11 “’141'4‘41‘ ’111 “11" 11111111 .11 ‘1111111111,: 11‘ 1111 41,11 111, 44444 1 11111111111111'111‘1 14,111,111, . 4 4,444, “1"“ 111“ 4‘11 “ 1 . 41111111 '114‘,111111,1’1'11111411‘4'44 11141111111144, 111141444: ', ,4 :1": 1'1’1111 111 111111114411 .1111? 1411‘ 111 :3? ' 1111444444414111111111414 411144444 4111 I 1’111‘ ‘ ,1, " ,1. 11.11.1111 '1‘4""'1 . 1 t 4 n I v f. r . .1131 '13? 'i g“ 2 1.411%“: flé‘ik ‘1 "‘11 ‘11" 111% 111111 ..3 $14111: ”:11 z 1‘ ‘1‘ . .4 3111. 1114 111"“ 1.13;" ,k _ ..~ ', 4 4, , ”11:19“; 1 . 1111..1. 1...,1. ' ‘- 11: .. "-‘e: ’1’; '1" L ' - ‘ " v‘,‘ if“ '._...,'4 '- “ 7 .1, 441,,411414, 1'31“ - ‘1“:‘4 L,f‘1‘~' 11,, 1 ‘ ”14:: "1‘14"“ I 1413119414, 441‘41411 1144.4 41‘ ‘1:".‘4 44,44 ""‘ 11 :,,4444441‘1'1,1,4 , L414} 44444 4 '1411411, 111. :11,“ "'11 1'1 ,14144,’ ' ’1' "'1‘: 1111' '1'1 1 ...11 , ‘,_ ,_ ,. . .1 '11" .1..‘.-11- .. .. l " ‘:1Ivo|" _. .11“' I' I 1 I “‘" " "" I11‘: 11‘4”“ '1 1111.1 “ ‘ .. “ p1‘ ,1 , ' , 4 ' .41 ‘ 7 ““11‘11‘1" ‘1411..‘. “A 1' "f . 1 ‘ 1;:«44 '13‘ ‘ 71% ‘ w 11‘ :‘, > ‘ ‘ 119-". 1.1' 1‘111‘11 111 '1, 1 ‘- .. '1’ ‘1‘ 414‘: 4194 4 4414,, 11mm” 11“”. ‘1431441114'5 ' 14‘ 1'44433. 41‘ ‘m,.‘ 4144‘ 4 ”fl 1“.“1 {<3 '14, LIBRARY Michigan State University Ill!llllllllllfllflll|llljlllljjlilllllsllllllllj] 37 This is to certify that the thesis entitled THE PORTRAYAL OF MEN AND WOMEN IN AMERICAN TELEVISION COMMERCIALS: A REPLICATION AND EXTENSION presented by Michael D. Dauria has been accepted towards fulfillment of the requirements for Masters degree in Sociology ({SRVW “Q. Cme Major professor Date January 24, 1979 0-7639 l. mama L? a * M5511 Mil-{ii fill/{WHY gnv l 61996. bx lfiz . «$46339 A. w: ’ 1% . w :---m "imifm‘ (f) MAYEB 1 2002 ‘ J p. JEN: a; {12' 1 our “WW“ ‘ OVERDUE FINES ARE 25¢ PER DAY PER ITEM Return to book drop to remove this checkout from your record. M 22 m, Fifi («It . g. 1 , oil's? “It 2 0 1°”- ; 39%:ch ' ,2..nd n~fi Q“ U 913;) THE PORTRAYAL OF MEN AND WOMEN IN AMERICAN TELEVISION COMMERCIALS: A REPLICATION AND EXTENSION BY Michael D. Dauria A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Sociology 1979 ABSTRACT THE PORTRAYAL OF MEN AND WOMEN IN AMERICAN TELEVISION COMMERCIALS: A REPLICATION AND EXTENSION BY Michael D. Dauria This study is a replication and extension of an earlier work conducted by McArthur and Resko (1975) con- cerning the portrayal of men and women in American tele- vision commercials. The characteristics of adult male and female models in randomly selected television commer- cials were systematically coded and content analyzed by the authors, and several significant sex differences were discovered. Focusing on the advertising viewed in millions of American homes, I also found that women are portrayed quite differently from their male counterparts. In my analysis of a sample of Fall 1978 network television ads. males and females manifested different behaviors which were followed by different consequences. This research suggests the possible influence of television commercials on sex stereotyped behaviors, which tend to portray women in an unfavorable light and reinforce sexual differences. TABLE OF CONTENTS Page LIST OF TABLES . . . . . . . . . . . . . . . . . . . iv INTRODUCTION . . . . . . . . . . . . . . . . . . . . Statement of the Problem . . . . . . . . . . . METHODOLOGY . . . . . . . . . . . . . . . . . . . . Sample . . . . . . . . . . . . . . . . . . . . COding O O O O O O O I O O O O O O O O O Validity and Objectivity of Measures . . . . . 1 Heads O\ l" H RESULTS 0 O O O O O O O O O O O O O O O O O O O O 0 13 Comparison with McArthur's and Resko's Findings 0 O I O O O O O O O O O O O O O O O O 13 Frequency of Male and Female Central Figures . . . . . . . . . . . . . 13 Basis for the Credibility of Male and Female Central Figures . . . . . . . . . 14 Role of Male and Female Central Figures . 14 The Location of Male and Female Central Figures . . . . . . . . . . . . . . l9 Arguments Given by Male and Female Central Figures . . . . . . . . . . . . 28 Rewards Offered by Authority Central Figures . . . . . . . . . . . . . . . . 31 Rewards Reaped by Product User Central Figures . . . . . . . . . . . . . . . 33 Product Types Associated with Male and Female Central Figures . . . . . . . . . 35 Time of Day . . . . . . . . . . . . . . . 41 Additional Findings . . . . . . . . . . . . . 44 The Number of Central Figures During News Programming . . . . . . . . . . . . 44 The Frequency of Central Figures on Saturday . . . . . . . . . . . . . . . . 46 The Voice-Over . . . . . . . . . . . . . . 47 ii DISCUSSION . APPENDIX . . BIBLIOGRAPHY iii LIST OF TABLES Table Page 1. Basis for Credibility of Male and Female Central Figures . . . . . . . . . . . . . . . 15 2. Role of Male and Female Central Figures . . . 17 3. The Relational vs. Independent Roles of Male and Female Central Figures . . . . . . . . . 18 4. The Relational vs. Independent Roles of Male and Female Central Figures - Homemaker and Interviewer-Narrator Excluded . . . . . . . . 20 5. Location of Central Figures . . . . . . . . . 21 6. Location of Central Figures by Home . . . . . 23 7. Location of Central Figures by Occupational setting 0 I O O O O O O C O O O O O O O I O O 25 8. Location of Product User Central Figures by Home O O O O O O I O O O I I O O O O O O O 2 6 9. Location of Product User Central Figures by Occupational Setting . . . . . . . . . . . . 27 10. Arguments Given by Male and Female Central Figues O O O O O I O O O O O I O O O O O O O 29 11. Comparison of Argument Given by Male and Female Central Figures - One Degree of Freedom Analysis . . . . . . . . . . . . . . 30 12. Rewards Offered by Authority Central Figures - Raw Data Only . . . . . . . . . . . 32 13. Rewards Reaped by Product User Central Figures 0 O O O O O O O O O O O O O O O O O O 34 14. Product Types Associated with Male and Female Central Figures . . . . . . . . . . . 36 iv Table 15. 16. 17. 18. Food Product Type Associated with Authority and Product User Central Figures 19. 20. 21. Comparison by Home Product Associated with Male and Female Central Figures - One Degree of Freedom Analysis . . . Home Product Type Associated with Authority and Product User Central Figures Body Product Type Associated with Authority and Product User Central Figures Sex of Central Figure by Time of Day Sex of Central Figure by Time of Day - One Degree of Freedom Analysis Frequency of Occurrence of VOice-Over Which Accompanies Commercials . Page 37 39 4O 42 43 45 48 INTRODUCTION Statement of the Problem The work reported here is a replication and exten- sion of an earlier study conducted by McArthur and Resko (1975). These authors sought to determine the extent to which stereotyped portrayals of the sexes could be found in the medium of television commercials. In their inves- tigation the main area of concern was the characteristics of the male and female models in these commercials which were "sold" along with the product. McArthur and Resko randomly selected television commercials which were viewed for the purpose of syste- matically identifying particular characteristics depicted in them. Their analysis of the portrayal of males and females in adult television commercials revealed a number of significant sex differences which are consistent with current sex-role stereotypes. They found that men and women presented to the viewing audience differed in regard to their frequency, credibility, roles, location, argu- ments given, rewards reaped, and product type. The only category which was not statistically significant in their analysis was rewards offered by authority central figures. They also reported differences in the presentation of males and females by time of day. Given the growing concern in our society over the undesirable consequences of stereotyped sex roles, it would seem important to study this area further by repli- cating and extending the available research. My research suggests that one field which needs to be investigated is the strong impact that the mass communications media has upon sex-role stereotyping. Perhaps the most generalized reflection of extant sex—role stereotyping can be found in this field, including sex role definitions. These media include newspapers, magazines, movies, and popular music, but the main focus here is on television commer- cials. It seems that blatant sex-role stereotyping is overwhelmingly the case in television commercials, which are very influential as a communications medium. The mass media has long been recognized as a trans- mitter of cultural stereotypes (Allport, 1958; Liebert, 1973; Duberman, 1975). In a classic