Journal of Advertising History History of Advertising Trust Unit 6, The Raveningham Centre, Raveningham, Norwich NR14 6NU Tel: +44 (0) 1508 548 623, Fax: +44 (0) 1508 548 478

Special Issue, 2002

Stereotypes of the Elderly in US Television Commercials From the 1950s to the 1990s

Darryl W. Miller, Ph.D.
University of Wisconsin, River Falls
Teresita S. Leyell, Ph.D.
Washburn University
and
Juliann Mazachek, Ph.D.
Washburn University

Critics have charged that American advertisers have often portrayed the elderly through negative stereotypes. These negative portrayals, they suggest, not only offend elderly consumers but also contribute to ageism. This study examined whether American advertisers have indeed used a great deal of negative stereotyping of the elderly as the critics have suggested. Employing a concept of stereotypes found in cognitive psychology, the authors examined trends in the portrayal of the elderly in television commercials produced in the United States from the 1950s to the 1990s. Results of the study do not support the contentions of the critics. Very little negative stereotyping was discovered. Analysis indicated trends in the appearance of several positive stereotypes and differences in the stereotyping of gender groups and age segments within the elderly group. Results are interpreted from both marketing and social science perspectives.

Critics of advertising have suggested that in television commercials in the United States, elderly people are most often portrayed using negative stereotypes. That is, they are shown as 'half–dead codgers, meddling biddies, grandfatherly authority figures or nostalgic endorsers of products that claim to be just as high–quality as they were in the good old days' (Abrams 1981, p. 27), and that too many commercials use stereotypes of the elderly, depicting them as 'feeble, foolish or inept, passing their time aimlessly in rocking chairs' (Goldman 1993, p. B4). Social psychologists suggest that such advertising reinforces these stereotypes, and thus dehumanises interpersonal relations and aggravates ageism (Pollay 1986; Palmore 1990). Some experts suggest that this negative stereotyping can damage the self–concepts of the elderly, and disrupt the socialisation of young people with respect to the elderly (Smith et al. 1984; Palmore 1990). The latter effects are particularly noticeable when television becomes children's primary source of contact with information about the elderly (Davis & Davis 1985).

Others have suggested that advertising does not have the power to alter or shape social values but merely reflects the values of its target audiences (Holbrook 1987). This suggests that if, as research indicates, attitudes towards the elderly in US society have become more positive over the past few decades (Kite & Johnson 1988; Palmore 1990), advertising images of the elderly should have also become more positive overall. Further, one might expect that many of the images of the elderly in advertising would be a function of marketing strategies designed to meet the needs of increasingly affluent seniors. Considering the increasing size and economic power of elderly market segments, it would seem unwise for advertisers to insult the elderly with negative stereotypes. One could suppose, however, that advertising messages employing humorous appeals targeted at young people might occasionally employ negative portrayals of older people so as to amuse the target audiences.

Are the contentions of the critics correct? If advertisers have often used negative stereotypes of the elderly and these images have the negative social consequences cited above, one might suggest that advertisers have been socially irresponsible. However, such a serious charge should be supported by evidence. Further, has there been a trend from less negative to more positive portrayals consistent with a shift towards more favourable attitudes towards the elderly and the increasing size of the elderly population? The purpose of this study is to shed some light on these issues by examining the stereotypes of the elderly used in television commercials that were broadcast in decades from the 1950s to the 1990s. Using a formal theory of stereotyping drawn from cognitive psychology, the study analyses the specific stereotypes that have appeared and the trends in their appearance. A convenience sample consisting of commercials drawn from several different sources was employed for this study. Portrayals of the elderly appearing in commercials were assessed for their consistency with stereotypes discovered by Hummert (1990, 1994) and her colleagues. Television is the chosen medium for the study for two reasons. First, among advertising media, it is typically recognised as having the most impact (cf. Shimp 2000). Second, the literature contains no studies involving longitudinal examinations of stereotypes of the elderly in the medium.

The discussion begins with an explanation of the theoretical approach taken for this study. Next, past research on the portrayal of the elderly in television advertising is examined. Finally, the results of the study are presented and discussed.

From the perspective of cognitive psychology, stereotypes are defined as knowledge structures (schemas) in long–term memory that contain an individual's beliefs about a particular concept (Ashmore & Del Boca 1981), in the current case about elderly people. These structures are organised in a hierarchical order. That is, each individual has a schema containing beliefs about the concept of 'elderly adult' which comprises several distinct but related sub–schemas. These sub–schemas, which consist of sets of traits that describe types of elderly people, are specific stereotypes (Hummert 1990; Hummert et al. 1994). Stereotypes of the elderly can be either positive or negative. Positive stereotypes represent sanitised and idealised images, while negative stereotypes produce demeaning and ridiculing portraits (Langmeyer 1984). Stereotypes are rooted in cultural beliefs and are produced by social interaction, but exist only in each individual's cognitive representations. Many people might hold the same stereotype at a conceptual level, sharing a set of traits generally, but their individual sets of traits may not match exactly (Hummert 1990; Hummert et al. 1994).

