variance, with 35.36% variation shared between EFFECTl and APPROP. The summary table includes ANOVA F—test results and shows that APPROP is a significant predictor of the "realistic" measure at greater than the .001 probability level. The ANCOVA table, which 97 presents a more detailed break-down of the explained and residual variance for EFFECTl, also shows the F—value of APPROP as a covariate to be significant, mirroring the F-test results in the regression table. Table 4.15 Regression and Analysis of Covariance Results with Dependent Variable "Realism" REGRESSION SUMMARY RESULTS STEP MULTR RSQ ADJRSQ F(EQU) SIGF RSQCH FCH SIGCH BETAIN CORR l-APPROP .6017 .3621 .3536 42.570 .000 .3621 42.570 .000 .6017 .6017 2—GROUP .6095 .3715 .3546 21.875 .000 .0095 1.114 .295 .0985 .1909 3-INTER1 .6230 .3882 .3630 15.439 .000 .0166 1.986 .163 —.4301 .3052 ANALYSIS OF COVARIANCE EFFECTl BY GROUP WITH APPROP SOURCE SS DF MS F SIG. OF F COVARIATES 6.104 1 6.104 42.570 .000 APPROP 6.104 1 6.104 42.570 .000 MAIN EFFECTS .160 l .160 1.114 .295 GROUP .160 1 .160 1.114 .295 EXPLAINED 6.264 2 3.132 21.875 .000 RESIDUAL 10.595 74 .143 TOTAL 16.859 76 .222 98 The next variable to enter the regression analysis was GROUP, with an R2 change value of .0095 which was not significant at the .05 probability level. The ANCOVA table also demonstrates that the proportion of variance explained by group membership fails to reach significance at the .05 level. Lastly entering the regression analysis for EFFECTl is the interaction term, with an insignificant R2 change value of .0166. Testing the Irritation Measure Next, - the regression and analysis of covariance procedures were conducted on EFFECT12--"irritation." Again, three independent variables were included in the regression procedure: GROUP, which was again dummy—coded, APPROP and INTERl. Table 4.16 includes the regression summary table for the criterion variable EFFECT12. As with EFFECTl, APPROP was the first independent variable to enter the step-wise regression equation and contributed the greatest amount of explanation, with a coefficient of determination of 11.25. Thus, 11.25% of the variance in EFFECTl was shared by variance in APPROP. The "appropriateness" variable, APPROP, showed an F-value of 9.510, which was significant at greater than the .01 probability level. 99 Table 4.16 Regression and Analysis of Covariance Results with Dependent Variable "Irritation" REGRESSION SUMMARY RESULTS STEP MULTR RSQ ADJRSQ F(EQU) SIGF RSQCH 1-APPROP .3355 .1125 .1007 9.510 .003 .1125 2—GROUP .3455 .1194 .0956 5.016 .009 .0069 3-INTER1 .3925 .1540 .1193 4.431 .006 .0347 FCH SIGCH BETAIN CORR 9.510 .576 2.990 .003 —.3355 -.3355 .450 -.0839 -.1347 .088 .6207 -.1601 ANALYSIS OF COVARIANCE EFFECTl BY GROUP WITH APPROP SOURCE SS DF MS COVARIATES 3.429 1 3.429 APPROP 3.429 1 3.429 MAIN EFFECTS .209 l .209 GROUP .209 1 .209 EXPLAINED 3.638 2 1.819 RESIDUAL 26.837 74 .363 TOTAL 30.475 76 .401 F 9.510 9.510 .576 .576 5.016 SIG. OF F .003 .003 .450 .450 .009 100 The second variable entering the regression equation was the group membership variable, which contributed a .0069 change in R2. This degree of change, however, was not significant at the .05 probability level. The ANCOVA output also shows that the proportion of variance explained by group membership was not significant. Finally, the interaction term was entered into the regression equation for EFFECT12, but did not contribute a significant amount of explanation. Table 4.16 shows that the R2 change resulting from interaction between group membership and appropriateness was .0347, which was not significant at the .05 probability level. SALES EFFECT HYPOTHESES The next hypotheses to be tested are the sales effect hypotheses. The two hypotheses—-product liking and likeliness to purchase——were designed to explore whether exposure to the "sharing" or "non-sharing" portrayals could influence the subjects’ liking of or likelihood to purchase the products featured in those commercials. 101 Product Liking Hypothesis The SPSS paired samples t—test function was used to test for differences between the advertised products’ pre—test and post-test liking scores. It was predicted that after viewing the "sharing" commercials, subjects’ liking ratings for the advertised products would increase. The t—test results for the "sharing" group, which are presented in Table 4.17, demonstrate that only commercial four showed a significant pre-test/post—test difference in the predicted direction. This difference was significant at greater than the .01 probability level. Commercial two also showed a significant difference between the two liking ratings, but this difference was not in the predicted direction. The individual means for commercial two show that the post—test liking score was lower than the pre—test score; this difference was significant at greater than .001 probability level. Thus, for the "sharing" group, the research hypothesis that the product liking scores will increase after exposure to the "sharing" portrayals is supported only for commercial four. The null hypothesis that the post—test liking scores will be the same or lower than the pre-test scores is supported for commercial two. For commercials one and three, the null hypothesis that the post—test scores will be the same or lower cannot be rejected. 102 Table 4.17 T—tests of "Sharing" Group’s Product Liking Scores VARIABLE NUMBER MEAN STANDARD T-VALUE DEGREES l-TAIL OF CASES DEV OF FREEDOM PROB. COMMERCIAL 1 Pre-test 40 2.010 .739 Post—test 40 1.922 .788 .84 39 .204 COMMERCIAL 2 Pre—test 40 1.972 .743 Post—test 40 1.621 .767 4.45 39 .000 COMMERCIAL 3 Pre-test 40 2.522 .743 Post-test 40 2.545 .714 -.40 39 .347 COMMERCIAL 4 Pre—test 40 2.975 .477 Post—test 40 3.119 .319 —3.04 39 ‘ .004 o T-tests were also used to explore the difference between the "non-sharing" commercials’ pre-test and post-test liking scores. For this experimental group it was predicted that the liking scores for the advertised products would be lower after viewing the "nonssharing" portrayals. Table 4.18 demonstrates that this relationship was true for commercial two, which showed that the post-test liking score was significantly lower than the pre-test score at greater than the .05 probability level. Commercial four also showed a significant difference between the pre—test and post—test liking scores; however, this difference was not in the 103 Table 4.18 T-tests of "Non-sharing" Group’s Product Liking Scores VARIABLE NUMBER MEAN STANDARD T-VALUE DEGREES l-TAIL OF CASES DEV OF FREEDOM PROB. COMMERCIAL l Pre—test 37 2.611 .703 Post-test 37 2.652 .664 -.69 36 .248 COMMERCIAL 2 Pre—test 37 2.162 .852 Post-test 37 1.969 .921 2.11 36 .021 COMMERCIAL 3 Pre—test 37 2.254 .706 Post-test 37 2.083 .809 1.58 36 .062 COMMERCIAL 4 Pre—test 37 1.998 .752 Post-test 37 2.173 .770 —1.73 36 .047 predicted direction. The individual means for commercial four show that the post-test liking score is actually higher than the pre-test score. This difference is significant at greater than the .05 level. Thus in the "non—sharing" group, the research hypothesis that the advertised products’ post-test liking scores will be lower than the pre—test scores is supported for commercial two. The null hypothesis that the post—test liking scores will be the same or higher is supported by commercial four. In addition, the null hypothesis cannot be rejected for commercials one and three. 104 Likeliness to Purchase Hypothesis The "sharing" group’s likeliness to purchase scores were also analyzed using paired samples t-tests. Again, it was predicted that after viewing the "sharing" commercials, the advertised products’ likeliness to purchase scores would increase. Table 4.19 presents the results of these tests and indicates that only one commercial showed a significant difference in the predicted direction. The post-test likeliness to purchase score for commercial one is significantly higher than the pre—test score at greater than the .05 probability level.‘ None of the other "sharing" commercials showed significant pre-test/post-test differences. Thus, in the "sharing" group, the research hypothesis that the likeliness to purchase scores will be higher after viewing the "sharing" portrayals is supported only for commercial one. For the remaining three commercials, the null hypothesis that the likeliness to purchase scores will be the same or lower after viewing the "sharing" commercials cannot be rejected. 105 Table 4.19 T—tests of "Sharing" Group’s Likeliness to Purchase Scores VARIABLE NUMBER MEAN STANDARD T-VALUE DEGREES l—TAIL OF CASES DEV 0F FREEDOM PROB. COMMERCIAL 1 Pre-test 40 1.335 .634 Post—test 40 1.536 .728 -2.41 39 .011 COMMERCIAL 2 Pre—test 40 1.576 .812 Post-test 40 1.527 .796 .67 39 .254 COMMERCIAL 3 Pre—test 40 2.519 .850 Post-test 40 2.447 .794 1.04 39 .153 COMMERCIAL 4 Pre—test 40 2.944 .605 Post-test 40 2.938 .613 .12 39 .452 For the "non—sharing" group, it was predicted that the post—test likeliness to purchase for the advertised products would be lower than the pre—test scores after viewing the "non—sharing" portrayals. However, as Table 4.20 demonstrates, none of the commercials’ pre—test/post-test scores were significantly different in the predicted direction. Two of the commercials’ post—test likeliness to purchase scores were significant, but in the opposite direction. The individual means for commercials one and four show that the post—test scores were higher after exposure to the 106 Table 4.20 T-tests of "Non—Sharing" Group’s Intent to Purchase Scores VARIABLE NUMBER MEAN STANDARD T-VALUE DEGREES l-TAIL OF CASES DEV OF FREEDOM PROB. COMMERCIAL 1 Pre-test 37 2.264 .947 Post—test 37 2.445 .796 —l.97 36 .029 COMMERCIAL 2 Pre—test 37 1.747 .987 Post—test 37 1.799 .960 —.49 36 .315 COMMERCIAL 3 Pre—test 37 2.049 .937 Post—test 37 1.917 .882 1.23 36 .114 COMMERCIAL 4 Pre—test 37 1.575 .837 Post-test 37 1.876 .925 -2.82 36 .004 "non-sharing" portrayals. The difference for commercial one was significant at greater than the .05 probability level, while the difference for commercial four was significant at the .01 level. In summary, the research hypothesis that likeliness to purchase scores will decrease after viewing the "non-sharing" portrayals is not supported by any of the commercials. The null hypothesis that the post-test likeliness scores will be the same or higher after viewing the "non—sharing" 107 portrayals is supported for commercials one and four. Lastly, the null hypothesis cannot be rejected for commercials two and three. Women’s Beliefs Hypothesis The last group of hypotheses under examination are the women’s beliefs hypotheses. The hypotheses are designed to explore whether exposure to "sharing" or "non-sharing" portrayals can influence subjects’ beliefs about working mothers in four belief areas: beliefs about working women’s job status, beliefs about task sharing, beliefs about role conflict, and role expectations. As stated in the "Questionnaire Development" section, four multi-item scales were developed to measure the four content areas. As was expected for the other scales, these items were expected to factor similiarly in their respective content areas and to factor similiarly to those scales after which they were modeled. This however did not occur. Table 4.21 graphically presents the factorization patterns of the belief scales for both experimental groups’ pre-test and post-test measures. Again, patterns can be seen that correspond to those expected. In factor two, the post-test measure shows that items 17 through 22——the role expectation items——did all load onto the same factor. For the task—sharing factor, items 11 and 12 loaded together and for role expectations, items 19 and 20 factored together, 108 but these items did not match the item selected for the scale after which they were modeled--attribute task sharing. Thus, it was not possible to build common scales across all measures as originally suggested to represent the various content areas. Scale Modifications Since analyses revealed differing factor structures for the four belief