study, Child, Potter, and Levine (1946) analyzed stories in thirty third grade textbooks used to teach reading. They found marked dif- ferences in the way male and female characters were por- trayed. Male characters were significantly more likely than females to be portrayed solving problems, physically exerting themselves, and engaging in constructive or pro- ductive behavior. Of course, in 1946 literature was the predominant media form. Today that place is held by tele— vision. Research has been somewhat sparse, however, in the area of television commercials and the way in which they serve as a transmitter of sex-role stereotypes, although studies by Maccoby and her associates (1957, 1958) have reported that film viewers identify with, spend more time looking at, and remember more about same-sexed characters. But there is research evidence bearing on the question of television and sex-role stereotyping (Hennessee and Nich- olson, 1972). Analysis of the portrayal of males and fe- males in the ten most popular children's commercial tele- vision programs (Sternglanz and Serbin, 1974) and in adult television commercials (Courtney and Whipple, 1974; Dom- inick and Ranch, 1972) have revealed a number of signifi- cant sex differences which are consistent with current #fsex-role stereotypes. These authors contend that media content_wou1d be‘lessmggxist if women characters were shown to have the same occupational distribution as male 2 characters, were shown advertising the same types of pro- 1 ducts as men, and were used for the voice-overs in adver- ; tisements to the same extent as men. is»; L‘ There is some research which supports their conten- tions. An early study of television programs (Head, 1954) revealed that men held more than two-thirds of the major roles, women only one-third. More recently, DeFleur's work (1970) indicates that the situation may have deter- iorated. He found that women comprise less than 20% of the roles having definite occupational activity. Similarly, Coleman (1971) found that television commercials carry direct and subtle messages about who does what kind of! work. For instance, coffee commercials often show the wife fussing around at breakfast, serving a husband who is obviously getting ready for work while she remains at home with a stack of dirty dishes. In regard to product types, Courtney and Whipple (1974) found that women were seven times more likely to appear in ads for personal hygiene products (deoderants, toothpaste, and soap) than not appear. Analysis revealed that 75% of all ads using females were for products found in the kitchen or bathroom. The behavior of the characters also supports sex- role stereotyping. Lands and Brennan (1974) analyzed the "type of voice” used by the narrator in television ads and found that virtually all of the males were either ”factual" or "aggressive-sales pitchy." Females were overwhelmingly characterized as either "seductive" or "soft spoken." Relatedly, Parker and Lemm (1974) found that 24% of the female characters were ”silent," a cate- gory that held no males. Our society does define stereotypically a host of traits as belonging almost exclusively to one or the other sex roles, as evidenced by research conducted in the mass media. In short, society creates a radical dichotomy of human types, despite both the many differences between individuals of the same gender and the many similarities 5 between people of opposite genders. Nonetheless, while the present findings indicate a high degree of sex ster- eotyping for televised models, it would seem desirable to have further documentation of such portrayal of the sexes, and one purpose of the present investigation is to provide such evidence. My study of the portrayal of men and women in American television commercials not only re- veals sex differences which replicate some of McArthur's and Resko's research on sex-role stereotyping, but also includes an examination of commercials on Saturday and during evening news programming. In addition, an anal- ysis is made of the voice-over which accompanies many com- mercials and which seems to confer the stamp of approval on the advertised product. METHODOLOGY Sample Television commercials were drawn from broadcasts of the three major networks in the Fall of 1978. CBS was sampled on a Tuesday (October 10), ABC on a Wednesday (October 11), and NBC on a Saturday (October 14). This was different from the original study in two respects: the order in which the networks were sampled and the days chosen for viewing. The latter was important for my study, since I wanted to see if there were any significant differ- ences in the portrayal of men and women in television ads on the weekend. Bach network was viewed for a total of seven hours (McArthur and Resko viewed each station for six hours). The time periods were as follows: 9:00 A.M.- 11:00 A.M. (morning);1:00 P.M. - 3:00 P.M. (afternoon); and 6:00 P.M. - 9:00 P.M. (evening). These time periods also differed from those reported in the original study (10:00 A.M. - 12:00 Noon; 1:30 P.M. - 3:30 P.M., and 8:00 P.M. - 10:00 P.M.). I had chosen these for the sake of simplicity; they all began on the hour. Furthermore, I added an addi- tional viewing hour, as previously noted, since I wanted to analyze those commercials shown during news programming. McArthur and Resko coded every other commercial. However, they neglected to say what constituted a commer- cial. Some ads were simply not commercials, such as an- nouncements of future program viewing or political ads, the latter which were very prevalent during this election year. Therefore, I made a further clarification and coded every other commercial in which some type of pro- duct (rather than self or network program) was displayed or discussed. Ads such as those just mentioned were not treated as commercial announcements and simply ignored. For instance, if the first ad which appeared during a com- mercial break was political in nature, it was ignored and the following ad was coded (if, indeed, a product was advertised in the latter). If not, the next commercial was then coded, and so on. A total of 242 commercials were viewed during these hours, of which 154 could be coded. Eighty-eight commercials were omitted because no adult central figures appeared in them or because they were identical duplications of previously coded commercials. Note that I sampled types of commercials and not duplica- tions. Frequency, therefore, was defined as the number of different ads appearing on television, with a subsequent analysis of the categories coded for each new ad; repeti— tions were not coded. Coding The coding procedures in my study were similar to those reported in the McArthur and Resko analysis, with some minor variations. An analysis of each coding cate- gory was most important, rather than an intracategory study. Notes were taken during the commercial, at which time the central figures and other aspects of the commer- cial to be coded were recorded. The audio tape was stopped after each coded commercial, and the information was coded according to the categories as described by McArthur and Resko. In case of doubt, the commercial was replayed. (The Appendix lists and defines these categories, including the components of each.) The variations in coding were as follows: Central figures. McArthur and Resko reported that when it was unclear which two figures were most central, a central figure of each sex was coded. In addition, if there were only two adults present in the ad, both were always coded. In my analysis I decided to code only one central figure, if I thought he or she was the "central focus" of the ad more so than any other adult figures. Thus, I did not always find it necessary to code at least two central figures in those ads which featured two or more adults, nor code an adult figure of each sex when it was unclear as to which adults were most central, as the authors defined this category. For instance, one ad for a brand of potato chips featured many adult figures, but only one was the central focus (celebrity Roy Clark) by virtue of his prominent visual exposure and credibil- ity base. My study dealt with these indecisions by choos- ing only those adults, whether one or two or of the same or opposite sex, who were the central focus of the commer- cial under investigation. Thus, I defined this coding category somewhat differently from the authors. (See Appendix for full details and descriptions.) 