In earlier advertising studies, stereotypes of the elderly in advertising had to be inferred from the coding of variables such as the setting of the scenario and the role played by the elderly individual. The cognitive psychology approach is advantageous because it allows a more explicit assessment of specific stereotypes that are actually widely held by individuals. Hummert et al. (1994), building on earlier research (Schmidt & Boland 1986; Hummert 1990), showed that people do indeed hold multiple stereotypes of the elderly. Using card–sorting and cluster analysis techniques, they found six positive and eight negative stereotypes (see Table 1). The positive stereotypes were the:

1. Perfect grandparent;
2. Golden ager
3. John Wayne conservative
4. Liberal matriarch/patriarch
5. Activist
6. Small–town neighbour.

The negative stereotypes were the:

1. Despondent
2. Vulnerable
3. Severely impaired
4. Shrew/curmudgeon
5. Recluse
6. Mildly impaired
7. Self–centred
8. Elitist

TABLE 1: STEREOTYPE OF THE ELDERLY REPORTED BY HUMMEL ET AL. (1994) AND THE CORRESPONDING TRAITS

This stream of research does not present clear evidence of whether people associate either positive or negative stereotypes as being more typical of the elderly. Hummert (1990, 1993) found that both young and elderly subjects rated negative stereotypes as less typical of elderly than they did positive stereotypes. However, in a later study, Hummert et al. (1995) found no significant difference in typicality ratings between positive and negative stereotypes made by young adult, middle–aged and elderly subjects.

These studies do reveal that negative stereotypes are associated more with older while positive stereotypes are associated more with younger elderly groups. Hummert and her colleagues (1993, 1995) found that respondents increasingly associated 'Mildly impaired', 'Severely impaired', 'Shrew/curmudgeon', 'Despondent', 'Recluse' and 'Vulnerable with greater age'. Positive stereotypes such as 'Golden ager', 'Perfect grandparent', 'Activist' and 'Liberal matriarch/patriarch' were most associated with the young elderly. 'John Wayne conservative' and 'Small–town neighbour' were most associated with the middle elderly (ages 70 to 79).

With respect to differences in stereotyping of elderly gender groups, Hummert et al. (1997) found that subjects associated fewer positive stereotypes (i.e. 'Golden ager', 'Perfect grandparent', 'John Wayne conservative') with photos of young and middle elderly women than they did with photos of young and middle elderly men. However, for the old elderly gender groups the pattern was reversed. Thus, it appears that Americans may hold elderly men (except for the very old) in higher esteem than they do elderly women.

Past research on the portrayal of the elderly in TV advertising in the US several studies have examined how the elderly have been portrayed in television commercials. Table 2 contains a summary of these studies. A chronological examination of this literature begins with Francher (1973), who examined 100 commercials that appeared in the early 1970s. He concluded that the elderly often appeared in commercials involving humorous appeals and were often depicted as blatantly comical.

TABLE 2: PAST RESEARCH ON THE PORTRAYALS OF THE ELDERLY IN TELEVISION COMMERCIALS IN THE US

A few years later, Harris and Feinberg (1977) conducted a content analysis of 80 commercials that aired in 1976. Although they found no portrayals of debilitated or incapacitated elderly, they did find that the elderly were depicted as having considerably more physical ailments than were younger people. They also found that the proportion of women cast as advice–givers declined much more with age than it did with men. They concluded that, overall, the elderly were presented in commercials in 'unflattering…unhealthy, unstylish, and uninteresting ways' (p. 467). Further, they concluded that portrayals of elderly women were especially harsh compared with elderly men, reflecting a decline in esteem for women with age.

Langmeyer's (1984) findings contrast somewhat with those cited above. She examined elderly individuals' perceptions of the role portrayals of the elderly in prime–time commercials. Using an adjective checklist of both positive and negative traits, 54 elderly subjects rated the elderly characters that appeared in 17 commercials. According to the author, the subjects generally saw the elderly characters as neutral, having neither negative nor positive traits. Females were sometimes rated as motherly, and males were often described as authoritative and dignified. Both genders were rated as friendly, serious and having positive attitudes. Very few of the respondents rated any of the portrayals negatively (i.e. as incompetent, comic, stupid, unfriendly, ill, etc.). Subjects' ratings indicated mild disagreement with the statement that elderly men and women portrayed in the commercials were unattractive. Further, these ratings showed no significant difference between the gender groups.