measures--one pre-test and one post—test for each experimental group——the same solution as previously employed in the study was applied to the women’s beliefs scales: one item was selected to represent each scale’s intent. Those items were selected based on one of two criteria: 1) if the scale was modeled after another scale in the study-—such as the task sharing belief scale, which was modeled after the commercial attribute task sharing scale—-then the same item was selected to represent the belief scale as was chosen to represent the scale after which it was modeled; 2) if the scale was not modeled after another in the study-—for example, the job status scale-—then the item that best represented the scale’s intent was chosen. Those items chosen to represent the women’s belief scales will be discussed in the following section. The items will be presented along with the revised hypotheses when necessary. Pre-test/Post-test Women’s Beliefs 109 Table 4.21 Factor Loading Comparisons ITEMS: 1 2 3 4 5 6 7 8 9 10 11 12 l3 14 15 16 17 18 19 20 21 22 TASK SHARING FACTOR [Pre—test GROUP 1: GROUP 2: 10 [Post—test GROUP 1: GROUP 2 F" 0 ROLE EXPECTATION FACTOR (Pre—test GROUP 1: GROUP 2: [Post-test GROUP 1: GROUP 2: JOB STATUS & ROLE CONFLICT FACTORS [Pre—test GROUP 1: l 2 5 6 GROUP 2: 5 6 (Post—test GROUP 1: 5 6 7 GROUP 2: l 2 11 12 ll 12 13 ll 12 11 12 13 —14 15 —16 -l8 19 20 22 19 20 -21 22 17 —18 19 20 -21 22 17 19 20 22 110 The following item was selected from the women’s beliefs section to operationalize the job status variable. This item was selected intent. on the basis that it best represents the scale’s Most working women have executive or managerial jobs. The revised research and null hypotheses are as follow: H-VIII: HO-VIII: After viewing the four commercials, subjects in both experimental groups will be more likely to agree that most working women have executive or managerial jobs. After viewing the four commercials, subjects in both experimental groups will be more equally or less likely to agree that most working women have executive or managerial jobs. The item selected to operationalize the task sharing beliefs variable was chosen on the basis that it corresponds to the selected commercial attribute task sharing item: Working fathers spend as much time with their children as do working mothers. The revised research and null hypotheses also reflect the emphasis on the child care aspect of task sharing: H-IX: After viewing the "sharing" portrayals, subjects will be more likely to agree that working fathers share child-caring with their working wives; after viewing the "non—sharing" commercials, subjects will be less likely to agree that working fathers share child—caring with their working wives. 111 HO-IX: After viewing the "sharing" portrayals, subjects will be equally or less likely to agree that working husbands and wives share child-caring; after viewing the "non-sharing" commercials, subjects will be equally or more likely to agree that working husbands and wives share child care. Likewise, the item selected to operationalize the role conflict beliefs item corresponds to the item chosen for commercial attribute role conflict measure. Working mothers easily balance their job responsibilities with their family responsibilities. For this item, the research and null hypotheses are identical to the commercial attribute role conflict item, except that the hypotheses are reversed: H-X: After viewing the "sharing" commercials, subjects will be less likely to agree that working mothers do not experience role conflict; after viewing the "non—sharing" commercials, subjects will be more likely to agree that working mothers do not experience role conflict. HO—X: After viewing the "sharing" commercials, subjects will be equally or more likely to agree that working mothers do not experience role conflict; after viewing the "non—sharing" commercials, subjects will be equally or less likely to agree that working mothers do not experience role conflict. Lastly, the role expectation variable was operation— alized with the item that corresponds to other items which emphasized the child care aspect of task sharing: A working father should not be expected to spend as much time with his kids as does his working wife. r , “p , 112 The revised research and null hypotheses also reflect this emphasis: H—XI: After viewing the "sharing" portrayals, subjects will be less likely to agree that a working father should not be expected to child care as much as his working wife; after viewing the "non—sharing" portrayals, subjects will be more likely to agree that a working father should not be expected to child care as much as his working wife. HO-XI: After viewing the "sharing" portrayals, subjects will be equally or more likely to agree that a working father should not be expected to share child care as much as his working wife; after viewing the "non—sharing" portrayals, subjects will be equally or less likely to agree that a working father should be expected to share child care as much as his working wife. Testing the Women’s Beliefs Hypotheses SPSS paired samples t—tests were employed to examine the four women’s belief hypotheses for pre-test/post—test differences. The results of these tests are presented in the following sub—sections. Job Status Belief Hypothesis "Job status" was the first variable to be examined. It was predicted that after viewing the four commercials, subjects in both experimental groups would be more likely to agree that working mothers have executive or managerial jobs. Table 4.22 presents the t~test results for this hypothesis and shows that although the post—test means for both groups 113 Table 4.22 T-tests of Job Status Belief Item GROUP/ NUMBER MEAN STANDARD T—VALUE DEGREES 1-TAIL OF CASES DEV OF FREEDOM PROB. "SHARING"/ Pre-test 40 1.416 .550 Post-test 40 1.487 .543 —l.45 39 .078 "NON—SHARING"/ Pre-test 37 1.339 .428 Post—test 37 1.343 .417 - .05 36 .479 were greater than their respective pre-test means, this difference was not large enough to reach significance. Hence the research hypothesis——that subjects in the "sharing" group will be more likely to agree and subjects in the "non—sharing" group will be less likely to agree that working mothers have executive or managerial jobs——is not supported. In turn, the null hypotheses cannot be rejected. Task Sharing Belief Hypothesis "Task sharing" was the next variable to be tested for pre—test/post—test differences. It was predicted that after viewing the four commercials, subjects in the "sharing" group would be more likely to agree and subjects in the "non— sharing" group would be less likely to agree that working mothers share child care tasks with their spouses. 114 Table 4.23 T—tests of Task Sharing Belief Item GROUP/ NUMBER MEAN STANDARD T-VALUE DEGREES l-TAIL 0F CASES DEV 0F FREEDOM PROB. "SHARING"/ Pre-test 40 1.634 .604 Post-test 40 1.739 .557 -l.33 39 .095 "NON—SHARING"/ Pre—test 37 1.625 .585 Post-test 37 1.712 .614 — .72 36 .239 Table 4.23 shows that no support was found for either part of this hypothesis; the t—values were insignificant for both experimental groups. Thus, the null hypothesis-- that subjects in the "sharing" group will be equally or less likely and subjects in the "non-sharing" group will be equally or more likely to agree that working mothers share child care responsibilities with their spouses——cannot be rejected. Role Conflict Belief Hypothesis For the "role conflict" variable, it was predicted that after viewing the four commercials, subjects in the "sharing" group would be less likely to agree that working mothers do not experience role conflict. It was also predicted that 115 Table 4.24 T-tests of Role Conflict Belief Item GROUP/ NUMBER MEAN STANDARD T—VALUE DEGREES l—TAIL OF CASES DEV OF FREEDOM PROB. "SHARING"/ Pre—test 40 1.817 .750 Post—test 40 1.832 .593 — .20 39 .420 "NON-SHARING"/ Pre-test 37 1.683 .731 Post—test 37 1.839 .661 —l.83 36 .038 after viewing the four commercials, subjects in the "non- sharing" group would be more likely to agree that working mothers do not experience role conflict. Table 4.24 reveals that this prediction was not supported for the "sharing" group, but was supported for the "non-sharing" group. In the "non-sharing" group, the pre- test/post-test scores’ difference was signficant at greater than the .05 probability level. Thus, the "non-sharing" research hypothesis—~that subjects in the "non-sharing" group would be more likely to agree that working mothers do not experience role conflict-~15 supported. However, no support was found for the "sharing" research hypotheses. Thus the null hypothesis——that subjects in the "sharing" group will be equally or more likely to agree that working mothers do not experience role conflict——cannot be rejected. 116 Table 4.25 Tstests of Role Expectation Item GROUP/ NUMBER MEAN STANDARD T-VALUE DEGREES l-TAIL 0F CASES DEV 0F FREEDOM PROB. "SHARING"/ Pre-test 40 1.122 .305 Post-test 40 1.346 .520 —2.67 39 .006 "NON-SHARING"/ Pre-test 37 1.135 .450 Post—test 37 1.256 .491 —1.16 36 .126 Role Expectations Hypothesis The final research hypothesis in the women’s beliefs section predicted that after viewing the four commercials, subjects in the "sharing" group would be less likely to agree that a working father should not be expected to child care as much as his working wife; conversely, it was predicted that after viewing the four commercials, subjects in the "non—sharing" group would more likely agree that a working father should not be expected to child—care as much as his working wife. Table 4.25 shows that for the "sharing" group, the pre— test/post-test difference was significant at greater than the .01 significance level, but in the opposite direction predicted. Thus, no support is found for the research hypothesis that after viewing the four commercials, subjects in the "sharing" group would be less likely to agree that a 117 working father should not be expected to share in child- caring. Likewise, the null hypothesis cannot be rejected. Table 4.25 also shows that in the "non—sharing" group, no support was found for the research hypothesis that after viewing the four commercials, subjects would be more likely to agree that working fathers should not be expected to child care as much as their working wives. The null hypothesis-- that subjects in the "non—sharing" group would be equally or less likely to agree that working fathers should not be expected to child care as much—-cannot be rejected. Analysis Summary This chapter presented four main sets of hypotheses and discussed the results of each hypothesis test. The first group of hypotheses to be tested were the commercial attribute hypotheses. The first of these hypotheses predicted that subjects in the "sharing" group would report a greater amount of task sharing than subjects in the "non- sharing" group would report. This hypothesis was revised, limiting the operational definition of "task sharing" to "child care task sharing." T—test results showed the research hypothesis was significant in the predicted direction. These findings provide empirical support for the researcher’s distinction between "sharing" and "non-sharing" portrayals. 