3212- Roles were defined as relational (spouse, parent, girlfriend/boyfriend, or housewife) or independent (worker, professional, celebrity, or narrator-interviewer) as in the original study. In addition, I included "un- known" in the category labeled as "other" if it was im- possible to determine the role of the central figure. This use of ”other" to incorporate "unknown" was also applied to those coding categories outlined by the authors which employed the former term. For instance, one commercial initially showed a woman as a baseball coach for a girl's team, but her role and the location were soon altered. The woman was now in the home with her daughter as she attempted to explain the benefits of purchasing a par- ticular brand of bread. , Location. The category "home" was specified fur- ther by including outdoors around the home, such as one's yard or patio. The category "store" also included res- taurant, since it is an establishment where one purchases goods. In one sense, home and yard could be viewed as 10 private locations, office or place of work as semi-private, and store or restaurant as public. Arguments. "Scientific argument" included only those arguments where actual facts, figures, statistical results, lists, or percentages were presented by the cen- tral figure. For instance, an ad for a headache remedy which stated "54% more pain reliever per tablet than or- dinary sinus remedies," or an ad for margarine with "25% less fat or calories" would be scientific in nature. Rewards. The "other" category included not only "unknown" but also "feels nice," "good-tasting," "comfort- able," "service," or ”dependability." Type of Product. The category "foodstuffs" also included beverages, such as tea, coffee, and soft drinks. I also discovered in my study that some of these coding categories were not mutually exclusive. At times there were simultaneous roles existing in one ad, if the commercial switched scenes, purpose, or location. For instance, in one ad a female central figure was initially portrayed as a worker, but then the ad switched to her home where she was seen as a parent. An awareness of simultaneous role portrayal seemed to be an important issue here, and is related to the notion that many roles and lccations are somewhat vague in today's ads, as discussed later. It was also found that at times all coding cate- gories were inappropriate and that tally was omitted. The central figure was simply "there." This explains why some 11 tabular presentations to follow contain more or less of the specified number of central figures actually coded during the analysis. Validity and Objectivity of Measures An attempt to obtain a valid and objective set of measures was provided by interrater agreement on the cod- ing of a sample of commercials which was viewed a few days prior to the main study. This measure of interrater agreement was based on the following formula: number of agreements number of agreements plus number of disagreements X 100 My fellow coder was a female doctoral student, and we in- dependently coded commercials on two successive days (Fri- day, October 6, and Saturday, October 7). The first cod- ing session was inadequate, however. Interrater agreement was quite strong for all of the coding categories except ”reward." Therefore, we discussed our problems with this category and attempted to clear up any other minor discrep- ancies at this time. Once we were convinced that the major issues were resolved and category-coding was more fully un- derstood, we decided to code commercials for an additional two hours (Sunday, October 8, from 8:00 P.M. - 10:00 P.M.). Interrater agreement was strong for all of the coding categories in the eight commercials which were coded. The average percentage of agreement between raters regarding the characteristics of the fourteen central figures (which 12 were all coded correctly--nine males and five females) was 79% for credibility, 79% for role, 100% for location, 79% for argument, 71% for reward, 100% for type of product, and 100% for voice-over. RESULTS Comparison with McArthur's and Resko's Findings Frequency of Male and Female Central Figures McArthur and Resko reported that in the 199 com- mercials which they had coded, a total of 299 central figures were noted. Males accounted for 57% of these .central figures, and females 43%, a significant differ- ence (x2 = 5.62, p < .02, df = 1). In my replication, however, this finding was not substantiated. Of the 154 commercials I had coded, a total of 210 central figures were marked. Recall that 242 commercials were viewed, but 88 of them were omitted because they did not contain any adult central figures or they were duplications of previously coded commercials. Males accounted for 49% of these central figures (n = 104) and females 51% (n = 106), a difference which was not statistically significant (x2 = .019, .80 < p < .90, df = 1). Therefore, in my an- alysis the number of female central figures coded was slightly higher, whereas McArthur and Resko found that the number of male central figures was much greater. Note that my results may have been due to the differences in viewing times, or the authors may have double-coded their 13 14 ads, a procedure I did not adopt. Basis for the Credibility of Male and Female Central Figures The authors reported that the difference in the credibility base of male and female central figures was highly significant (x2 = 88.75, p < .001, df = 1). They found that 70% of the males were portrayed as authorities, while only 30% were portrayed as product users. Only 14% of the female central figures were portrayed as authori- ties, while the remaining 86% were cast as product users. Consistent with their results, I found a significant dif- ference in the credibility base, although it was not quite as high as that which was reported in their study (x2 = 14.07, p < .001, df = 1). This can be seen by looking at Table 1. In my study 52% of the males were portrayed as authorities and 48% were portrayed as product users. This constitutes a substantial increase in the number of product users since the original study, even though this table does not specify which types of products. For adult female central figures, 27% were portrayed as authorities while 73% were cast as product users. Role of Male and Female Central Figures A significant 2 x 9 (sex by role) chi square anal- ysis performed by McArthur and Resko indicated that male and female central figures were depicted in different 2 roles (x = 111.74, p < .001, df = 8). I found strikingly 15 Table 1. Basis for Credibility of Male and Female Central Figures* Sex of Central Figure Male Female Product User .48 .73 .60 Basis for Credibility of (49) (76) (125) Central Figure Authority .52 .27 .40 (54) (28) (82) .50 .50 1.00 (103) (104) (207) (x2 = 14.07, p < .001, df = 1) *Data given in proportions of male and female cen- tral figures by basis for credibility. Raw data given in parentheses. 16 2 = 108.397, similar results, as can be seen in Table 2 (x p < .001, df = 8). It must be pointed out, however, that the expected frequency in three of the eighteen cells was less than five. Most statistical texts agree that re— sults are still valid if expected frequencies are greater than five in at least 80% of the cells. This was con- firmed in my replication, where 83% of the cells had an expected frequency greater than five. The authors collapsed their data into a one degree of freedom matrix to determine exactly where this sex difference lay in regard to roles. They found that female central figures were more likely to be portrayed in a role which defined them in terms of their relationship to others-- a spouse, parent, girlfriend, or housewife. Males were more likely to be portrayed in a role which defined them independently of others--a worker, professional, celebrity, or interviewer-narrator (x2 = 60.74, p < .001, df = 1). As had been observed by McArthur and Resko, my replication confirmed the relational vs. independent roles of female and male central figures (x2 = 34.604, p < .001, df = 1). Table 3 shows that 70% of the female central figures were portrayed in some type of relational role, whereas 77% of the male central figures were defined in an independent - role. Specific figures were not reported in the original study. To insure that the obtained sex difference in roles was not merely a restatement of differences in the Table 2. 