Three years later, Swayne and Greco (1987) reported their content analysis of commercials broadcast on three major television networks in 1986. They were interested in the prominence and nature of the roles played by the elderly. Their results indicated that most of the elderly depicted played minor roles in the commercials. The researchers found very few negative portrayals. Only 13% of the elderly people were depicted as comical or humorous and only 6.5% were depicted as confused or feeble. On the other hand, 65% of the elderly were portrayed as advisers, which, the authors suggested, is consistent with several positive stereotypes.

Later, Robinson et al. (1995) sought to replicate the Swayne and Greco study. They examined more than 800 commercials which appeared within prime–time network programming in 1994. Consistent with previous results, they found that most elderly people were cast in minor roles. As before, very few negative portrayals were found. Only 13% were portrayed as comical or humorous, and only 4% as feeble or confused. A much lower percentage of elderly (30%) were cast as advisers than was the case in the previous study.

Finally, Peterson & Ross (1997) examined 1,874 commercials broadcast on network, local and cable television in 1991. They were interested in determining whether older people were portrayed in desirable or undesirable ways depending upon the apparent target audience of the commercials. 'Desirable' portrayals were defined as those in which elderly models display mental and physical competence. They defined 'undesirable' portrayals as those in which the elderly models exhibit mental or physical incompetence exemplified by appearing to be impaired, helpless, uninformed, weak or lazy.

Overall, Peterson & Ross found that elderly people (aged 65 and older) were portrayed in desirable ways. However, the proportion of desirable and undesirable portrayals varied by the target audience of the commercials. The highest proportion of desirable portrayals (71.5%) and lowest proportion of undesirable portrayals (28.5%) appeared in commercials targeted at people aged 45 and older. Commercials not targeted at any specific age group contained the lowest proportion of desirable portrayals (54%) and the highest proportion of undesirable portrayals (46%). Commercials targeted at younger people contained 62.7% desirable and 37.3% undesirable portrayals. Even though in each case the majority of portrayals are desirable, the proportion of undesirable portrayals is rather high. They also found that the proportions of desirable portrayals of elderly people were less than those for people aged 46 to 64, and especially for those aged 45 and under. The proportions of undesirable portrayals increased with age.

The studies conducted in the 1970s present a very negative view of the portrayals of the elderly in television commercials. Overall, later studies present a much less negative view. This may reflect a trend towards less negative portrayals. Still, it remains difficult to draw meaningful conclusions regarding trends in the stereotyping of the elderly in television commercials from these studies because a variety of empirical approaches were used and none employed a longitudinal design or a formal theory of stereotyping. To better examine trends in the stereotyping of the elderly in advertising, a firm theoretical basis and a longitudinal sample is required.

Such a study, involving stereotypes of the elderly in print ads, has recently appeared. Miller et al. (1999) examined stereotypes of the elderly in advertising based upon the findings of Hummert and her colleagues. They examined 1,944 ads in issues of Life, Better Homes and Gardens and Popular Mechanics published from 1956 to 1996. Very little negative stereotyping of the elderly was discovered. Of the ads that included elderly people, 1.6% contained the 'Shrew/curmudgeon' and 5.7% contained the 'Mildly impaired'. Positive stereotypes more commonly appeared in these ads. Fourteen per cent included the 'Perfect grandparent' and 9.3% contained the 'John Wayne conservative'. By far the most prevalent stereotype was the 'Golden ager', which appeared in 37% of the ads. Trend analysis revealed statistically significant increases in the proportion of the 'Mildly impaired' and the 'Perfect grandparent'. Analysis also revealed that the 'Mildly impaired', 'Perfect grandparent' and 'John Wayne conservative' were associated more with the older elderly than they were with the younger elderly. The 'Golden ager' was associated more with the younger elderly than it was with the older elderly.

As its authors suggest, the results of this study seem reasonable if viewed from a marketing perspective. The increasing trend in the appearance of the 'Mildly impaired' stereotype could relate to marketers promoting products designed to alleviate some of the physiological problems associated with ageing to a growing elderly market. Further, it seems reasonable to suggest that marketers would be reluctant to offend elderly consumers with portrayals of negative stereotypes. This might explain why very few examples of the 'Severely impaired', 'Shrew/curmudgeon', and 'Despondent' were found.

What, then, given this range of opinion and empirical results, would one expect to find in an analysis of stereotypes appearing in television commercials over the past few decades? If one believes the critics of advertising, might one expect to find a great deal of negative stereotyping? However, if one assumes that advertising images reflect social reality and/or marketing strategies, might one expect to find mostly positive stereotyping and an increasing (decreasing) trend in the proportion of positive (negative) stereotypes? Also, would one expect that negative (positive) stereotypes have been associated more (less) with the old elderly than with the young elderly and more (less) with women than with men? Finally, would the pattern of stereotypes of the elderly appearing in a longitudinal sample of television commercials be consistent with the findings of the Miller et al. (1999) print ad study?