118 The second hypothesis proposed that working mother characters in "sharing" portrayals would be judged as experiencing more role conflict that the characters in "non— sharing" commercials. This hypothesis was also found to be significant in the predicted direction. The last commercial attribute hypothesis suggested that the amount of child care task sharing in the "sharing" portrayals would be evaluated as more similiar to the subjects’ own level of child care sharing than that demonstrated in the "non—sharing" portrayals. This difference was not found to be significant, and the commercial means indicated the opposite relationship. The next hypotheses to be examined were the commercial effectiveness hypotheses. The first hypothesis in this section proposed that "sharing" portrayals would be judged as more effective than "non—sharing" portrayals. This hypothesis was revised so that a single item—-"realism"--was used to represent the "effectiveness" concept. Thus, the hypothesis as a whole was revised to predict that "sharing" portrayals would be judged as more realistic than the "non— sharing" portrayals. This difference was found to be significant in the predicted direction. The second commercial effectiveness hypothesis predicted that "non—sharing" portrayals would be rated by subjects as more irritating than the "sharing" portrayals. This scale 119 was also reduced to one representative item, but did not necessitate that the hypothesis be revised. In testing this hypothesis, the means for "sharing" and "non-sharing" groups’ realism scores were found to be significantly different. However, the individual commercials’ means for both the "realistic" and the ”irritating" ratings were found to contain interaction between the group membership and both criterion variables. Another variable——"appropriateness"-- was introduced to the analysis in an attempt to better explain the within—group variations. Regression analyses and analyses of covariance revealed that "appropriateness" was a significant predictor of both "realism" and "irritation," while group membership was an insignificant predictor. Next, the sales effect hypotheses were examined. The first hypothesis—-product liking-—predicted that after viewing the "sharing" commercials, subjects would rate the advertised products more favorably. This hypothesis was significant in the predicted direction for commercial four; however, the liking rating for commercial two was significantly lower after viewing the four commercials, which was opposite the predicted direction. The pre—test/post—test differences for commercials one and three were insignificant in either direction. This same hypothesis also predicted that after viewing the "non—sharing" portrayals, subjects would rate the advertised products more favorably. This hypothesis was 120 significant for commercial two in the predicted direction, but was significant in the opposite direction for commercial four. The remaining commercials’ pre-test/post—test differences were insignificant. The second sales effect hypothesis--likeliness to purchase-—predicted that after viewing the "sharing" commercials, subjects would be more likely to purchase the advertised products. This predicted difference was only significant for commercial one. Also included in this hypothesis was the prediction that after viewing the "non- sharing" portrayals, subjects would be more likely to purchase the advertised products. This difference was insignificant for all the commercials. The only pre- test/post-test differences which were significant were in the opposite direction than predicted, for commercials one and four. The final group of hypotheses to be discussed were the women’s beliefs hypotheses. These hypotheses suggested that viewing "sharing" and "non—sharing" portrayals of working mothers may influence subjects’ beliefs about working mothers in four areas: job status, beliefs about task sharing, beliefs about role conflict, and role expectations. The only belief item which showed a pre—test/post—test difference in the predicted direction was the role conflict hypothesis for the "non—sharing" group. T-test results showed that subjects in believe that of the other 121 after viewing the "non-sharing" portrayals, the "non—sharing" group were less likely to working mothers experience role conflict. None pre—test/post-test differences were signficant. CHAPTER V CONCLUSION Introduction This chapter provides a discussion of the present study’s limitations, conclusions, and recommendations. Limitations will be presented first, followed by the conclusion asection. Next presented are recommendations for two groups: advertisers targeting married working mothers and researchers examining consumer reactions to multiple-role portrayals of married working mothers. This section is followed by the chapter summary. Limitations One limitation of this study involves the subject sample drawn from the population of female Michigan State University employees. Because the sample is unrepresentative geographically, results are not generalizable to all types of working women that an advertiser may be trying to reach. However, because the sample excluded professional and academic employees, the sample does represent the majority of working women in terms of job status and annual salary. 122 123 The conclusions based on this sample may not be useful to advertisers specifically targeting professional or managerial women; however, this sample is applicable for advertisers who solely or additionally target non-professionals and non- managers. A second limitation of the study’s sample is that the subjects were gathered non-randomly on a quota basis. Because participation in the study required an hour to an hour and a half of the the participant’s time, it is possible that those individuals who were interested enough to participate are different from those who were not interested. Because potential participants were told the session involved a study of women’s attitudes toward advertising that features female characters, it is possible that the volunteers would be more interested than non—volunteers in issues relating to women. However, this aspect of representativeness may not be crucial since previous studies found feminism to insignificantly predict women’s judgements of working women portrayals. Other limitations involve the experiment’s commercial exposures. All the commercials were professionally produced, on-air commercials. These commercials provided the advantage of being equal in terms of professional production; however, the on—air nature of the executions means that subjects may have viewed the commercials previously. 124 Additionally, the experiment did not utilize a rotated design. Thus, order and testing effects cannot be controlled. However, while the commercial evaluation scores were analyzed for between—group differences, they were not compared individually between groups--for example, comparing the first commercial in group one to the last commercial in group two. Therefore, one can assume that any order or testing effects would be the same for both experimental groups. The external validity of this study, as with most laboratory experiments, is also limited. The subjects viewed four commercials which they evaluated, and completed pre-test and post-test measures. However, viewing and evaluating the individual commercials is not highly externally valid, since viewing conditions can only be simulated. Subjects are apt to pay more attention to the message in the lab than during normal viewing: the commercials were not imbedded in program content and environmental distractions were eliminated. In terms of the sales effect and belief measures’ validity, it is not possible to say that the effects of consecutively viewing and evaluating four same-portrayal messages are equal to the cummulative effects of viewing many varied-portrayal messages in the natural viewing environment. However, relationships can be suggested as long as the limitations are considered. 125 Conclusions The first major conclusion of this study is that subjects did evaluate the "sharing" and "non-sharing" portrayals of working mothers as significantly different along two attributes: task sharing and role conflict. As discussed in the introduction chapter, multiple role portrayals that feature task sharing between wives and husbands are a variation of previous portrayals which showed the working mother singulary performing multiple roles. This study’s data indicate that viewers do see the "sharing" characters as participating in more task sharing than the "non-sharing" characters. In addition, the "sharing" working mother character was seen as experiencing significantly more role conflict than the "non—sharing" character. Aforementioned sociological literature stated the majority of married working mothers do experience role conflict and they do participate in role- sharing. Hence portrayals which feature characters who experience a greater degree of role conflict and who participate in role sharing more closely approximate married working mothers’ actual multiple*role experiences. Moreover, these portrayals reflect the changing social structure of role responsibilities for working mothers and their spouses, 126 and demonstrate that many dual-earner couples are sharing previously "female" domestic roles to reduce interrole conflict. Based on these intitial findings, one would predict that "sharing" portrayals would be judged as more similiar to working mother subjects’ own experiences with task sharing. However, the between-group tests indicate that one cannot conclude the "sharing" portrayals were in fact more similiar. Working mother subjects did judge the "sharing" portrayals as more similiar to their own degree of task sharing than the "non-sharing" portrayals, but the difference was not large enough to achieve significance. One explanation is that the insignificant between-group difference is more a function of the perceived similiarity item’s wording than it is a measure of similiarity between commercial character and viewer. Some subjects expressed that although they could judge the extent of role interactions actually exhibited in the commercial, they did not feel they could generalize from those isolated behaviors to an overall statement about the advertising characters’ frequency of such behaviors. For example, in the case of the perceived similiarity measure, respondents could judge whether the characters participated in task sharing during the commercial’s 60 seconds, but were not able to generalize from that observation to a statement about "how much" those characters task share. Moreover, the perceived similiarity 127 item asked the subjects to compare "how much" they task share to "how much" the commercial characters task share, a comparison for which they had no basis. This explanation suggests that the item may have failed to measure what it was intended to measure and, instead, revealed that subjects cannot make inferences about portrayed characters from one "slice" of the characters’ lives, regardless of the portrayal type. The study’s next major conclusion concerns the commercial effectiveness measures. These data show that subjects rated "sharing" portrayals as significantly more realistic than "non-sharing" portrayals. In light of the study’s previous findings, this relationship implies that perhaps working mothers view the "sharing" portrayals as more real because the messages portray a working mother character who is less phenomenal, who experiences conflict between her roles but manages them realistically by task sharing with her spouse. One might also conclude that portraying a working mother character who does not "do it all" is a more realistic strategy than portraying a character who accomplishes all roles with ease. Surprisingly, results did not indicate that the portrayal types were different in terms of the second effectiveness measure——irritation. The data showed that neither portrayal type was significantly more irritating than the other; furthermore, an interaction effect was found to 128 exist between appropriateness and both criterion variables, realism and irritation. Further analyses revealed that portrayal appropriateness for the product category was a significant predictor of both the "realistic" and "irritation" scores. These results indicate that when controlling for appropriateness, the presence or absence of task—sharing is an insignificant predictor of commercial effectiveness—-for realism gpd irritation. These results negate the previous findings of significantly different realism scores: the relationship between group membership ("sharing" versus "non—sharing") and realism appears to be spurious, since the relationship becomes insignificant when controlling for appropriateness. One can conclude that a multiple—role portrayal’s realism or irritation value is most strongly related to the appropriateness of that portrayal for the given product category. The next conclusion offered concerns the sales effect measures. One cannot generally conclude that viewing "sharing" portrayals increased liking for the advertised products nor that viewing "non-sharing" portrayals decreased liking for the advertised products. The same conclusive weakness applies to the likeliness to purchase measure. For both liking and likelihood to purchase, the advertised products’ pre—test/post-test t-tests yielded varied results that did not match the predicted group differences. 