17 Role of Male and Female Central Figures Sex of Central Figure Male Female Spouse .50 .50 .14 (15) (15) (30) Parent .26 .74 .16 (9) (25) (34) Homemaker .00 1.00 .05 (0) (10) (10) Worker .70 .30 .13 (19) (8) (27) Professional .85 .15 .06 R°le °f (11) (2) (13) central Celebrity .75 .25 .06 (9) (3) (12) Interviewer/ .81 .19 .12 Narrator (21) (5) (26) Boyfriend/ .13 .87 .04 Girlfriend (l) (7) (3) Other .36 .64 .24 (18) (32) (50) .50 .50 1.00 (103) (107) (210) 108.397, p < .001, df = 8) 18 Table 3. The Relational vs. Independent Roles of Male and Female Central Figures Sex of Central Figure Male Female Relational .30 .70 .51 (spouse, parent, ' (25) (57) (82) girlfriend/ boyfriend, TYPE housewife) of Independent .77 .23 .49 (worker, Role professional, (60) (18) (78) celebrity, interviewer- narrator) .53 .47 1.00 (85) (75) (160) (x2 = 34.604, p < .001, df = 1) 19 credibility base for male and female central figures, McArthur and Resko also performed a sex by role analysis in which housewives and interviewers-narrators were ex- cluded, since these roles seemed to predominate the rela- tional vs. independent roles for women and men, respective- ly. They found that women continued to be defined primar- ily in terms of their relationship to others--spouse, parent, or girlfriend, while men tended to be defined inde- pendently of others--worker, professional, or celebrity (x2 = 3.94, p < .05, df = 1). These results were also con- firmed in my analysis (x2 = 19.61, p < .001, df = 1). Table 4 shows that 65% of the female central figures were more often portrayed in a role which defined them in relation to others , whereas E75% of the male central figures were portrayed in a role which defined them independently of others. The Location of Male and Female Central Figures A significant 2 x 4 (sex by location) chi square analysis indicated that male and female central figures were depicted in different locations for the McArthur and Resko study (x2 = 14.54, p < .01, df = 3). In my replica- tion, Table 5 shows these same sex differences for loca- tion (x2 = 22.55, p < .001, df = 3). A larger number of females were seen in the home, whereas most males were seen in an occupational setting. It is interesting to note that 41% of the total locations were unknown or other, which suggests that many locations in ads are 20 Table 4. The Relational vs. Independent Roles of Male and Female Central Figures - Homemaker and Interviewer-Narrator Excluded Sex of Central Figure Male Female Relational .35 .65 .58 (spouse, parent, (25) (47) (72) girlfriend/ Type boyfriend) of Independent .75 .25 .42 (worker, Role professional, (39) (13) (52) celebrity) .52 .48 1.00 (64) (60) (124) (x2 = 19.61, p < .001, df = 1) 21 Table 5. Location of Central Figures Sex of Central Figure Male Female Home .35 .65 .36 (27) (50) (77) (Store .61 .39 .11 (14) (9) (23) Location of Occupa- .88 .12 .12 Central tional Setting (22) (3) (25) Figure Other .48 .52 .41 (41) (45) (86) .49 .51 1.00 (104) (107) (211) 2 (x = 22.55, p < .001, df = 3) 22 somewhat nebulous. This was also true for roles, where roughly one-fourth were either other or unknown (see Table 2). There may be a definitive trend for roles and locations to be vague in today's television commercials. This "vagueness" may suggest a change in the portrayal of men and women in future ads. By introducing uncertainty in these ads, media advertisers may be able to circumvent current sex-role stereotypes. This was alluded to earlier in my coding procedures when I discussed simultaneous roles. These roles also suggest that vagueness may con- stitute uncertainty as well as ambiguity in commercial advertising. The category subdivisions were also collapsed into a 2 x 2 matrix in accordance with whatever subdivisions seemed to be contributing most of the overall effect. McArthur's and Resko's report revealed that female central figures were depicted in the home proportionately more often than were male central figures (x2 = 8.24, p < .01, df = l). Consistent with their findings, my analysis found sex differences by home location as well. As reported in Table 6, I found that 65% of the female central figures were depicted in the home, whereas only 35% of the male central figures were so defined (x2 = 9.81, p < .01, df = 1). McArthur and Resko did not report any figures to substantiate their results, which would lend themselves to comparative analyses. The authors also noted that male central figures 23 Table 6. Location of Central Figures by Home Sex of Central Figure Male Female Home .35 .65 .36 (27) (50) (77) Location of All other .58 .42 .64 Central categories (77) (57) (134) Figure .49 .51 1.00 (104) (107) (211) (x2 = 9.81, p < .01, df = 1) 24 were depicted in an occupational setting proportionately more often than were females (x2 = 8.65, p < .01, df = 1). My findings replicate their data as well, and were as significant as their results (x2 = 17.01, p < .001, df = 1). As can be seen in Table 7, only 12% of the female central figures were located in an occupational setting, whereas 88% of the male central figures were so defined. (One should note, however, that occupational setting consti- tuted only 12% of the total locations coded.) McArthur and Resko again did not report individual figures in their study. McArthur and Resko discovered that the tendency for females to be depicted more frequently in the home and for males to be depicted more frequently in an occupational setting held true even when only male and female product users were analyzed (x2 = 3.00, p < .10, df = 1, and x2 = 13.14, p < .001, df = 1, respectively). These results, however, were not confirmed in my replication for home (x2 = 1.24, .20 < p < .30, df = l) but only for occupa- tional setting (x2 = 15.39, p < .001, df = 1) when the credibility base under investigation was that of product user. These results are reported in Tables 8 and 9. Thus, for the home setting, there was something inherent in the role of product user which confined one to the home. Male and female central figures in a home setting did not differ in regard to product usage as a basis for credibil- ity. That is, it was not the sex of the central figure 25 Table 7. Location of Central Figures by Occupational Setting Sex of Central Figure Male Female Occupa- .88 .12 .12 tional Setting’ (22) (3) (25) Location of All other .44 .56 .88 Central categories (82) (104) (186) Figure .49 .51 1.00 (104) (107) (211) (x2 = 17.01, p < .001, df = 1) 26 Table 8. Location of Product User Central Figures by Home Sex of Central Figure Male Female Home .33 .67 .41 (17) (34) (51) Location of All other .43 .57 .59 Central categories (32) (42) (74) Figure .39 .61 1.00 (49) (76) (125) 2 (x = 1.24, .20 < p < .30, df = 1) 27 Table 9. Location of Product User Central Figures by Occupational Setting Sex of Central Figure Male Female Occupa- .82 .18 .14 tional Setting (14) (3) (17) Location of All other .34 .66 .86 Central categories (35) (73) (108) Figure .39 .61 1.00 (49) (76) (125) (x2 a 15.39, p < .001, df = 1) 28 which accounted for differences in home location, although I did find sex differences for occupational setting as in McArthur's and Resko's study. Table 9 shows that 82% of the male product users were located in an occupational setting as compared with 18% of the female central figures who were also product users. Arguments Given by Male and Female Central Figures A significant 2 x 3 chi square analysis (sex by ar- gument) was reported by McArthur and Resko for differences in arguments given by central figures in support of a pro- duct (x2 = 9.21, p < .01, df = 2). When category sub- divisions were collapsed into a 2 x 2 matrix, their re- sults also revealed that males were significantly more likely than females to give any type of argument (x2 = 27.69, p < .001, df = 1). In my replication I did not find any significant sex differences by type of argument in either the 2 x 3 analysis or in the one degree of free- dom matrix (x2 = 3.97, .10 < p < .20, df = 2 and x2 .274, .50 < p < .70, df = 1). Table 10 shows that the arguments given by male and female central figures were fairly constant. Only the argument categorized as "scientific" differed signifi- cantly for males and females, but one must note that such arguments constituted only 9% of the total sample. When my figures were collapsed into a 2 x 2 matrix, the null hypothesis again was not rejected. Table 11 29 Table 10. Arguments Given by Male and Female Central Figures Sex of Central Figure Male Female Scientific .72 .28 .09 (13) (5) (l8) Argument Non- .48 .52 .55 scientific Given by (55) (60) (115) Central No .47 .53 .36 Argument (36) (40) (76) Figure .50 .50 1.00 (104) (105) (209) (x2 = 3.97, .10 < p < .20, df = 2) 30 Table 11. Comparison of Argument Given by Male and Fe- male Central Figures - One Degree of Freedom Analysis Sex of Central Figure Male Female Scientific .51 .49 .64 or