Obtaining a random sample of television commercials broadcast over the past five decades would be extremely difficult if not virtually impossible, especially for commercials from earlier decades. Archives of these commercials are few, incomplete, variously organised and widely scattered about the USA. Given these difficulties, a convenience sample was employed for this study. Researchers tried to draw commercials from as many different archives and from as many different product categories as possible in an attempt to increase the quality of the sample.

Commercials were drawn from several sources. One was the John W. Hartman Center for Sales, Advertising and Marketing History at Duke University. The Hartman Center includes: (1) the Video Resources Collection containing commercials from the 1950s and 1960s; (2) a collection of commercials from the 1980s and early 1990s donated by USA Today; and (3) the Wayne P. Ellis Collection of Eastman Kodak commercials, which were produced from the 1950s to the 1980s. Another source of commercials was the State Historical Society of Wisconsin Visual Material Archive. The archive includes: (1) the Television Bureau of Advertising Collection, consisting of commercials from the 1950s to the early 1990s; (2) the National Association of Broadcasters Collection, which includes commercials from the 1960s and 1970s; (3) a collection donated by WISC–TV, consisting of commercials from the late 1950s to the early 1970s; and (4) the Mitchell Collection, which includes commercials from the early 1950s to the early 1970s.

A third source consisted of commercials appearing within 15 hours of network programming recorded in the 1980s. The programming was recorded for personal entertainment purposes and the researchers had no prior knowledge of the commercials that were captured on these tapes. The researchers also included commercials contained in 36 hours of network programming broadcast during September 1998. Six hours of programming appeared in the 6 a.m. to noon, noon to 6 p.m. and 6 p.m. to midnight segments, respectively. Each was recorded on a different day of the week. The overall sample (Table 3) consisted of 1,662 commercials stratified by decade.

TABLE 3: SAMPLE OF TELEVISION COMMERCIALS BY DECADE

The three authors coded each commercial independently. The first task was to determine whether a commercial contained people. There was complete agreement on this task and commercials without people underwent no further analysis. The next task was to determine whether each of the remaining 1,454 commercials included at least one elderly person as a central character. Following Hummert et al. (1994, 1995, 1997), elderly was defined as 60 years of age or older. As in previous studies (Swayne & Greco 1987; Robinson et al. 1995), people were classified as elderly based on the appearance of factors such as grey hair, wrinkled faces and hands, retirement scenarios, ambulatory aids such as canes, walkers and wheelchairs, presence of middle–aged children and/or presence of grandchildren.

A central character was defined as one who spoke within the commercial, appeared within several scenes, and/or was directly involved in the depicted scenario. Because of the difficulty in coding them, background characters and images that appeared in the commercial for brief flashes were excluded from the analysis. For this coding task there was a 98% proportion of interjudge agreement (86 disagreements from 4,362 total judgements), rendering a proportional reduction of loss (PRL) reliability coefficient (Rust & Cooil 1994) of 1.00. Commercials that did not contain elderly people as central characters were coded no further. A total of 69 commercials contained portrayals of elderly people as central characters. The decade of the commercial and product category were also recorded.

From this point, the unit of analysis switched from a commercial to the elderly persons depicted in a commercial. A total of 101 portrayals of elderly people who appeared in 69 commercials were coded further. The gender and age category, i.e. young elderly (aged 60–74) or old elderly (aged 75 and over) were coded. There was complete agreement among the judges on the gender–coding task. For the elderly age category task, the proportion of interjudge agreement was 85% (44 disagreements from 303 total judgements). This rendered a PRL reliability coefficient of 0.96. Table 4 shows the distribution of product categories across the 69 commercials.

TABLE 4: SAMPLE OF COMMERCIALS BY PRODUCT CATEGORY

Next, the depiction of each elderly person was coded as to whether it exhibited each of 97 traits used in the Hummert et al. (1994) study. Elements of each portrayal were examined to determine their consistency with the various traits. These included the nature of the elderly person's physical appearance (e.g. healthy–looking vs. not healthy–looking), apparent emotional state (e.g. happy vs. sad and lonely), activity depicted (active vs. sedentary) and the nature of the social situation (e.g. alone vs. with people such as family, friends and/or co–workers). In addition, the audio content of each commercial, including what was said by, to and/or about each of the elderly, aided the coding of traits. Judges were allowed to be inductive if necessary. For example, an AT&T commercial from 1992 depicts an elderly man calling an elderly close friend to boast about a large fish that he had caught that day. During their conversation, the two engage in some good–natured needling of one another. Based upon this scenario, one might judge that the fisherman and his friend are fun–loving, happy, lively and sociable.