129 Two main explanations for these results can be offered. First, one could cite the commercial exposure artificiality: four artificially-viewed, same-message portrayals may not be sufficient to affect a subject’s att itude or behavioral intent toward the product, although exposure to portrayals such as these in the natural environment may, over time, have such effects. A second explanation is that previous exposures to the commercials may affect subjects’ responses. Research has found that for some low— brand familiarity alone may increase 1 involvement products iking for the brand (Zajonc, 1980). Several of the present study’s advertised products are low-involvement products. Thus, if a subject had not been previously exposed to commercials for these products, one exposure might increase brand familiarity enough to affect brand liking or, even, intent to purchase-— regardless of the portrayal type. Conversely, for those commercials that subjects had viewed previously, one additional exposure may not affect the attitudes they expressed in the pre—test. The last measure, women’s beliefs, results. Neither the "sharing" nor showed pre-test/post—test differences belief item. Both groups showed that status expectation was higher after commercials, but the difference was not also produced varied "non—sharing" groups for the job status the subjects’ job viewing the four significant. These 130 results suggest that perhaps continuous exposure to executive or managerial portrayals of working women in the natural viewing environment would increase women’s beliefs about the status of other working women; this study, however, does not substantiate that conclusion. One must also consider that the experiment did not control for previous exposures to executive or managerial portrayals of working women. It is possible that previous media portrayals of women in professional positions had already raised subjects’ job status beliefs. The pre-test data, however, negate this explanation by showing that the subject sample as a whole did not agree that the majority of working women have executive or managerial jobs. Results for the next two women’s belief measures, task sharing and role conflict beliefs, are also varied. Neither the "sharing" nor the "non—sharing" group showed a significant difference in pre—test/post-test task sharing scores. Likewise, the "sharing" group did not show significantly different pre—test/post—test belief scores for the role conflict variable. However, in the "non-sharing" group, the post-test role conflict belief score was significantly higher than this group’s pre-test score, which was the predicted relationship. As a whole, the sample showed pre~test role conflict belief scores that indicated they did not believe working mothers easily balance multiple roles. However, these data showed that after viewing the 131 "non—sharing" portrayals, subjects were more likely to believe that working mothers easily balance work and family responsibilities. One explanation for these results is that the "sharing" portrayals merely confirmed subjects’ pre-existing beliefs that working mothers d9 experience role conflict and, thus, did not alter their beliefs significantly. However, the subjects in the "non—sharing" group viewed portrayals which featured working mothers performing multiple roles effortlessly and with no assistance. Thus the "non-sharing" messages contradicted rather than confirmed the subjects’ original belief, and shifted the subjects toward believing working mothers do not experience interrole conflict. Hence one can conclude that subjects who viewed the belief-congruent "sharing" messages did not report significantly different post—test beliefs. However, after subjects viewed the "non-sharing" messages, which contradicted their original beliefs, they were less likely to believe that working mothers experience role conflict. One can conclude the following: portrayals that feature characters who experience role conflict match working mother subjects’ belief that working mothers do experience role conflict. Conversely, one can conclude that after working mothers view commercial characters performing multiple roles with less role confict, subjects may be more likely to agree that working mothers do not experience role conflict. The final measure, role expectations, also produced varied results. For subjects viewing "non—sharing" portrayals, no significant pre—test/post-test difference was observed. However in the "sharing" group, the post—test expectation score was lower than the pre-test score, a difference which was significant but not in the predicted direction. These results lead to the conclusion that viewing either "sharing" or "non—sharing" portrayals does not significantly increase subjects’ belief that working fathers should spend as much time with their children as their working wives do. More specifically, the results indicate that after viewing portrayals that feature a working mother who task shares with her spouse, subjects’ will be significantly less likely to expect that working fathers spend as much time with their children as do their working wives. This latter result opposes the predicted relationship. One explanation relates to the task sharing expectation item’s wording. The subject sample as a whole stated in the pre-test that they agreed that fathers should spend time with their kids. However, the item states, "working fathers should not be expected to child care as much as their working wives." Perhaps if the item merely stated that working fathers should task share with their wives——rather 133 than stating that they should spend as much time as their wives--then the results might have more closely reflected the predicted differences. RECOMMENDATIONS Future Research The recommendations for future research on commercial portrayals of married working mothers encompass two main areas. The first recommendations concern alternative sampling and research designs. It is suggested that future studies consider alternative populations from which to draw samples. This study chose to involve only non—professional and non-managerial working women. However, advertisers of high-cost, high—involvement products often target markets characterized by higher socioeconomic status and greater buying power. Thus advertisers of these products may want to consider exclusively examining professional and managerial women’s preferences for multiple role portrayal types. Additionally, researchers may want to investigate the effectiveness of multiple—role portrayals for markets other than married working mothers. Because working fathers are also portrayed in commercials that fit the "sharing" distinction, this group may be found to express preferences for various portrayal types that feature multiple—role fathers. 134 Regarding research design, this study supports the recommendation that future research attempt to control for previous exposure to commercial portrayals of dual-earner characters. Researchers can control for this variable by using professionally—produced commercials that have not been aired in the\ area from which the sample is drawn. Researchers can also use specially-produced test commercials. This strategy allows for developing identical commercials that contain only role portrayal. variations and thus eliminate content differences between stimuli. Previous exposure can also be defined as a variable and measured. This strategy would involve asking the subjects to report the number of times s/he remembers seeing the commercial previously. However, one must be aware that the measure lacks content validity; the subjects’ ability to recognize a previously—seen commercial may not accurately represent the degree to which they were affected by previous exposures. This study next recommends strategies for developing evaluative measures. Analyses revealed that portrayals’ appropriateness for the product category best predict the commercial realism and irritation scores. Thus it is suggested that future researchers control for product categories. Future studies that utilize same-category products will be able to more thoroughly investigate consumer preferences for commercial attributes—~such as role conflict 135 and task sharing-—and will be able to more conclusively examine the relative effectiveness of different portrayal types. Another recommendation concerns the use of multi-item scales to measure the attributes and effectiveness of working mother commercial portrayals. This recommendation suggests that extensive pre-tests be conducted to clarify the items’ interpretability. More importantly, in the case of content- varied commercials as were used in the present study, pre- testing is advised to identify which items will apply to all commercials. It is suggested that items requiring the subject to make evaluative inferences about the portrayed characters and their situations—-such as this study’s perceived similiarity measure-—be specifically pre—tested for applicability. Recommendations for Advertisers The conclusions stated in this manuscript have several implications for advertisers targeting the married working mother market. Consumers and industry members have criciticized some multiple—role portrayals as being unrealistic and exaggerated. Sociologists have stated that the majority of working mothers experience conflict between their multiple role responsibilities and that many task share with their husbands to alleviate this stress. 136 Although this study found two categories of portrayal types to be significantly different in terms of task sharing and role conflict,~ the study’s findings do not suggest that one portrayal type is more effective or less irritating than the other. Furthermore, this study found no indication that either portrayal type has the potential to influence brand liking or intent to purchase. Results did indicate, hOWever, that subjects’ judgement of commercials’ realism and irritation value were related to appropriateness of the portrayal type for the product category. The findings suggest that manufacturers of products that have specific value to dual—earner families may want to consider targeting these individuals with multiple- role appeals. Time—saving kitchen products and household products, day care services and home—cleaning services, "escapist" oriented industries such as travel and air line services-—all are industries in which multiple—role messages may help advertisers appeal to dual-earner couples’ specific and targetable needs. However, aside from product—appropriate portrayals, advertisers have no empirical incentive to choose the "sharing" type portrayals over "non-sharing" portrayals. Nonetheless, advertisers may want to consider that many multiple—role women experience role conflict and that many task share with their husbands. It is also worth acknowledging that as both mothers and fathers continue 137 working outside the home, the trend toward role diversification will most likely continue. Thus advertisers who are concerned with constructing "slice of life" commercials that appeal to the dual-earner market may want to consider portraying characters who reflect the experiences of real-life individuals. SUMMARY This chapter presented the limitations, conclusions, and recommendations for a study of married working mothers’ reactions to commercial portrayals featuring married working mothers. Three main conclusions are evident. First, these data support the conclusion that married working mother subjects view "sharing" commercials as