non- Argument sc1ent1f1c (68) (65) (133) . No .47 .53 .36 Given by Argument Central (36) (40) (76) Figure .50 .50 1.00 (104) (105) (209) (x2 - .274, .50 < p < .70, df = 1) 31 shows that both male and female central figures were as likely to give an argument, scientific or otherwise. In McArthur's and Resko's study, 30% of the female central figures gave no argument at all as compared with only 6% of the male central figures. As Table 11 shows, I found that 53% of the female central figures and 47% of the male central figures gave no argument at all, results which were quite different from those reported by the authors. Rewards Offered by Authority Central Figures In the original study male and female authorities did not differ in the rewards which they offered to the viewer for using the product they were advertising (x2 = 5.36, .30 < p < .50, df = 4). In my replication, however, a difference in rewards seemed to be evident. A signifi- cant chi square analysis was found, but unfortunately it was statistically inaccurate. That is, less than 80% of the cells had an expected frequency greater than five (only 75%, or six out of the eight expected cell frequencies, were adequate). A test for significance was not warranted since the approximation of the sample statistic to the chi square distribution was not very close; comparisons would be virtually meaningless. Nonetheless, Table 12 presents the raw data for this coding category. 32 Table 12. Rewards Offered by Authority Central Figures - Raw Data Only Sex of Central Figure Male Female Social (2) (9) (11) enhancement Type of Self— (7) (5) (12) Reward Offered enhancement by Central Practical (22) (5) (27) Figure Other (23) (9) (32) (54) (28) (82) 33 Rewards Reaped by Product User Central Figures McArthur and Resko found no significant sex differ- ences in the general categories of rewards accruing to males and females. Male and female central figures were equally likely to receive social enhancement, self-en- hancement, practical, or other rewards (p > .25). Strik- ingly similar results were found in my analysis (x2 = 2.38, .30 < p < .50, df = 3). Table 13 presents a listing of the data and shows these consistencies, particularly for the practical and other categories. For the prac- tical category, 50% of the males and 50% of the females reaped such rewards. In the other category, the results were 42% and 58% for males and females, respectively. Since McArthur and Resko did not present any tabular re- sults, comparisons and specific interpretations could not be made. The authors did discover sex differences in the type of reward received within the subcategory of social en- hancement (x2 = 21.21, p < .001, df = 5). When the data were collapsed into a one degree of freedom matrix, they found that females were more likely than males to obtain the approval of family and the opposite sex as reward for using a given product, while males more frequently obtained the approval of their friends, social advancement, and career advancement. However, the authors found no signif- icant differences in the type of self-enhancement or 34 Table 13. Rewards Reaped by Product User Central Figures Sex of Central Figure Male Female Social .34 .66 .29 enhancement (13) (25) (38) Type °f Self- .30 .70 .25 Reward enhancement (10) (23) (33) Reaped by Practical .50 .50 .12 Central (8) (8) (16) Figure Other .42 .58 .34 (19) (26) (45) .38 .62 1.00 (50) (82) (132) (x2 = 2.38, .30 < p < .50, df = 3) 3S practical rewards received by males and females who used a given product. In my replication I am unable to report any specific results. The expected cell frequencies for Ithe subcategories of social enhancement, self-enhancement, and practical rewards were below the minimum acceptable level for appropriate chi square analysis. There were simply not enough subjects in these categories to permit meaningful comparisons. Product Types Associated with Male and Female Central Figures A 2 x 4 (sex by product type) chi square analysis by McArthur and Resko indicated that male and female pro- duct users were associated with different types of pro- ducts (x2 = 8.97, p < .05, df = 3). In my replication this finding was also substantiated. In fact, the re- sults were almost identical (x2 = 8.00, p < .05, df = 3). Table 14 lists these differences in product types for male and female central figures. The data were then col- lapsed into a 2 x 2 matrix to find out whether or not fe- male product users were more likely than males to be identified with home products. These results were con- firmed in my study (x2 = 4.01, p < .05, df = 1), as well as in McArthur's and Resko's analysis (x2 = 6.12, p < .02, df = 1), the latter at a slightly higher significance level. They reported that 33% of the female product users were portrayed using home products as compared with 13% of the male product users. As shown in Table 15, I found that 36 Table 14. Product Types Associated with Male and Female Central Figures Sex of Central Figure Male Female Body .47 .53 .22 Product (22) (25) (47) Type of Home .37 .63 .22 Product Product. (17) (29) (46) Associated Foodstuff .51 .49 .36 with Central (38) (37) (75) Figure Other .67 .33 .20 (28) (14) (42) .50 .50 1.00 (105) (105) (210) (x2 = 8.00, p < .05, df = 3) 37 Table 15. Comparison by Home Product Associated with Male and Female Central Figures - One Degree of Freedom Analysis Sex of Central Figure Male Female Home .37 .63 .22 Type of Product (17) (29) (46) Product All other .54 .46 .78 Associated Products (88) (76) (164) with Central Figure .50 .50 1.00 (105) (105) (210) (x2 = 4.01, p < .05, df = 1) 38 63% of the females and 37% of the males were portrayed as home product users. In regard to product types, McArthur and Resko also reported that there was a general and consistent tendency for males to appear as authorities on a product which was used primarily by females. They found that while males comprised only 16% of the home product users, they ac- counted for 86% of the authorities on these products (x2 = 34.41, p < .001, df = 1). Similarly, males accounted for 78% of the authorities on body products, but only 33% of the body product users (x2 = 20.99, p < .001, df = l). I found similar trends for home products (x2 = 11.21, p < .001, df = 1) and body products (x2 = 10.30, p < .01, df = l) as seen in Tables 16 and 17, and these results were as de- finitive as those reported by McArthur and Resko. In my study of home product types, males comprised-18% of the home product users and accounted for 67% of the authori- ties on these products (Table 16). Thus, these aspects of male central figures have not changed considerably from the results reported in the original study. As for body product types, Table 17 shows that males comprised 32% of the body product users and 85% of the authorities on these products. Again, the authority vs. product user dichotomy for body product type does not seem to be changing as far as the male central figure is concerned. The only product type which was not substantiated in my replication was that of food products. McArthur and Table 16. 39 Home Product Type Associated with Authority and Product User Central Figures Sex of Central Figure Male Female product .18 .82 .61 Basis for User (5) (23) (28) Credibility Authority .67 .33 .39 of Central (12) (6) (18) Figure .37 .63 1.00 (17) (29) (46) (x2 = 11.21, p < .001, df = l) Table 17. 40 Body Product Type Associated with Authority and Product User Central Figures Sex of Central Figure Male Female Product .32 .68 .72 User Basis for (11) (23) (34) Credibility of Authority .85 .15 .28 Central (11) (2) (13) Figure .47 .53 1.00 (22) (25) (47) (x2 = 10.30, p < .01, df = 1) 41 Resko reported that men comprised 95% of the authorities on food products, but only 40% of the food product users (x2 = 20.99, p < .001, df = 1). As one can see in Table 18, I failed to find the same results. Males comprised only 48% of the authorities on food products and 52% of the food product users (x2 = .11, .70 < p < .80, df = 1). Of course, there is the possibility that I defined the food category differently. Time of Day The analysis by McArthur and Resko revealed that differences in the presentation of male and female central figures were quite constant over time. That is, whatever differences there were were as likely to occur in the morn— ing and afternoon as in the evening. The authors did not report any statistical results to substantiate their ar- gument. In my analysis, however, there were differences across time: the presentation of male and female central figures differed by time of day (x2 = 6.59, p < .05, df = 2) as shown in Table 19. McArthur and Resko did find, however, that 70% of the central figures in evening hours were male as compared with 52% in the morning and afternoon. The reason for this comparison, the authors argued, was that in the evening males were most likely to be watching television, while in the morning and afternoon most viewers were female. One can see that the differences were most evident when morning Table 18. 