Data analysis and reliability

The score for each trait (1 = consistent; 0 = not consistent) was summated across the individual ratings of three judges. The next step in the analysis was to conduct factor analysis on the summated traits to determine which stereotypes would emerge. However, before factor analyses could be conducted, the number of trait scores had to be reduced by about half. This would render a 2:1 ratio of observations to variables – a minimally acceptable ratio for conducting factor analysis (Hair et al. 1992). Accordingly, the means of positive and negative traits were listed separately and rank ordered. Next, the lower half was deleted from each list. A series of varimax factor analyses designating from three to twelve factor solutions was conducted on the remaining 51 summated trait variables. The seven–factor solution most closely reflected the stereotype clusters of Hummert et al. (1994). The resulting factors (Table 5) were quite similar to the 'Despondent', 'Mildly impaired', 'Shrew/curmudgeon', 'Perfect grandparent', 'John Wayne conservative' and 'Golden ager' stereotypes. The traits that comprised the 'Golden ager' stereotype in the results of Hummert et al. (1994) were split between two factors. These stereotypes were relabelled 'Adventurous golden ager' and 'Productive golden ager' in this study.

TABLE 5: STEREOTYPES OF THE ELDERLY SCALES

A measure of each stereotype was created by summating the 44 trait scores with positive loadings of 0.40 or higher within each factor (Gorsuch 1974). Reliability coefficients for the resulting scales were: 'Despondent' (= 0.91), 'Mildly impaired' (= 0.85), 'Shrew/curmudgeon' (= 0.74), 'Perfect grandparent' (= 0.88), 'John Wayne conservative' (= 0.76), 'Adventurous golden ager (= 0.85) and 'Productive golden ager' (= 0.82). Next, each scale was divided into three equal segments. Observations whose scores fell in the highest range were categorised as consistent with the respective stereotype. Those whose scores fell in the middle range were categorised as somewhat consistent, and those whose scores fell in the lowest range were categorised as not consistent with the corresponding stereotype.

Table 6 shows frequencies and percentages for the appearance of portrayals consistent with and somewhat consistent with each of the seven stereotypes across decades. Of the 101 elderly people depicted, 79 (78.2%) were at least somewhat consistent (i.e. the total frequency of consistent and somewhat consistent) with a positive stereotype (i.e. 'Perfect grandparent', 'John Wayne conservative', 'Adventurous golden ager' and/or 'Productive golden ager'). On the other hand, only 12 (11.9%) were at least somewhat consistent with a negative stereotype (i.e. 'Despondent', 'Mildly impaired' and/or 'Shrew/curmudgeon'. Six (5.9%) elderly were at least somewhat consistent with both a positive and a negative stereotype. The stereotype in each of these cases was the 'Mildly impaired', which appeared simultaneously with the 'John Wayne conservative' in five cases and with the 'Perfect grandparent' in one case.

TABLE 6: FREQUENCIES, PERCENTAGES AND TREND COEFFICIENTS FOR STEREOTYPES OF THE ELDERLY (TOTAL SAMPLE)

One can see in Table 6 that the most commonly depicted stereotype was the 'Adventurous golden ager'. A commercial for T. Rowe Price Brokerage Service from the 1990s presents an example. The commercial showed an energetic and pleasant elderly woman who, because of her solid financial position, was able to retire and volunteer her time hiking and camping with children. Thirty–three portrayals were somewhat consistent and eight were consistent with this stereotype. Thus, 41 (40.6%) were at least somewhat consistent with the 'Adventurous golden ager'. The next most commonly depicted stereotype was the 'Perfect grandparent'. A typical 'Perfect grandparent' depiction appeared in a commercial for Kodak film from the early 1970s. The scene showed an elderly man happily playing with his grandson in a park. Of the 101 elderly people depicted, 23 were somewhat consistent and ten were consistent with this stereotype. Thus, 32.7% of the elderly depicted were at least somewhat consistent with the 'Perfect grandparent'.

The third most commonly depicted stereotype was the 'Productive golden ager', an example of which appeared in a McDonald's commercial from the 1980s. The commercial portrayed an elderly man coming out of retirement to work part–time at a McDonald's restaurant. At first, the teenagers working there appeared to be quite sceptical about the prospect of having the elderly man join the crew, but his diligence and good humour won them over. Twenty–four of the depictions were somewhat consistent and seven were consistent with this stereotype (30.7% at least somewhat consistent). The fourth most commonly used stereotype was the 'John Wayne conservative'. Sixteen images were somewhat consistent and four were consistent (19.8% in total) with this stereotype. One, a commercial for V–8 vegetable juice, showed a member of the American Legion in his cap and vest sampling a cup of the product.