portraying significantly more task sharing and role conflict than "non— sharing" commercials. Thus one can conclude that subjects do view the two portrayal types differently. Analyses comparing the two portrayal types’ levels of effectiveness initially found that "sharing" portrayals were rated as significantly more realistic than the "non~sharing" portrayals. These findings suggest that by portraying characters who task share and who are judged to experience role conflict, the "sharing" portrayals are reflecting the "realities" of many multiple— role couples’ lives. However, these data also support the conclusion that the relationship between portrayal type and 138 realism is spurious: group membership was an insignificant predictor when controlling for portrayal appropriateness. The same conclusion applies to the irritation measure, as group membership was an insignificant predictor after controlling for appropriateness. The sales effect measures produced varied results and were generally inconclusive. One cannot conclude that viewing "sharing" or "non-sharing" commercials significantly affects product liking or likeliness to purchase. The women’s belief results also varied. Of the four belief measures, the only significant pre—test/post—test difference was with the "non—sharing" group’s role conflict belief scores. These results support the following conclusion: viewing "non—sharing" portrayals have the potential to make working mother subjects more likely to agree that married working mothers do ppt experience role conflict. In summary, the study revealed that portrayals of married working mothers can be categorized according to the commercials’ portrayal of interrole conflict and sharing between spouses. One cannot conclude based on this study that one portrayal type is more effective or more likely to improve subjects’ attitudes toward the product. The study did reveal, however, that portrayal appropriateness is an important factor in identifying more realistic and less irritating portrayals. APPENDICES APPEND IX A 139 Participant I.D. # Women’s Attitudes Toward Advertising'Studyh-1985 A. We’re interested to know how much you like the following products. Based on what you know about these products, please evaluate each product using a scale ranging from 0 to 10, where zero means you do not like the product at all, and ten means you like the product totally. Circle the appropriate number on the scale to the right. Not at all Totally 1. Sunrise dishwashing detergent ' 0 l 2 3 4 5 6 7 8 10 2. TEST PRODUCT 0 l 2 3 4 5 6 7 8 9 10 3. Bounce fabric softener 0 l 2 3 4 5 6 7 8 9 10 4. Spray-n-Wash 0 1 2 3 4 5 6 7 8 9 10 5. Del Monte catsup 0 l 2 3 4 5 6 7 8 9 10 6. Lipton soup 0 l 2 3 4 5 6 7 8 9 10 7. TEST PRODUCT 0 1 2 3 4 5 6 7 8 9 10 8. Velveeta cheese 0 l 2 3 4 5 6 7 8 9 10 9. Pepsi—cola 0 l 2 3 4 5 6 7 8 9 10 10. Emeraude cologne 0 1 2 3 4 5 6 7 8 9 10 11. Jean Nate body splash 0 1 2 3 4 5 6 7 8 9 10 12. TEST PRODUCT 0 l 2 3 4 5 6 7 8 9 10 13. Raintree body lotion 0 l 2 3 4 5 6 7 8 9 10 14. L’eggs sheer energy pantyhose 0 1 2 3 4 5 6 7 8 9 10 15. Sealy Posturpedic mattress 0 l 2 3 4 5 6 7 8 9 10 16. TEST PRODUCT 0 l 2 3 4 5 6 7 8 9 10 17. TEST PRODUCT 0 1 2 3 4 5 6 7 8 9 10 18. TEST PRODUCT 0 l 2 3 4 5 6 7 8 9 10 19. Ford Mustang 0 l 2 3 4 5 6 7 8 9 10 20. Commodore computer 0 l 2 3 4 5 6 7 8 9 10 140 Not at all Totally 21. Lincoln-Mercury Cougar XR7 0 1 2 3 4 5 6 7 8 9 10 22. Cedar Point amusement park 0 l 2 3 4 5 6 7 8 9 10 23. TEST PRODUCT 0 l 2 3 4 5 6 7 8 9 10 24. Detroit zoo 0 1 2 3 4 5 6 7 8 9 10 B. Next, we’re interested to know how likely you would be to purchase the following products. Based on what you know about these products, please evaluate each product using a scale ranging from 0 to 10, where zero means you are not at all likely to purchase the product, and ten means you are totally likely to purchase the product. Circle the appropriate number on the scale to the right. Not at all Totally 1. Sunrise dishwashing detergent 0 l 2 3 4 5 6 7 8 10 2. TEST PRODUCT 0 l 2 3 4 5 6 7 8 9 10 3. Bounce fabric softener 0 1 2 3 4 5 6 7 8 9 10 4 Spray-n—Wash 0 l 2 3 4 5 6 8 9 10 5. Del Monte catsup 0 l 2 3 4 5 6 7 8 9 10 6 Lipton soup 0 1 2 3 4 5 6 7 8 9 10 7. TEST PRODUCT 0 1 2 3 4 5 6 7 8 9 10 8. Velveeta cheese 0 1 2 3 4 5 6 7 8 9 10 9. Pepsi—cola 0 l 2 3 4 5 6 7 8 9 10 ‘10. Emeraude cologne 0 l 2 3 4 5 6 7 8 9 10 11. Jean Nate body splash 0 l 2 3 4 5 6 7 8 9 10 12. TEST PRODUCT 0 l 2 3 4 5 6 7 8 9 10 13. Raintree body lotion 0 l 2 3 4 5 6 7 8 9 10 14. L’eggs sheer energy pantyhose 0 l 2 3 4 5 6 7 8 9 10 15. Sealy Posturpedic mattress 0 l 2 3 4 5 6 7 8 9 10 16. TEST PRODUCT 0 1 2 3 4 5 6 7 8 9 10 17. TEST PRODUCT 0 1 2 3 4 5 6 7 8 9 10 18. TEST PRODUCT 0 1 2 3 4 5 6 7 8 9 10 141 Not at all Totally 19. Ford Mustang 0 l 2 3 4 5 6 7 8 9 10 20. Commodore computer 0 1 2 3 4 5 6 7 8 9 10 21. Lincoln-Mercury Cougar XR7 0 1 2 3 4 5 6 7 8 9 10 22. Cedar Point amusement park 0 1 2 3 4 5 6 7 8 9 10 23. TEST PRODUCT 0 l 2 3 4 5 6 7 8 9 10 24. Detroit zoo 0 l 2 3 4 5 6 7 8 9 10 C. Next, we’d like to know what some of your beliefs are concerning women’s roles. Please indicate the extent to which you agree with the following statements using a scale of zero to ten, where zero means you do not agree at all and ten means you agree totally. Not at all Totally 1. Most working women have 0 l 2 3 4 5 6 7 8 9 10 executive or managerial jobs. 2. Most working mothers have 0 l 2 3 4 5 6 7 8 9 10 executive or managerial jobs. 3. Few working women make more 0 1 2 3 4 5 6 7 8 9 10 than $13,000 a year. 4. Few working mothers make more 0 1 2 3 4 5 6 7 8 9 10 than $13,000 a year. 5. Most working women make more 0 1 2 3 4 5 6 7 8 9 10 than $20,000 a year. 6. Most working mothers make more 0 1 2 3 4 5 6 7 8 9 10 than $20,000 a year. 7. Working mothers easily balance 0 l 2 3 4 5 6 7 8 9 10 their job responsibilities with their family responsibilities. 8. Working mothers are able to 0 l 2 3 4 5 6 7 8 9 10 spend the desired amount of time at work and also spend enough time with their child/children. 9. Working mothers do not find 0 1 2 3 4 5 6 7 8 9 10 having a job outside the home stressful. 10. ll. 12. l3. l4. 15. 16. 17. 18. 19. 20. 142 Not at all Working fathers spend as much time with their children as do working mothers. Working women receive help from their spouse when it comes to household chores such as dish— washing or clothes—washing. Working fathers help their working wives when it comes to fixing family meals. Working fathers spend as much time "taxi—ing" their kids as do working mothers. Working fathers do not spend as much time attending their kids’ activities as do working mothers. Working fathers spend as much time helping their child or children with homework as does the working mother. Working fathers do not grocery shop for their families as often as do working mothers. A working father should not be expected to spend as much time with his kids as does his working wife. A working father should be expected to spend as much time helping his children with homework as does his working wife. A working father should not be expected to spend as much time dish—washing and clothes—washing as does his working wife. A working father should not be expected to be responsible for preparing family meals. 0 l 0 l 0 1 0 1 0 l 0 1 0 l 0 l 0 l 0 1 0 l 2 Totally 9 10 9 10 9 10 9 10 9 10 9 10 9 10 9 10 9 10 9 10 9 10 143 Not at all Totally 21. A working father should help 0 l 2 3 4 5 6 7 8 9 10 perform household chores such as grocery shopping. 22. A working father should not be 0 1 2 3 4 5 6 7 8 9 10 expected to spend as much time "taxiing" their child/children around as does his working wife. Would you please tell us what you think the average working woman’s annual salary is? *ttPLEASE INDICATE WHEN YOU HAVE FINISHED THE PREVIOUS SECTION AND WAIT FOR INSTRUCTIONS BEFORE CONTINUING WITH QUESTIONNAIRE. 144 Commercial #1 D. Now we’d like you to evaluate the commercial you’ve just seen. Please indicate the extent to which you agree with the statements using a scale of zero to ten, where zero means agree at all and ten means you agree totally. 10. 11. Not at all The working mother in this ad has trouble balancing her work responsibilities with her family responsibilities. This working mother is able to spend the desired amount of time at her job while being able to spend enough time with her children. The working mother in this ad does not find having a job outside the home stressful. The working mother in this ad is dissatisfied with her role as mother. The working mother in this ad is satisfied with her job. This working mother’s husband spends as much time taking care of their child/children does his working wife. This working mother’s husband does no dish-washing or clothes—washing. This working mother’s husband does not spend as much time "taxi-ing" their child/children around as does his working wife. This working mother’s husband does grocery—shopping for their family. This working mother’s husband does not prepare family meals. This working mother spends as much time as husband attending their child/children’s activities. 0 l 2 3 4 5 6 7 following you do not Totally 8 9 10 8 9 10 1 1 8 9 10 8 9 10 8 9 10 8 9 10 8 9 10 8 9 10 8 9 10 8 9 10 8 9 10 12. 13. 14. 15. 16. 17. 18. 19. 20. E. the ad at all, and ten means please tell us how well you think the following statements 145 Not at all The working mother in this ad 0 1 2 spends more time than her husband helping their child/ children with their homework. My life is very similiar to that 0 l 2 of the working woman featured in this ad. I would like my life to be 0 l 2 similiar to that of the working mother in this ad. I experience more job-related 0 l 2 stress than the working mother in this ad. The working mother in this ad 0 1 2 is happier being a mother than I am. My spouse and I do not share 0 l 2 child-care responsibilities as much as the working mother and her spouse in this ad do. My spouse helps out with 0 l 2 dish~washing and clothes—washing as much as this working mother’s spouse does in the commercial. My spouse does not help prepare 0 l 2 family meals as often as the husband in this ad does. This working mother’s husband 0 1 2 grocery shops for their family more often than does my husband. Totally 9 10 9 10 9 10 9 10 9 10 9 10 9 10 9 10 9 10 Next, we’d like you to rate this ad along a few other dimensions. Using a ten-point scale where zero means the statement doesn’t describe ad you have just seen. 1. 2. 3. Not at all the statement describes the ad totally, describe the Totally Portrays a realistic situation 0 1 2 3 4 5 6 7 8 9 10 Appeals to women like me 0 l 2 3 4 5 6 7 8 9 10 Something I will remember 0 1 2 3 4 5 6 7 8 9 10 146 Not at all Totally 4. Provides me with useful 0 1 2 3 4 5 6 7 8 9 10 information 5. Convinces me about the 0 l 2 3 4 5 6 7 8 9 10 product’s effectiveness 6. Makes me feel good about myself 0 1 2 3 4 5 6 7 8 9 10 7. Makes me feel good about 0 1 2 3 4 5 6 7 8 9 10 my spouse 8. Makes me want to try the product 0 l 2 3 4 5 6 7 8 9 10 9. Irritating to working mothers 0 l 2 3 4 5 6 7 8 9 10 10. Irritating to housewives 0 l 2 3 4 5 6 7 8 9 10 ll. Irritating to working fathers 0 1 2 3 4 5 6 7 8 9 10 12. Irritating to someone like me 0 l 2 3 4 5 6 7 8 9 10 13. Gives me a bad feeling about 0 l 2 3 4 5 6 7 8 9 10 myself 14. Upsetting to me 0 l 2 3 4 5 6 7 8 9 10 15. Gives me a bad feeling about 0 1 2 3 4 5 6 7 8 9 10 my spouse F. Now we’d like to know how appropriate you think this commercial portrayal is for the product category it is advertising. Please rate its appropriateness on a ten-point scale where zero means the portrayal is not at all appropriate and ten means the portrayal is totally appropriate. Not at all Totally 0 l 2 3 4 5 6 7 8 9 10 ***PLEASE INDICATE WHEN YOU HAVE FINISHED THE PREVIOUS SECTION AND WAIT FOR INSTRUCTIONS BEFORE CONTINUING WITH QUESTIONNAIRE. 147 Commercial #2 G. using a scale of zero to ten, We’d like you to evaluate the commercial you’ve just seen. indicate the extent to which you agree with the following statements where zero means you do not agree at all and ten means you agree totally. 