42 Food Product Type Associated with Authority and Product User Central Figures Sex of Central Figure Male Female Product .52 .48 .67 User Basis for (26) (24) (50) Credibility Authority .48 .52 .33 of Central (12) (13) (25) Figure .51 .49 1.00 (38) (37) (75) (x2 = .11, .70 < p < .80, df = l) 43 Table 19. Sex of Central Figure by Time of Day Sex of Central Figure Male Female Morning .46 .54 .19 (18) (21) (39) Afternoon .40 .60 .38 Time (32) (48) (80) °f Evening .59 .41 .43 Day (54) (37) (91) .50 .50 1.00 (104) (106) (210) (x2 = 6.59, p < .05, df = 2) 44 and afternoon data were combined and compared with the evening. These results were confirmed in my replication (x2 = 6.19, p < .02, df = 1). Morning and afternoon cen- tral figures were collapsed into one category and compared with data from those commercials viewed in the evening. Table 20 shows that 59% of the central figures in evening hours were male as compared with 42% in the morning and afternoon. On the other hand, females comprised only 41% of the evening central figures and 58% of the morning and afternoon figures. Females differ in the morning and afternoon as compared with the evening. Obviously, women are not seen as often as central figures during those hours when television is most likely to be watched by men. Additional Findings The Number of Central Figures During News Programming McArthur and Resko did not sample television commer- cials during news programming, which usually runs from 6:00 P.M. - 6:30 P.M. locally and from 6:30 P.M. - 7:00 P.M. nationally. As reported earlier, they had found a tendency for central figures to be predominantly male in the evening, which was also confirmed in my replication. However, I had decided to carry this one step further to (see whether or not there was a significant difference in the number of male and female central figures shown during the six o'clock news hour. This seemed to be an important point, especially since most men are home at this time 45 Table 20. Sex of Central Figure by Time of Day - One Degree of Freedom Analysis Sex of Central Figure Male Female Morning .42 .58 ‘ .57 and Afternoon (50) (69) (119) Time Evening .59 .41 .43 of (S4) (37) (91) Day .50 .50 1.00 (104) (106) (210) (x2 = 6.19, p < .02, df = 1) 46 and frequently watch news programs. Therefore, I had ex- pected a preponderance of male central figures during this time. This was not the case, however (x2 = 1.48, .20 < p < .30, df = l). A chi square goodness-of—fit test revealed that 61% (n = 20) of the central figures during the news hour were male as compared with 39% (n = 13) for females. Thus, there were no significant sex differences for those commercials sampled during news programming on Tuesday, Wednesday, and Saturday. The Frequency of Central Figures on Saturday The original study did not sample television commer- cials on the weekend, whereas I had watched programs for a total of seven hours on a Saturday. As noted earlier, I found a tendency for the central figures to be predom- inantly male in the evening. This was also confirmed in my analysis of Saturday evening programming only. A chi square goodness—of-fit test found that there was a signif- icant difference between the number of male and female central figures seen during Saturday evening (x2 = 9.53, p < .01, df = l). I found that 76% of the evening central figures were male (n = 26) as compared with 24% for females (n = 8). Recall I had discovered that 59% of the evening central figures were male and only 41% were female when the entire three day sample was taken into consideration. I expected a greater number of male central figures to be shown on Saturday afternoon, since the World Series 47 game and other sports events were scheduled at this time. The chi square results, unfortunately, did not lend them- selves to significant meaningful comparisons. However, the results led me to believe that there was a significant difference between the number of male and female central figures shown during Saturday afternoon. In my small sample, 68% of the central figures in the afternoon were male, whereas only 32% were female. I also could not analyze those ads televised during Saturday morning cartoon shows. Only three commercials out of the twenty-two surveyed could be coded according to the criteria outlined by McArthur and Resko. The other nineteen had no adult central figures in them. They often included children only, or cartoon and fantasy figures, such as Fred Flintstone, Count Chocula, and Ronald Mc- Donald. The Voice-Over The voice-over, as the term suggests, refers to the voice of the off-camera announcer which often accompanies commercially advertised products. I found that the voice urging the viewer to buy the product was nearly always male; women were rarely used in this regard (x2 = 139.86, p < .001, df = 3). Of the 163 commercials which contained some type of voice-over, 62% were accompanied by a male voice, whereas the female voice was heard in only 2% of the cases (see Table 21). 48 Table 21. Frequency of Occurrence of voice-Over Which Accompanies Commercials Percentage of Total Raw Total Male .62 - 101 Type of Female .02 2 Voice- Chorus .10 17 Over Other .26 43 N = 163 (x2 = 139.86, p < .001, df = 3) DISCUSSION The present investigation of commercial advertising has revealed sex differences which, for the most part, replicate and extend the evidence of sex stereotyping re- ported by McArthur and Resko. One of the most striking differences, however, was that females appeared as fre- quently as males in my ads. Given that women comprise over one-half of our population, one would expect that approximately one-half of the central figures in the media would be women (McArthur and Resko, p. 217). Their spec- ulation was confirmed in my analysis. The number of central figures was almost equally divided between the sexes. This was quite unexpected, since men were found to outnumber women in McArthur's and Resko's analysis, and this finding seemed to serve as a basis for validat- ing their subsequent research. Perhaps the differences in our results could be explained by my clarification of the coding category "central figure." The authors seemed to me to be very unstructured in their use of this cate- gory; I had thought it would be best to code only those figures who were the central focus of the ad. In the future a category entitled "group ads" or "group figures" may be more significant in determining 49 50 role analyses. This may incorporate ”minor" figures as well as "central" figures only, and the former would un- doubtedly have some relevancy to the "group figures" category. An example would be the "I want to give the world a Coke" ad, where numerous faces representing var- ious ethnic groups and nationalities are depicted. It should also be noted that my coding problems could have been minimized if the authors had sent me a copy of their codebook procedures. I had tried repeatedly to obtain such material from them, but to no avail. None- theless, I think my methodological analysis was warranted. There were notable differences in the quality as well as in the quantity of the portrayal of males and fe- males in most of the coding categories, which were con- sistent with McArthur's and Resko's findings. More females are portrayed as authority figures in today's commercials, although this trend is still not as high as that reported for male authority central figures. The male continues" to be the authority figure, although my analysis revealed that this trend seems to be declining, whereas females continue to be portrayed predominantly as product users. It seems to me that women product representatives are most often seen performing domestic tasks involving the adver- tised product. That is, they demonstrate the product fea- tures by simply using the product. Women seem to win the smiles of their husbands, children, and guests with the help of the product. Male product representatives, on the 51 ¢*-A other hand, most often demonstrate but do not actually use I the advertised product while making product feature claims. My role and location analyses also confirmed McArthur's