Much less commonly depicted was the 'Mildly impaired' (six somewhat consistent (5.9%)). This stereotype appeared in an AT&T commercial from the early 1990s. The commercial showed an unhappy and apparently unhealthy elderly man sitting on a bench near a garden. His son tried to humour him and persuade him to get up and move around, but the old man refused. Portrayals somewhat consistent with the 'Despondent' appeared four times (4.0%). One, a commercial for Western Bank from the 1970s, showed an unhappy elderly woman sitting in a rocking chair upon a porch of a ramshackle old house. The voice–over described the woman as poverty–stricken, desperate and unhappy. Finally, two images somewhat consistent (2.0%) with the 'Shrew/curmudgeon' appeared. One, in a commercial for MJB coffee from the 1960s, portrayed a rather snobbish and critical elderly woman attending a party. Her behaviour and demeanour indicated that she disapproved of the hostess, her party and the coffee. She did approve of the coffee, of course, once she had tasted it. There were a total of 12 images (11.9%) at least somewhat consistent with a negative stereotype.

Several trends in the appearance of the stereotypes from the 1950s to the 1990s are indicated. Spearman's rank (), a nonparametric correlation coefficient, may be used as a test of trend (Gibbons 1985). Analysis (Table 6) indicated a marginally significant increasing trend in the appearance of two stereotypes: the 'Adventurous golden ager' (= 0.80, p ≤ 0.10) and the 'John Wayne conservative' (= 0.80, p ≤ 0.10). In addition, analysis indicated a marginally significant decrease in the appearance of the 'Perfect grandparent' (= –0.80, p ≤ 0.10). The trend coefficient for 'Productive golden ager' is positive (0.40) but not significant. Trend analysis was not conducted on 'Despondent', 'Shrew/curmudgeon' and 'Mildly impaired' since there were too few cases to interpret the results confidently. However, no trend emerged (0.00) when the negative stereotypes were combined into a single category.

Next, the focus turns towards whether there have been differences in the stereotyping of the young–elderly and old–elderly. The task of coding each image as either young–elderly or old–elderly produced 22 disagreements among judges. These discrepancies were resolved by coding each portrayal consistent with the judgement of the two judges in the majority. The authors believed that it was better to accept a lower level of reliability than to discard one–fifth of the observations for this phase of the analysis.

Results (Table 7) indicate that, compared with the old–elderly, the young–elderly were less likely to be depicted as the 'John Wayne conservative' (Z = –3.38, p ≤ 0.01) and with a negative stereotype (Z = –2.11, p ≤ 0.05). Alternatively, the young–elderly were more likely to be depicted as the 'Productive golden ager' (Z = 1.96, p ≤ 0.05) than the old–elderly.

TABLE 7: FREQUENCIES, PERCENTAGES, TREND COEFFICIENTS AND DIFFERENCES BETWEEN STEREOTYPES OF ELDERLY AGE SEGMENTS

Young Elderly                                             Old Elderly

Trend analysis (Table 7) for the young–elderly revealed marginally significant increasing trends in the appearance of the 'Adventurous golden ager' (= 0.80, p ≤ 0.10) and the 'Productive golden ager' (= 0.80, p ≤ 0.10). A negative trend coefficient (0.40) for 'Perfect grandparent' emerged within the young–elderly group but was not significant. There were too few cases of 'John Wayne conservative' and the various negative stereotypes within the young–elderly group to conduct trend analysis.

For the old–elderly, a positive trend coefficient (0.50) emerged for 'Adventurous golden ager' and a negative trend coefficient emerged for 'John Wayne conservative' (–0.50). However, neither was statistically significant. There were too few cases for the remaining stereotypes within the old–elderly group to conduct trend analysis.

Finally, analysis (Table 8) indicated that men were less likely to be portrayed as the 'Perfect grandparent' than were women (Z = –2.25, p ≤ 0.05). For elderly men, positive trend coefficients emerged for 'Adventurous golden ager' (0.60), 'John Wayne conservative' (0.63) and the negative stereotypes overall (0.20). Negative trend coefficients emerged for 'Productive golden ager' and 'Perfect grandparent' (–0.33). However, none of these coefficients are statistically significant. There were too few cases of the various negative stereotypes within the gender group to conduct trend analysis.

TABLE 8: FREQUENCIES. PERCENTAGES, TREND COEFFICIENTS AND DIFFERENCES BETWEEN STEREOTYPES OF ELDERLY GENDER GROUPS

                                                             Elderly Men                                   Elderly Women

For elderly women, there was a marginally significant increase in the appearance of the 'Productive golden ager' (= 0.78, p ≤ 0.10). In addition, positive but non–significant trend coefficients emerged for 'Adventurous golden ager' (0.40) and 'John Wayne conservative' (0.40). A negative trend coefficient for 'Perfect grandparent' (–0.60) emerged but was not statistically significant. As with elderly men, there were too few cases of the various negative stereotypes within the gender group to conduct trend analysis.