10. 11. Not at all The working mother in this ad has trouble balancing her work responsibilities with her family responsibilities. This working mother is able to spend the desired amount of time at her job while being able to spend enough time with her children. The working mother in this ad does not find having a job outside the home stressful. The working mother in this ad is dissatisfied with her role as mother. The working mother in this ad is satisfied with her job. This working mother’s husband spends as much time taking care of their child/children does his working wife. This working mother’s husband does no dish—washing or clothes—washing. This working mother’s husband does not spend as much time "taxi-ing" their child/children around as does his working wife. This working mother’s husband does grocery—shopping for their family. This working mother’s husband does not prepare family meals. This working mother spends as much time as husband attending their child/children’s activities. Please Totally 0 1 2 3 4 5 6 7 8 9 10 10 10 10 10 10 10 10 10 10 10 148 Not at all Totally 12. The working mother in this ad 0 l 2 3 4 5 6 7 8 9 10 spends more time than her husband helping their child/ children with their homework. 13. My life is very similiar to that 0 l 2 3 4 5 6 7 8 9 10 of the working woman featured in this ad. 14. I would like my life to be 0 1 2 3 4 5 6 7 8 9 10 similiar to that of the working mother in this ad. 15. I experience more job-related 0 1 2 3 4 5 6 7 8 9 10 stress than the working mother in this ad. 16. The working mother in this ad 0 l 2 3 4 5 6 7 8 9 10 is happier being a mother than I am. 17. My spouse-and I do not share 0 l 2 3 4 5 6 7 8 9 10 child—care responsibilities as much as the working mother and her spouse in this ad do. 18. My spouse helps out with 0 l 2 3 4 5 6 7 8 9 10 dish—washing and clothes-washing as much as this working mother’s spouse does in the commercial. 19. My spouse does not help prepare 0 l 2 3 4 5 6 7 8 9 10 family meals as often as the husband in this ad does. 20. This working mother’s husband 0 l 2 3 4 5 6 7 8 9 10 grocery shops for their family more often than does my husband. H. Next, we’d like you to rate this ad a long a few other dimensions. Using a ten-point scale where zero means the statement doesn’t describe the ad at all, and ten means the statement describes the ad totally, please tell us how well you think the following statements describe the ad you’ve just seen. Not at all Totally 1. Portrays a realistic situation 0 1 2 3 4 5 6 7 8 9 10 2. Appeals to women like me 0 1 2 3 4 5 6 7 8 9 10 3. Something I will remember 0 l 2 3 4 5 6 7 8 9 10 149 Not at all Totally 4. Provides me with useful 0 l 2 3 4 5 6 7 8 9 10 information 5. Convinces me about the 0 1 2 3 4 5 6 7 8 9 10 product’s effectiveness 6. Makes me feel good about myself 0 l 2 3 4 5 6 7 8 9 10 7. Makes me feel good about 0 l 2 3 4 5 6 7 8 9 10 my spouse 8. Makes me want to try the product 0 1 2 3 4 5 6 7 8 9 10 9. Irritating to working mothers 4 0 l 2 3 4 5 6 7 8 9 10 10. Irritating to housewives 0 l 2 3 4 5 6 7 8 9 10 ll. Irritating to working fathers 0 1 2 3 4 5 6 7 8 9 10 12. Irritating to someone like me 0 1 2 3 4 5 6 7 8 9 10 13. Gives me a bad feeling about 0 l 2 3 4 5 6 7 8 9 10 myself 14. Upsetting to me 0 1 2 3 4 5 6 7 8 9 10 15. Gives me a bad feeling about 0 1 2 3 4 5 6 7 8 9 10 my spouse 1. Now we’d like to know how appropriate you think this commercial portrayal is for the product category it is advertising. Please rate its appropriateness on a ten—point scale where zero means the portrayal is not at all appropriate and ten means the portrayal is totally appropriate. Not at all Totally 0 l 2 3 4 5 6 7 8 9 10 ***PLEASE INDICATE WHEN YOU HAVE FINISHED THE PREVIOUS SECTION AND WAIT FOR INSTRUCTIONS BEFORE CONTINUING WITH QUESTIONNAIRE. 150 Commercial #3 J. Once again, we’d like you to evaluate the commercial you’ve just seen. Please indicate the extent to which you agree with the following statements using a scale of zero to ten, where zero means you do not agree at all and ten means you agree totally. Not at all Totally l. The working mother in this ad 0 l 2 3 4 5 6 7 8 9 10 has trouble balancing her work responsibilities with her family responsibilities. 2. This working mother is able to 0 l 2 3 4 5 6 7 8 9 10 spend the desired amount of time at her job while being able to spend enough time with her children. 3. The working mother in this ad 0 l 2 3 4 5 6 7 8 9 10 does not find having a job outside the home stressful. 4. The working mother in this ad 0 1 2 3 4 5 6 7 8 9 10 is dissatisfied with her role as mother. 5. The working mother in this ad 0 1 2 3 4 5 6 7 8 9 10 is satisfied with her job. 6. This working mother’s husband 0 1 2 3 4 5 6 7 8 9 10 spends as much time taking care of their child/children does his working wife. 7. This working mother’s husband 0 l 2 3 4 5 6 7 8 9 10 does no dish-washing or clothes—washing. 8. This working mother’s husband 0 l 2 3 4 5 6 7 8 9 10 does not spend as much time "taxi-ing" their child/children around as does his working wife. 9. This working mother’s husband 0 l 2 3 4 5 6 7 8 9 10 does grocery—shopping for their family. 10. This working mother’s husband 0 1 2 3 4 5 6 7 8 9 10 does not prepare family meals. 11. This working mother spends as 0 l 2 3 4 5 6 7 8 9 10 much time as husband attending their child/children’s activities. 151 Not at all Totally 12. The working mother in this ad 0 l 2 3 4 5 6 7 8 9 10 spends more time than her husband helping their child/ children with their homework. 13. My life is very similiar to that 0 l 2 3 4 5 6 7 8 9 10 of the working woman featured in this ad. 14. I would like my life to be 0 l 2 3 4 5 6 7 8 9 10 similiar to that of the working mother in this ad. 15. I experience more job—related 0 l 2 3 4 5 6 7 8 9 10 stress than the working mother in this ad. 16. The working mother in this ad 0 l 2 3 4 5 6 7 8 9 10 is happier being a mother than I am. 17. My spouse and I do not share 0 1 2 3 4 5 6 7 8 9 10 child—care responsibilities as much as the working mother and her spouse in this ad do. 18. My spouse helps out with 0 l 2 3 4 5 6 7 8 9 10 dish-washing and clothes-washing as much as this working mother’s spouse does in the commercial. 19. My spouse does not help prepare 0 1 2 3 4 5 6 7 8 9 10 family meals as often as the husband in this ad does. 20. This working mother’s husband 0 1 2 3 4 5 6 7 8 9 10 grocery shops for their family more often than does my husband. K. Next, we’d like you to rate this ad a long a few other dimensions. Using a ten-point scale where zero means the statement doesn’t describe the ad at all, and ten means the statement describes the ad totally, please tell us how well you think the following statements describe the ad you’ve just seen. Not at all Totally 1. Portrays a realistic situation 0 l 2 3 4 5 6 7 8 9 10 2. Appeals to women like me 0 l 2 3 4 5 6 7 8 9 10 3. Something I will remember 0 l 2 3 4 5 6 7 8 9 10 152 Not at all Totally 4. Provides me with useful 0 l 2 3 4 5 6 7 8 9 10 information 5. Convinces me about the 0 1 2 3 4 5 6 7 8 9 10 product’s effectiveness 6. Makes me feel good about myself 0 l 2 3 4 5 6 7 8 9 10 7. Makes me feel good about 0 1 2 3 4 5 6 7 8 9 10 my spouse 8. Makes me want to try the product 0 l 2 3 4 5 6 7 8 9 10 9. Irritating to working mothers 0 1 2 3 4 5 6 7 8 9 10 10. Irritating to housewives 0 l 2 3 4 5 6 7 8 9 10 11. Irritating to working fathers 0 1 2 3 4 5 6 7 8 9 10 12. Irritating to someone like me 0 l 2 3 4 5 6 7 8 9 10 13. Gives me a bad feeling about 0 1 2 3 4 5 6 7 8 9 10 myself 14. Upsetting to me 0 l 2 3 4 5 6 7 8 9 10 15. Gives me a bad feeling about 0 l 2 3 4 5 6 7 8 9 10 my spouse L. Now we’d like to know how appropriate you think this commercial portrayal is for the product category it is advertising. Please rate its appropriateness on a ten—point scale where zero means the portrayal is not at all appropriate and ten means the portrayal is totally appropriate. Not at all Totally 0 l 2 3 4 5 6 7 8 9 10 ***PLEASE INDICATE WHEN YOU HAVE FINISHED THE PREVIOUS SECTION AND WAIT FOR INSTRUCTIONS BEFORE CONTINUING WITH QUESTIONNAIRE. 153 Commercial #4 M. Once again, we’d like you to evaluate the commercial you’ve just seen. Please indicate the extent to which you agree with the following statements using a scale of zero to ten, where zero means you do not agree at all and ten means you agree totally. Not at all 1. The working mother in this ad has trouble balancing her work responsibilities with her family responsibilities. 2. This working mother is able to spend the desired amount of time at her job while being able to spend enough time with her children. 3. The working mother in this ad does not find having a job outside the home stressful. 4. The working mother in this ad is dissatisfied with her role as mother. 5. The working mother in this ad is satisfied with her job. 6. This working mother’s husband spends as much time taking care of their child/children does his working wife. 7. This working mother’s husband does no dish-washing or clothes—washing. 8. This working mother’s husband does not spend as much time "taxi—ing" their child/children around as does his working wife. 9. This working mother’s husband does grocery-shopping for their family. 10. This working mother’s husband does not prepare family meals. 11. This working mother spends as much time as husband attending their child/children’s activities. 1 2 3 4 5 6 7 8 Totally 9 10 9 10 9 10 9 10 9 10 9 10 9 10 9 10 9 10 9 10 9 10 12. 13. 14. 15. 16. 17. 18. 19. 20. N. l. 2. 3. 154 The working mother in this ad spends more time than her husband helping their child/ children with their homework. My life is very similiar to that 0 of the working woman featured in this ad. I would like my life to be 0 similiar to that of the working mother in this ad. I experience more job—related _ 0 stress than the working mother in this ad. The working mother in this ad 0 is happier being a mother than I am. My spouse and I do not share 0 child-care responsibilities as much as the working mother and her spouse in this ad do. My spouse helps out with 0 dish-washing and clothes—washing as much as this working mother’s spouse does in the commercial. My spouse does not help prepare 0 family meals as often as the husband in this ad does. This working mother’s husband 0 grocery shops for their family more often than does my husband. Not at all 1 2 1 3 Totally 9 10 9 10 9 10 9 10 9 10 9 10 9 10 9 10 9 10 Next, we’d like you to rate this ad a long a few other dimensions. Using a ten-point scale where zero means the statement doesn’t describe the ad at all, and ten means the statement describes the ad totally, please tell us how well you think the following statements describe the ad you’ve just seen. Not at all Portrays a realistic situation 0 Appeals to women like me Something I will remember 0 Totally l 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10 l 2 3 4 5 6 7 8 9 10 155 Not at all Totally 4. Provides me with useful 0 1 2 3 4 5 6 7 8 9 10 information 5. Convinces me about the 0 1 2 3 4 5 6 7 8 9 10 product’s effectiveness 6. Makes me feel good about myself 0 l 2 3 4 5 6 7 8 9 10 7. Makes me feel good about 0 1 2 3 4 5 6 7 8 9 10 my spouse 8. Makes me want to try the product 0 l 2 3 4 5 6 7 8 9 10 9. Irritating to working mothers 0 l 2 3 4 5 6 7 8 9 10 10. Irritating to housewives 0 l 2 3 4 5 6 7 8 9 10 ll. Irritating to working fathers 0 l 2 3 4 5 6 7 8 9 10 12. Irritating to someone like me 0 l 2 3 4 5 6 7 8 9 10 13. Gives me a bad feeling about 0 l 2 3 4 5 6 7 8 9 10 myself 14. Upsetting to me ‘ 0 l 2 3 4 5 6 7 8 9 10 15. Gives me a bad feeling about 0 1 2 3 4 5 6 7 8 9 10 my spouse 0. Now we’d like to know how appropriate you think this commercial portrayal is for the product category it is advertising. Please rate its appropriateness on a ten-point scale where zero means the portrayal is not at all appropriate and ten means the portrayal is totally appropriate. Not at all Totally 0 1 2 3 4 5 6 7 8 9 10 ***PLEASE CONTINUE WITH QUESTIONNAIRE ------ > 156 P. Now we’re interested to know how much you like the following products. Based on what you know about these products, please evaluate each product using a scale ranging from 0 to 10, where zero means you do not like the product at all, and ten means you like the product totally. Circle the appropriate number on the scale to the right. Not at all Totally 1. Sunrise dishwashing detergent 0 1 2 3 4 5 6 7 8 10 2. TEST PRODUCT 0 l 2 3 4 5 6 7 8 9 10 3. Bounce fabric softener 0 1 2 3 4 5 6 7 8 9 10 4 Spray—n—Wash 0 l 2 3 4 5 6 7 8 9 10 5. Del Monte catsup 0 l 2 3 4 6 7 8 9 10 6 Lipton soup 0 l 2 3 4 5 6 7 8 9 10 7. TEST PRODUCT 0 l 2 3 4 5 6 7 8 9 10 8. Velveeta cheese 0 l 2 3 4 5 6 7 8 9 10 9. Pepsi-cola 0 l 2 3 4 5 6 7 8 9 10 10. Emeraude cologne 0 l 2 3 4 5 6 7 8 9 10 11. Jean Nate body splash 0 l 2 3 4 6 7 8 9 10 12. TEST PRODUCT 0 l 2 3 4 5 6 7 8 9 10 13. Raintree body lotion 0 l 2 3 4 5 6 7 8 9 10 14. L’eggs sheer energy pantyhose 0 l 2 3 4 5 6 7 8 9 10 15. Sealy Posturpedic mattress 0 l 2 3 4 5 6 7 8 9 10 16. TEST PRODUCT 0 1 2 3 4 5 6 7 8 9 10 17. TEST PRODUCT 0 l 2 3 4 5 6 7 8 9 10 18. TEST PRODUCT 0 1 2 3 4 5 6 7 8 9 10 19. Ford Mustang 0 l 2 3 4 5 6 7 8 9 10 20. Commodore computer 0 1 2 3 4 5 6 7 8 9 10 21. Lincoln-Mercury Cougar XR7 0 l 2 3 4 5 6 7 8 9 10 22. Cedar Point amusement park 0 1 2 3 4 5 6 7 8 9 10 23. TEST PRODUCT 0 l 2 3 4 5 6 7 8 9 10 24. Detroit zoo 0 1 2 3 4 5 6 7 8 9 10 Q. the following products. please evaluate each product using a scale ranging from 0 to 10, 157 Next, we’re interested to know how likely you would be to purchase Based on what you know about these products, where zero means you are not at all likely to purchase the product, and ten means you are totally likely to purchase the product. appropriate number on the scale to the right. 