and Resko's findings. The sexes appeared in different roles and locations. The television commer- cials of America have portrayed women in limited or unde- sirable roles, such as young mother, middle-aged house- wife, or old lady. These ads appear to overlook the married or single woman who is intelligent, sensitive, employed, supports herself, has talents and hobbies, or is skilled at her profession. Females in my ads were often presented in terms of their relationship to others; males tended to be portrayed in a role which defined them in- dependently of others. I would say the sexes differed in their behavior as well as in their roles, which, in turn, is related to their credibility base. For instance, males more often manifested expertise concerning the advertised products. While males were thus presented as more know- ledgeable than females, females were more often portrayed in search of knowledge. Perhaps this explains their higher representation among product users in ads. _A§ Lands and Brennan (1974) state, advertisers obviously‘be:_ lieve that men are_E£usted and believed much more often ‘ than women. w The observed sex differences in behavior found in these ads parallel those differences reported by Child, Potter, and Levine (1946) in their study of children's 52 textbooks. Just as female characters in children's books were less likely than the males to be knowledgeable, so were the female figures in television commercials less likely than the males to possess expertise. While data on roles and location indicated what were seen in television commercials, it might be more appropriate to discuss what was not seen. Women lawyers, doctors, business executives, scientists, engineers, and professors were conspicuously absent from these ads. And although there are millions of working wives in this country, a commercial was seldom shown featuring them.) In the world of the television commercial, women are house-1 wives or low-level employees. Seldom do they combine em- ployment with management of their homes and personal lives. When location is considered, the female is oftentimes pictured in the home, helping to sell some product found in the kitchen or bathroom. Those women who are shown away from home relate to people in a service role, either as a stewardess on an airplane or as a secretary in an office. And occasionally, an attractive model is seen advising other women how they too can look beautiful. These results confirm those reported by Coleman (1971) and Courtney and Whipple (1974) concerning work roles and location of product usage, respectively. Women were more likely to be featured in the home, displaying products often used in the kitchen or bathroom. Occupations are sex stereotyped as well in these 53 ads, which parallels those results found by Head (1954), DeFleur (1970) and Coleman (1971). This implies that the occupations which are sex-typed "female" not only are con- x sistent with_the culture's stereotypes toward women,_but they are occupations that are unlikely to lead to rights to distribute’ggarce_goods and resources; Women were morehlikely\to—be’portrayed as consumers than producers, Their roles as housewife and primary purchaser override those of authority and producer, the latter most often reserved for the male figure. These findings suggest that males in these ads have more varied ”worldly" roles, emphasizing their importance or dominance in the sphere of employment or occupation, while females are circumscribed by their domesticity, passivity, submissiveness, or sexuality, and overrepre- sented in family or home occupations. It seems to me that if these results are viewed from the perspective of some- one in the feminist movement, their criticisms are well taken. The image of the female, as shown in these ads, is in line with conventional stereotypes. These nations are supported by Duberman (1975) when she noted that oc- cupations are linked to sex roles. Women are not portrayed up” as autonomous human beings, but are primarily sex-typed. Commercials presenting the image of the "modern" woman are virtually nonexistent; they reinforce instead the assump- tion that a woman's only valid function is that of wife, __,._—- mother, or servant 5f men. Thus, occupational opportunities 54 should be available for each sex according to interest and talent. The sex labeling of occupations has restricted the creative contributions of many talented individuals. In addition to differences in the quantity and qual- ity of their behavior, male and female central figures in the original study differed in the consequences of their behavior, but I failed to find similar results except for rewards reaped by product user central figures. The dif- ferences between their results and mine may be explained, in part, by the strong specifications I required for the category "scientific argument" in my coding procedures. They were probably not as stringent in their analyses. Perhaps this was a grave error on my part, but, of course, it would have been helpful if they had provided some con- crete examples of their coding format. Furthermore, their reluctance to outline coding categories more succinctly may have been the reason for some of my insignificant tabular results, especially in regard to the reward cate- gories. As far as product types associated with male and female central figures were concerned, the only differ- ence between their results and mine was in the category "food product." McArthur and Resko found a consistent tendency for males to appear as authorities on food pro- ducts which were used primarily by females. My replica- tion, however, represented a definite reversal trend. Males may not always be the authorities on a product which 55 is used primarily by females. Perhaps the roles of males and females in television commercials are changing in re- gard to the relationship between the sex of the characters in the commercials and the product category advertised, as my results for food products suggest. Nonetheless, females in today's ads are most often shown representing cleaning products, cosmetics, and home appliances, whereas Hales are most often shown repre- ‘ senting cars, travel, banks, and industrial products. The lack of women in the latter supports the stereotype that a woman has no interest or ability in so masculine an area as mechanics or machines, as noted by Dominick and Ranch (1972). She is still primarily shown in the kitchen or bathroom.of the home serving her husband or children. Finally, my replication of McArthur's and Resko's study of commercials by time of day revealed that there are differences across time in the presentation of male and female central figures. The authors did not find such results, but failed to provide any specific figures for comparative purposes. Concerning the additional analysis I did, I found no significant sex differences for central figures por- trayed during news programming. This was quite unexpected, since I had predicted a preponderance of male central fig- ures during the news hour when men are most likely to be home from work. Perhaps my results are a reflection of the small sample size as well. 56 Unrelated to this but also involving the news is the problem with products advertised during the news broad- cast, especially the local news. Oftentimes, various por- tions of the news (sports or weather, for instance) have only one sponsor, and this could result in a select group of products advertised. This is indirectly related to the simple fact that there are two kinds of ads which can be seen throughout the day: national and local. Local res- taurant and furniture ads are but one example of this. In looking at Saturday programming, the results led me to believe that male and female central figures dif- fered by time of day. As was found during evening program- ming on a weekday, more males appeared in television ads during Saturday evening than females. Perhaps television sponsors believe that the number of male viewers is higher during the evening hours, especially on weekends when men do not ordinarily work. But this does not explain why there was not a preponderance of male central figures during news programming when men are just as likely to be home. Again, this may reflect television sponsorship, or it may be that commercials during these times are dir- ecting their efforts toward "family" matters rather than specific male or female products. The number of furniture ads shown during the news hour is a case in point. This is the type of ad which is likely to contain as many male as female