This study is the first to examine stereotypes of the elderly in television advertising using a longitudinal design and a theoretical perspective of stereotypes drawn from cognitive psychology. This allowed analysis of specific stereotypes rather than simply positive or negative portrayals in a more general sense. Results are consistent with the view that over the past few decades the elderly have been portrayed in television commercials in a positive manner. Very few negative stereotypes were found. While specific advertisers may have used demeaning and ridiculous stereotypes at times, overall it seems that advertisers have probably not been socially irresponsible in their portrayal of the elderly. Most of the negative stereotyping involved the 'Mildly impaired', none of which were presented in a ridiculing manner. One could argue that using this stereotype is not socially irresponsible anyway. Thus, these results are consistent with the position that perhaps the critics have overstated the case.

Positive stereotypes such as the 'Perfect grandparent', 'Adventurous golden ager' and 'Productive golden ager' seem to have been the most commonly used. This is similar to the pattern of typicality ratings reported by Hummert (1993), i.e. that people believe that positive stereotypes are more typical of the elderly than are negative stereotypes. The increasing trend in the appearance of the 'Adventurous golden ager' along with the decreasing trend in the appearance of the 'Perfect grandparent' seem to reflect social reality. With the improved health and greater wealth of the elderly, they have been able to remain more active and to pursue roles beyond the traditional role of grandparent. It seems logical that advertisers would employ strategies designed to promote products that elderly people could use in these more diverse roles.

The young–elderly appeared more often than did the old–elderly as the 'Productive golden ager'. This result seems reasonable since it is likely that the young–elderly are less likely to be retired than are the old–elderly. The increasing trend for the young–elderly to appear as the 'Adventurous golden ager' is consistent with the overall trend. The old–elderly appeared more often than did the young–elderly as the 'John Wayne conservative' and for negative stereotypes combined. This result is consistent with Hummert's (1993) typicality ratings.

Elderly women appeared significantly more as the 'Perfect grandparent' than did men. This seems to reflect the traditional role of women as nurturers of children. The increasing trend for women as the 'Productive golden ager' may reflect elderly women's increased variety of roles beyond the traditional role as homemaker.

The results of this study are generally consistent with the print ad study conducted by Miller et al. (1999). Both found that advertising images of the elderly over the past several decades has been quite positive. In each case the 'Golden ager' and 'Perfect grandparent' stereotypes were the most commonly discovered. Both found that the young–elderly were more likely to be portrayed as the 'Golden ager', and that the old–elderly were more likely to be depicted as the 'John Wayne conservative'. The two studies found opposite trends in the appearance of the 'Perfect grandparent'. The print ad study found a increasing trend but the current study found a decreasing trend. Future research could investigate reasons for this difference. Perhaps it is a function of media target audience or product category variations between the two samples.

Results of this study must be interpreted within the context of the sample. As explained above, because of the difficulties involved in obtaining a random sample of television commercials broadcast over a period of five decades, this research was conducted with a convenience sample. Samples of this type, of course, may contain idiosyncrasies that can potentially influence the results. For example, the relatively high percentage of commercials for Kodak products within the sample may have confounded the results. However, with the possible exception of the decreasing trend for the 'Perfect grandparent', the results do not seem to indicate that such confounding has occurred. This conclusion is based on the consistency of these results with those found by the Miller et al. (1999) study which employed the same theoretical perspective, a similar methodology and a random sample of magazine ads. A future study could address this potential weakness with an alternative sample.

Another concern relates to the low ratio of cases to variables in the factor analysis of the summated traits. According to Hair et al. (1992), as a general rule this ratio should be four or five to one. However, they go on to say that the rule is somewhat conservative, and researchers must often factor–analyse data where only a two–to–one ratio of cases to variables is available. In any case, they remind us that findings based on factor analyses employing low ratios of cases to variables must be interpreted cautiously.

Future research could also examine patterns of stereotypes across product categories and target audiences. As stated in the introduction, it would seem unwise for advertisers to use negative stereotypes when targeting elderly consumers. However, when targeting young people, advertisers might not have adhered to similar restraint. Future research could also examine patterns of stereotype usage across ethnic and racial groups, or in other countries.

This research was funded by grants from Washburn University and the John W. Hartman Center for Sales, Advertising and Marketing History at Duke University. The authors would like to thank and acknowledge the assistance of Ellen G. Gartrell, Director of the Hartman Center and David Benjamin, Photo Archivist at the State Historical Society of Wisconsin. Finally, the authors would also like to thank the editor and the anonymous reviewers for their thoughtful and constructive comments during the review process, and Dr Lawrence J. Marks of Kent State University who reviewed an early draft of the paper.

References

Abrams, B. (1981) Advertisers start recognizing cost of insulting the elderly. The Wall Street Journal, 5 March, p. 27.