10. 11. 12. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. Sunrise dishwashing detergent TEST PRODUCT Bounce fabric softener Spray-n-Wash Del Monte catsup Lipton soup TEST PRODUCT Velveeta cheese Pepsi-cola Emeraude cologne Jean Nate body splash TEST PRODUCT . Raintree body lotion L’eggs sheer energy pantyhose Sealy Posturpedic mattress TEST PRODUCT TEST PRODUCT TEST PRODUCT Ford Mustang Commodore computer Lincoln-Mercury Cougar XR7 Cedar Point amusement park TEST PRODUCT Detroit zoo Not at all 1 0 0 0 2 3 bfi-b-hh-D-fi-hh-DrP-b Ark-D J: UIUIOIO'IOI 0! 030390) 03030903 (330)030) 030') ~: «3 ~: ~: ooooooooco oooooooooooooo oooooooooooo Circle the Totally 9 10 (D 10 10 10 10 10 10 10 10 10 (DCDGDQCOCDCOCDCD 10 10 (D 10 10 10 10 wcoaococo 10 9 10 9 10 9 10 9 10 9 10 9 10 9 10 R. women’s roles. following statements 158 Next, we’d like to know what some of your beliefs are Please indicate the extent to which you agree with the using a scale ranging from zero to ten, where zero means you do not agree at all and ten means you agree totally. 10. Not at all Most working women have executive or managerial jobs. Most working mothers have executive or managerial jobs. Few working women make more than $13,000 a year. Few working mothers make more than $13,000 a year. Most working women make more than $20,000 a year. Most working mothers make more than $20,000 a year. Working mothers easily balance their job responsibilities with their family responsibilities. Working mothers are able to spend the desired amount of time at work and also spend enough time with their child/children. Working mothers do not find having a job outside the home stressful. Working fathers spend as much time with their children as do working mothers. Working women receive help from their spouse when it comes to household chores such as dish— washing or clothes—washing. Working fathers help their working wives when it comes to fixing family meals. Working fathers spend as much time "taxi—ing" their kids as do working mothers. 0 1 2 3 Totally 7 8 9 10 7 8 9 10 7 8 9 10 7 8 9 10 7 8 9 10 7 8 9 10 7 8 9 10 7 8 9 10 7 8 9 10 7 8 9 10 7 8 9 10 7 8 9 10 7 8 9 10 concerning 159 Not at all Totally 14. Working fathers do not spend as 0 l 2 3 4 5 6 7 8 9 10 much time attending their kids’ activities as do working mothers. 15. Working fathers spend as much 0 l 2 3 4 5 6 7 8 9 10 time helping their child or children with homework as does the working mother. 16. Working fathers do not grocery 0 l 2 3 4 5 6 7 8 9 10 shop for their families as often as do working mothers. 17. A working father should not be 0 l 2 3 4 5 6 7 8 9 10 expected to spend as much time with his kids as does his working wife. 18. A working father should be 0 1 2 3 4 5 6 7 8 9 10 expected to spend as much time helping his children with homework as does his working wife. 19. A working father should not be 0 l 2 3 4 5 6 7 8 9 10 expected to spend as much time dish-washing and clothes—washing as does his working wife. 20. A working father should not be 0 l 2 3 4 5 6 7 8 9 10 expected to be responsible for preparing family meals. 21. A working father should help 0 l 2 3 4 5 6 7 8 9 10 perform household chores such as grocery shopping. 22. A working father should not be 0 1 2 3 4 5 6 7 8 9 10 expected to spend as much time "taxiing" their child/children around as does his working wife. Would you please tell us what you think the average working woman’s annual salary is? 160 S. Now we’re interested to know a few more of your personal beliefs regarding women’s roles. Please indicate whether the following statements apply to you using a scale of zero to ten, where zero means the statement does not apply at all and ten means the statement applies totally. Not at all Totally 1. My job responsibilities prevent 0 l 2 3 4 5 6 7 8 9 10 me from spending time with my child/children. 2. I’m not able to spend enough 0 l 2 3 4 5 6 7 8 9 10 time with my child/children due to my job. 3. My responsibilities as a mother 0 l 2 3 4 5 6 7 8 9 10 limit the time I can devote to my job. 4. I would like to be able to spend 0 1 2 3 4 5 6 7 8 9 10 more time at my job. 5. Being a mother is the most 0 l 2 3 4 5 6 7 8 9 10 important thing in my life. 6. My job is the most important 0 1 2 3 4 5 6 7 8 9 10 thing in my life. 7. I would not choose to work 0 1 2 3 4 5 6 7 8 9 10 for pay outside the home if I didn’t have to. 8. I am satisfied with my job. 0 l 2 3 4 5 6 7 8 9 10 9. I am satisfied with my role 0 l 2 3 4 5 6 7 8 9 10 as a mother. 10. I feel that no matter how hard 0 l 2 3 4 5 6 7 8 9 10 I work, I still can’t accomplish everything I need to in a day. 11. I would like my spouse to help 0 l 2 3 4 5 6 7 8 9 10 out more with duties such as dish—washing, clothes-washing. 12. I would like my spouse to help 0 1 2 3 4 5 6 7 8 9 10 out more preparing family meals. 13. I would like my spouse to help 0 1 2 3 4 5 6 7 8 9 10 out more in in taking care of our child/children. 14. My spouse and I share household 0 1 2 3 4 5 6 7 8 9 10 responsibilities, such as cooking and cleaning, to my satisfaction. 161 Please answer the following three questions so they total 100 percent: 1. What percent of child—caring do you perform? ____________ 2. What percent does your spouse perform? __________ 3. What percent is performed by others, i.e. day-care services, baby-sitters? ___________ Finally, we’d like to ask you a few things about yourself. 1. What is your area of residence? ____________ Lansing ____________ East Lansing ____________ Okemos other: _______________ __ 2. What is your type of residence? a) __________ __ house ____________ apartment other: _________________ b) ____________ own ____________ rent other 3. What is your occupation? clerical/technical employee supervisor other: 4. In what campus department do you work? 5. What is your marital status? single married divorced separated other: 10. 11. 14. 162 Have you ever been married before? a) If you are married, is your husband employeed? b) If so, what is your spouse’s occupation? c) On the average, how many hours a week does your spouse work? d) Does he work at home or outside your home, i.e. at an office? at home outside home a) How many children live in your household? b) What is/are their age(s)? How many hours of prime—time television did you watch yesterday? What magazines do you subscribe to or regularly read? 163 16. What is your age? 17. a) What is your annual salary? b) What is your total household income? 18. Today’s date: APPENDIX B MICHIGAN STATE UNIVERSITY 164 OFFICE OF THE PROVOS'I' ADMINISTRATION BUILDING EAST LANSING 0 MICHIGAN 0 48824-1046 June 26, 1985 MEMORANDUM To: Selected MSU Staff Women S 0* From: Lou Anna Kimsey Simon, Assistant Provost Subject: Research Study Often the University is asked to release the names of employees who have certain characteristics (e.g. gender, marital status). Under law and policy, we do not release this type of information to a third party without the permission of the employee. The public record is considered to be the telephone directory. As chairperson of the University Committee on the , VI Release of Confidential Information, requests for an exception to this policy .'§:- come to me. Recently, a research proposal was presented by Ms. Sherri J. Crispin and Dr. Bruce Vanden Bergh for review through the appropriate University review procedures, including the University Committee on Research Involving Human Subjects (UCRIHS). The focus of the research is to provide information about working women's attitudes toward advertisements that feature female characters. The UCRIHS found that the study conformed to the applicable protocols for use of human subjects in research activities. Since the study also involves contacting selected staff women holding clerical-technical and supervisory positions to determine their interest in participating in the study, permission for the release of confidential information is required. After consultation with the Office of Personnel and Employee Relations, it was agreed that I would write selected staff women and inform them of the opportunity to voluntarily participate in this study. Only the names of those women who indicate on the enclosed form that they want to participate in the study will be given to Ms. Crispin and Dr. Vanden Bergh. Attached is a brief description of the project. If you would like for me to release your name to Ms. Crispin and Dr. Vanden Bergh, please send the enclosed form to me. If you have any questions, please let me know. LAKS:jmc ' Attachment CC: Dr. Butterfield Dr. Bredeck Dr. Groty MSI 'Ix an AI/irmalivc Artinn /Equal Opportunity lnxlllutmn 165 Dear MSU Employee: We in the Department of Advertising at MSU are conducting a study to examine how women feel about commercials that feature women as main characters. We are seeking participants who are women employed on campus as clerical/technical workers or as supervisors. We are specifically interested in such employees who are also mothers. If you are a C/T employee or a supervisor and are a mother, then you qualify as a participant. As previously stated, we are interested in working women’s attitudes toward commercials that feature female main characters. We will be asking participants to evaluate a list of products and also to evaluate several television commercials. Finally, participants will be asked to complete a questionnaire about womens’ social, personal, family, and work roles. We hope that by finding out how women viewers feel about advertising portrayals of women, that advertising messages can be improved in the future. It has been our experience that people find their participation in research of this sort to be positive and interesting. We hope you will consider taking part. The study will take place the week of July 29 to August 3, 1985 in the Communication Arts building here on campus at a time of your convenience. The study involves attending one session that will last between 1 to 1-1/2 hours. As an MSU employee, you have several options for participating. If you wish to participate during regular business hours, you may either take personal leave time, or you may get permission for release time from your supervisor or department head. Of course, you have the option of participating in the evening or on a Saturday if you wish. Your