central figures, a fact which I noted while cod- ing. 57 This notion concerning the time of day when men and women are most likely to be home somewhat parallels the results of Hennessee's and Nicholson's (1972) study when they found that the life style and problems of working women are virtually ignored in commercials. They found that ads for cars, banks, and insurance accounted for 3.2% of the daytime commercials, as opposed to 19.1% in prime time when men are most likely to be at home. These sex- segregated ads seem to be saying that women are incapable of making important decisions alone. Recall that some insignificant results were obtained due to the small expected cell frequencies, especially dur- ing Saturday cartoon programming. Perhaps my findings in- dicate that the commercially produced television ads viewed by children carry different messages about the appro- priate behavior for males and females. I would say that males and females appear in different prOportions and roles during Saturday morning shows. This suggests that for today's children, television ads may be an important source in the learning of stereotyped sex roles. Accord- ing to Liebert (1973), they may be an important source of children's expectations and prejudices. The voice-over accompanying many television commer- cials seemed to substantiate many of the results obtained by statistical analysis. For instance, my findings sug- gest that females lack credibility. Their small voices are often followed by a male voice-over, the voice of 58 authority, which confers the stamp of approval on the pro- duct. The male voice urging the viewer to buy the pro- duct is usually gentle, wise, helpful, and a bit seduc- tive. It seems to me that male authority is a built-in assumption, and it teaches women to look up to men as ex- perts. Female voice-overs, on the other hand, were most likely to occur during cartoon shows when dolls were the advertised product. A soft, soothing, "motherly" voice was evident here, offering comfort and reassurance. Thus, my results are consistent with Lands' and Brennan's (1974) study which found that females were overwhelmingly characterized as "soft spoken." These results are also similar to those concerning stereotyped images of women as reported by Courtney and Whipple (1974). They found that men are overwhelmingly present as the voice-over in television commercials, ac- counting for 89% of the total. This male presence was found to be as dominant in daytime commercials, presum- ably addressed primarily to women, as it was in evening commercial programming. It is interesting to note that oftentimes it was the voice-over, rather than an on-camera central figure, which gave factual, concrete, evidence in favor of using the product (scientific argument) or served as the reward or credibility base. Note the following scientific argu- ments given by voice overs: 59 Aspirin is what pediatricians recommend almost two to one over any other pain or fever reducer. Rolaids consumes forty-seven times its weight in excess stomach acid. This may explain the lack of scientific arguments given by central figures, since voice-overs frequently served in this capacity. One implication of the present findings is that if one wishes to diminish the sex stereotyped behavior which /,1 - r_ is so prevalent in our society, a change in the repre- sentation of males and females in television commercials --. .H—b-tv— is a useful step forward. This study also suggests that appropriate changes in television's portrayal of the sexes could serve to increase socially desirable, nonstereotyped behaviors on the part of both sexes. In the future studies could analyze television com- mercials by voice-over only to see if it continues to be the voice of authority. An analysis conducted later in the evening could be undertaken as well to determine whether or not such sex differences are as evident in these later time periods. The sample of programming and coding categories could also be broadened and defined more suc— cinctly to alleviate insignificant results. APPENDIX APPENDIX LISTING OF CODING CATEGORIES AND THEIR COMPONENTS Central Figure - Those adults appearing in the ad who were the central focus by virtue of either speaking or having prominent visual exposure. Credibility - Suggests plausibility, but more clearly stresses worthiness of belief. Credibility of cen- tral figure was categorized as product user when depicted as a user of the product being advertised; the basis for credibility was categorized as author- ity when depicted primarily as someone who "has all the facts" about the advertised product. Role - Refers to the manner in which the central figure was cast. Category includes: Relational - spouse, parent, girlfriend/boyfriend, housewife, and other (unknown) Independent - worker, professional, celebrity, in- terviewer/narrator, and other (unknown) Location - Locale in which the central figure was depicted. Category includes: Private - home or yard Semi-private - office, place of work,or occupational setting Public - store, restaurant, or park Other - (unknown) Argument - A coherent series of reasons offered by a cen- tral figure on behalf of a product. 60 61 Category includes: Scientific - actual facts, figures, statistical re- sults, lists, or percentages presented by the central figure Nonscientific - opinions and personal testimonials in favor of using the product No Argument - central figure merely displayed a product or was being persuaded by another to use it Reward - That which is given, offered, or received for some service or attainment. Categorygincludes: Social enhancement - opposite sex approval, family approval, friend's approval, social enhancement, career advancement, and other or unknown Self-enhancement - psychological improvement, attrac- tiveness, cleanliness, health, and other or un- known Practical reward - saving time, labor, or money Other - feels nice, good-tasting, comfortable, ser- vice, dependability, or unknown Type of Product - Product associated with central figure.- Categogy includes: Body product - appearance aids, body hygiene-clean- liness products, clothing, and health products Home product - exterior and interior household goods, household cleaners, and laundry and dish detergents Foodstuffs - includes beverages as well, such as tea, coffee, and soft drinks Other - pet food products, sporting and recreational items, automobiles and automotive products, in- surance, services, and other BIBLIOGRAPHY BIBLIOGRAPHY Allport, Gordon W., The Naturgjof Prejudice, Garden City, New York, Doubleday andICompany, Inc., 1958. Chafetz, Janet Saltzman, Masculine/Feminine or Human? An Overview of the Sociology of Sex RoIes, Itasca, III- inois, F. E. Peacock Publishers, Inc., 1974. Child, Irvin L., Potter, Elmer H., and Estelle M. Levine, "Children's Textbooks and Personality Development: An Exploration in the Social Psychology of Educa— tion," Psychological Monographs, 60:1-54, 1946. Colemen, Janet, "Advertising," You: The Magazine for the New WOman, July, 1971. Courtney, Alice E., and Thomas W. Whipple, "WOmen in TV Commercials," Journal of Communication, 24:110-18, 1974. DeFleur, Melvin L., "Occupational Roles as Portrayed on Television,“ Public Opinion Quarterly, 28:57-74, 1964. Dominick, Joseph R., and Gail E. Rauch, "The Image of Women in Network TV Commercials,” Journal of Broad- Duberman, Lucile, Gender and Sex in Society, New York, Praeger Publishers, 1975. Head, Sydney W., "Content Analysis of Television Drama Programs," Film Quarterly, 9:175-94, 1954. Hennessee, Judith Adler, and Joan Nicholson, "N. O. W. Says: TV Commercials Insult Women," The New York Times Magazine, pp. 12-13+, May 28, 1972. Liebert, Robert M., The Early Window: Effects of Television on Children and Youth, New York, Pergamon Press, Inc., 1973. Maccoby, Eleanor E., and William Cody Wilson, "Identifica- tion and Observational Learning from Films," Journal of Abnormal and Social Psychology) 55:76-87, 1957. 62 63 Maccoby, Eleanor E., Wilson, William C., and Reger V. Burton, "Differential Movie-Viewing Behavior of Male and Female Viewers," Journal of Personality, 26:259-67, 1958. McArthur, Leslie Zebrowitz, and Beth Gabrielle Resko, "The Portrayal of Men and WOmen in American Tele- vision Commercials," The Journal of Social Psychology, 97:209-220, 1975. Sternglanz, Sarah H., and Lisa A. Serbin, "Sex Role Ster- eotyping in Children's Television Programs," 23- velopmental Psychology, 10:710-15, 1974. ICHIGRN STRTE UNIV. LIBRQRIES lllllllllllllllllHIHlllllllllllllllllllllllllllllllllllHll 31293100635378