Ashmore, R.D. & Del Boca, F. K. (1981) Conceptual approaches to stereotypes and stereotyping. In Cognitive Processes in Stereotyping and Intergroup Behavior (Hamilton, D. L., ed.). Hillsdale, New Jersey: Erlbaum, pp. 1–35.

Davis, R. H. & Davis, J. A. (1985) TV's Image of the Elderly: A Practical Guide for Change. Lexington, Massachusetts: Lexington Books.

Francher, J. S. (1973) 'It's the Pepsi generation…' Accelerated aging and the television commercial. International Journal of Aging and Human Development, 4, 3, pp. 245–255.

Gibbons, J.D. (1985) Nonparametric Statistical Inference. New York: Marcel Dekker.

Goldman, K. (1993) Seniors get little respect on Madison Avenue. Wall Street Journal, 20 September, p. B4.

Gorsuch, R. L. (1974) Factor Analysis. Philadelphia: W. B. Saunders.

Hair, J. F. Jr., Anderson, R. E., Tatham, R. L. & Black, W. C. (1992) Multivariate Data Analysis. New York: Macmillan.

Harris, A. J. & Feinberg, J. F. (1977) Television and aging: is what you see what you get? The Gerontologist, 17, 5, pp. 464–468.

Holbrook, M. B. (1987) Mirror, mirror, on the wall, what's unfair in the reflections on advertising? Journal of Marketing, 51, 3, pp. 95–103.

Hummert, M. L. (1990) Multiple stereotypes of elderly and young adults: a comparison of structure and evaluation. Psychology and Aging, 5, 2, pp. 182–193.

Hummert, M. L. (1993) Age and typicality judgments of stereotypes of the elderly: perceptions of elderly vs. young adults. International Journal of Aging and Human Development, 37, 3, pp. 217–227.

Hummert, M. L., Garstka, T. A. & Shaner, J. L. (1997) Stereotyping of older adults: the role of target facial cues and perceiver characteristics. Psychology and Aging, 12, 1, pp. 107–114.

Hummert, M. L., Garstka, T. A., Shaner, J. L. & Strahm, S. (1994) Stereotypes of the elderly held by young, middle–aged, and elderly adults. Journal of Gerontology, 49, 5, pp. P240–P249.

Hummert, M. L., Garstka, T. A., Shaner, J. L. & Strahm, S. (1995) Judgments about stereotypes of the elderly. Research on Aging, 17, 2, pp. 168–189.

Kite, M. E. & Johnson, B. T. (1988) Attitudes toward older and younger adults: a meta–analysis. Psychology and Aging, 3, 3, pp. 233–244.

Langmeyer, L. (1984) Senior citizens and television advertisements: a research note. Current Issues and Research in Advertising, 6, pp. 167–179.

Miller, P.N., Miller, D. W., McKibbin, E. M. & Pettys, G. D. (1999) Stereotypes of the elderly in print advertisements 1956–1996. International Journal of Aging and Human Development, 49, 4, pp. 319–337.

Palmore, E. B. (1990) Ageism: Negative and Positive. New York: Springer Publishing.

Peterson, R. T. & Ross, D. T. (1997) A content analysis of the portrayal of mature individuals in television commercials. Journal of Business Ethics, 16, 4, pp. 425–433.

Pollay, R. W. (1986) The distorted mirror: reflections on the unintended consequences of advertising. Journal of Marketing, 50, 1, pp. 18–36.

Robinson, T. E., Duet, R. & Smith, T. V. (1995) The elderly in advertising: a content analysis of prime–time television commercials. In Proceedings of the 1995 Conference of the American Academy of Advertising (Madden, C. S., ed.). Waco, Texas: American Academy of Advertising, pp. 1–11.

Rust, R. T. & Cooil, B. (1994) Reliability measures for qualitative data: theory and implications. Journal of Marketing Research, 31, 1, pp. 1–14.

Schmidt, D.F. & Boland, S. M. (1986) Structure of perceptions of older adults: evidence for multiple stereotypes. Psychology and Aging, 1, 3, pp. 255–260.

Shimp, T. A. (2000) Advertising and Promotion. Fifth edition. Fort Worth: Dryden.

Smith, R. B., Moschis, G. P. & Moore, R. L. (1984) Effects of advertising on the elderly consumer: an investigation of social breakdown theory. In 1984 AMA Educators' Proceedings (Belk, R. W., Peterson, R., Albaum, G. S., Holbrook, M. B., Kerin, R. A., Malhotra, N. K. & Wright, P., eds). Chicago: American Marketing Association, pp. 1–5.

Swayne, L. E. & Greco, A. J. (1987) The portrayal of older Americans in television commercials. Journal of Advertising, 16, 1, pp. 47–54.

http://www.warc.com

© HAT. All rights reserved.