participation will be kept strictly confidential, and your responses will be identifiable by a name—coded number to which only we will have access. After the research is completed, we would be happy to notify you of the general results and to answer any questions you may have about the research. If you think you.would like to participate, please fill out the enclosed response form. All you need to do then is to fold the response form, staple it and send it through campus mail to the Office of the Assistant Provost, whose address is on the back of the form. Thanks for your time, and we hope to hear from you very soon. Sincerely, Sherri J. Crispin, Masters candidate in Advertising Bruce Vanden Bergh, Director of Graduate Studies, Advertising 166 *t*** RESEARCH PARTICIPANT RESPONSE SHEET ***** PLEASE RETURN BY FRIDAY, JULY 17, 1985 I do. . . I do not. . .wish to voluntarily participate in a study of women’s attitudes toward commercials that feature female characters, to be conducted by members of the MSU Advertising department. I understand that the university will release my name to the aforementioned researchers only if I agree to participate. NAME (print) ADDRESS TELEPHONE (where you can most easily be reached.) HOURS WE SHOULD TELEPHONE SIGNATURE: DATE: APPENDIX C MKEHGANSTATEIDHVERETY 167 OFFICE OF THE PROVOST EAST LANSING ' MICHIGAN 0 48824-1046 ADMINISTRATION BUILDING July 25, 1985 MEMORANDUM To: Select MSU Staff Women K; From: Lou Anna Kimsey Simon, Assistant Provost Subject: Research Study The attached memorandum from Ms. Crispin is being communicated to you from my office to assure confidentiality under University policy. This follow-up communication is designed to remind you of the opportunity to voluntarily participate in this study. Again, only the names of those women who indicate on the enclosed form that they want to participate in the study will be given to the representatives of the Department of Advertising. If you have any questions, please call me (3—380). LAKS:jc Attachments MSU is an Affirmatuve Action/Equal Opportunity Institution 168 July 23, 1985 Dear MSU Staff Woman, You may have received a letter in early July sent from the Office of the Assistant Provost informing you of the opportunity to participate in a study of women’s attitudes toward advertising. We in the Advertising Department thank the women who have agreed to participate; still, we need a few more participants. We specifically need participants who are married women with one or more children under 18 years old living at home. If you fit this description, we need you! The study will take place from August 12 to August 31 in the Communication building here on campus at a time of your convenience. The study involves attending one session that will last between 1 to 1—1/2 hours. If you have already responded, thank you. If you have not yet responded and are interested in participating in the study, please fill out the enclosed green response form. Simply fold the response form, staple it, and send it through campus mail to the Office of the Assistant Provost; it’s already addressed for you. If you did not receive the original mailing and would like more information about the study, please contact Dr. Simon, Assistant Provost, at 353-5380. Once again, we appreciate your attention and hope to hear from you very soon. Thank you, Sherri J. Crispin Masters candidate in Advertising 169 WOMEN’S ATTITUDES TOWARD ADVERTISING STUDY--1985 ***** RESEARCH PARTICIPANT RESPONSE SHEET ***** PLEASE RETURN BY WEDNESDAY, JULY 31, 1985 I do. . . I do not. . .wish to voluntarily participate in a study of women’s attitudes toward commercials that feature female characters, to be conducted by members of the MSU Advertising department. I understand that the university will release my name to the aforementioned researchers only if I agree to participate. NAME (print) DEPARTMENT TELEPHONE (where you can most easily be reached.) HOURS WE SHOULD TELEPHONE Please check the following that apply to you: I am married I have one or more children under 18 years old living in my home. Age(s) of child/children: SIGNATURE: DATE: APPENDIX D 170 SAMPLE PROFILE Sample size = 77 WORKiFAMILY DEMOGRAPHICS Median age of subject (in years): 34. Occupation: Clerical/Technical 88. Supervisor 12. Area of Residence: Lansing 33. East Lansing 9. Okemos/Haslett 13. Lansing Areas 14. East Lansing Areas 22. Farther outskirts 9. Married Before: Yes 22.0% No 78.0 Spouse’s Occupation: Professional 17.0% Managerial 12.0 Technical 13.0 Clerk 17.0 Skilled 14.0 Unskilled 17.0 Unemployed 6.0 Other 4.0 Number of children in hh: 014)me p—I 0 000000 39 Average total household income: $39,125.0 Average annual salary: $15,050.0 Type of residence: House 87.0% Apartment 8.0 Other 5.0 Own or rent: Own 77.0% Rent 23.0 Median Years Married: 9.8 Spouse’s work week (median hours): 40.4 Where spouse works: Home 5.0% Outside home 91.0 Both 4.0 Median age of children: lst child 7.3 2nd child 11.3 3rd child 14.0 171 SAMPLE PROFILE ( continued ) MEDIA-USAGE DEMOGRAPHICS Median hours of TV watched yesterday: 1.6 Median number of days/week watch local news: 3.0 Median number of days/week watch national news: 3.0 Median number of days/week read daily newspaper: 4.4 Types of magazines subscribed to or regularly read (categories not mutually exclusive): Business/News 49.4% Home 35.1 Women’s 28.6 Career 15.6 Sports/Hobby 14.3 Parent/Child 9.1 Reader’s Digest 9.1 Black 6.5 Other 18.2 ENDNOTES AND BIBLIOGRAPHY 10. 11. 12. 13. 14. 15. 172 ENDNOTES Bartos, Rena "What Every Marketer Should Know About Women," in James V. McNeal and Stephen W. McDaniel (eds.), Consumer Behavior: Classical and Contemporary Dimensions, (Boston: Little, Brown, and Co., 1982), p. 385. The Information Please Almanac 1986, Houghton Mifflin Company Boston, pp. 56-57. "Working Mothers and Their Families: A Fact Sheet," Commission on Working Women, Washington D.C, 1985. Ibid. Bartos, p. 388. Ibid. "Child Care Fact Sheet," Commission on Working Women, Washington D. C. 1986. Working Mothers Fact Sheet. Bartos, p. 388. Doan, Michael, "Business Shifts Its Sales Pitch for Women," U.S. News & World Report, July 6, 1981, p. 46. Ibid. Hayes, Cynthia A., "The Supermom Dilemma," Ad Week, July 1985, p. 34. Confidential Interview, May 15, 1985. Lieber, E.K., "The Professional Woman: Coping in a Two-Career Family," Sociology and Social Research, July 1982, p. 156. English, Mary McCabe, "A Cut Above the 'Slice—of—Life’ Ads," Advertising Age, October 3, 1983, p. M-10 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 173 Kanner, Bernice, "Comment: Ad Pitches," Working Woman, September 1984, 9, p. 75. Bartos, p. 388. The Information Please Almanac 1986, p. 60. Ibid. Kanner, p.75. Faver, Catherine A., "Life Satisfaction and the Life Cycle: The Effects of Values and Roles on Women’s Well-Being," Sociology and Social Research, July 1982, 66 (4), p. 436. Ibid., p. 442. Ibid., p. 447. Lieber, p. 159. Maret, Elizabeth and Barbara Finlay, "The Distribution of Household Labor Among Women in Dual-earner Families," Journal of Marriage and the Family, May 1984, 46 (2), p. 357 Ibid., p. 361. Ibid., p. 362. Holahan, Carole K. and Lucia A. Gilbert, "Conflict Between Major Life Roles: Women and Men in Dual Career Couples," Human Relations, 32 (6), 1979, p. 452. Ibid., p. 465. Henry, Susan, "Juggling the Frying Pan and the Fire; The Portrayal of Employment and Family Life in Seven Women’s Magazines, 1975-1982," The Social Science Journal, 21 (4), October 1984, p. 88. Ibid., p. 89. 32. 33. 34. 35. 36. 174 Ibid., p. 102. Aaker, Divid A. and Donald E. Bruzzone, "Causes of Irritation in Advertising," Journal of Marketing, 49, Spring 1985, p. 47. Whipple, Thomas W. and Alice E. Courtney, "How to Portray Women in TV Commercials," Journal of Advertising Research, 20 (2), April 1980, p. 58. Kirk, Roger E., Experimental Design: Procedures for the Behavioral Science, Belmont, CA: Brooks/Cole Publishing Company, 1968, p. 64—65. Ibid. 175 BIBLIOGRAPHY Blood, R.0. and D.M. Wolfe Husbands and Wives: The Dynamics of Married Living. Glencoe, I1: Free Press, 1960. Bernard, J. Women and Wives and Mothers. Chicago: Aldine, 1975. Campbell, Angus The American Way of Mating: Marriage si, children only maybe. Psychology Today, 8, 37-43. Courtney, Alice and Sarah Wernick Lockeretz A Woman’s Place: An Analysis of Role Portrayals by Women in Magazine Advertisements. Journal of Marketing Research, February 1971, 92-95. Courtney, Alice and Thomas W. Whipple Women in TV Commercials. Journal of Communication, Spring 1974, 24, 110—118. Duker, Jacob M. and Lewis R. Tucker Women’s Libbers Versus Independent Women: A Study of Preferences for Women’s Roles in Advertisements. Journal of Marketing Research, November 1977, 14, 469—475. Faver, Catherine A. Women Careers and Family: Generational and life—cycle effects on achievement orientation. Journal of Family Issues, 1981, 2, 91—112. Life Satisfaction and the Life—Cycle: The Effects of Values and Roles on Women’s Well—Being. Sociology and Social Researgh, 1982, 66 (4), 435—451. Franzwa, Helen H. Working Women in Fact and Fiction. Journal of Communication, Spring 1974, 24, 104-9. 176 Gies, F.L., Virginia Brown, and Joyce Jennings (Walsted) and Natalie Porter TV Commercials as Achievement Scripts for Women. Sex Roles, 1984, 10 (7/8), 513-525. Henry, Susan Juggling the Frying Pan and the Fire; The Portrayal of Employment and Family Life in Seven Women’s Magazines, 1975-1982. The Social Science Journal, October 1984, 21 (4), 88. Holahan, Carole K. and Lucia A. Gilbert Conflict Between Major Life Roles: Women and Men in Dual Career Couples. Human Relations, 1979, 32 (6), 451-467. Kanter, Rosabeth M. Work and Family in the United States: A Critical Review and Agenda for Research and Policy. New York: Russell Sage, 1977a. Kelly, J. Patrick, Paul J. Solomon, and Marion Burke Male and Female Responses to Women’s Roles in Advertising. Proceeds of the Annual Conference of the American Academy of Advertising, East Lansing, Michigan: (American Academy of Advertising, 1977), 98—104. Kirk, Roger E. Experimental Design: Procedures for the Behavioral Science. Belmont, CA: Brooks/Cole Publishing Company, 1968. Leiber, Esther K. The Professional Woman: Coping in a Two-Career Family. Educational Horizons, 1980, 58 (3), 156-161. Loughlin, Beverly The Women’s Magazine Short—Story Heroine. Journalism Quarterly, Summer 1983, 60, 138—42. Lungebeel, Barbara Derrick Defining Short—Story Patterns in Good—Housekeeping. Journalism Quarterly, Autumn 1975, 52, 584—50. 177 MacKenzie, Scott B. and Richard J. Lutz Monitoring Advertising Effectiveness: A Structural Equation Analysis of the Mediating Role of Attitude Toward the Ad. Working Paper 117, Center for Marketing Studies, UCLA, 1982. Maret, Elizabeth and Barbara Finlay The Distribution of Household Labor Among Women in Dual-earner Families. Journal of Marriage and the Family, May 1984, 357—364. Mitchell, Andrew A. and Jerry C. Olson The Role od Attitude Toward the Ad as a Mediator of Advertising Effectiveness: A Test of Competing Explanations. Journal of Marketing Research, August 1981, 18, 318—332. Rapoport, R. and R.N. Rapoport The Dual Career Family. Human Relations, 1969, 22, 3—30. Shimp, Terrence A. Attitude Toward the Ad as a Mediator of Consumer Brand Choice. Journal of Advertising, 10 (2), 9-15. Swartz, Teresa A. Role Portrayal Preferences for Print Advertisements. Proceedings of the 1983 Convention of the American Academy of Advertising, Columbia, SC: University of South Carolina, 1983, 112—115. Venkatesan, M. and Jean Losco Women in Magazine Ads: 1959-1971. Journal of Advertising Research, October 1975, 14, 49-54. Whipple, Thomas W. and Alice E. Courtney How to Portray Women in TV Commercials. Journal of Advertising Research, April 1980, 20 (2), 53—59. 178 Wortzel, Lawrence H. and John M. Frisbie Women’s Role Portrayal Preferences in Advertising: An Empirical Study. Journal of Marketing, October 1974, 38, 41—46. Zajonc, Robert B. Feeling and Thinking: Preferences Need No Inferences American Psychologist, February 1980, 35, 151-175. "mm111mm“