In 2001 Disability Studies Quarterly dedicated an entire issue to the topic of disability and advertising. As part of that issue Ganahl and Arbuckle presented a content analysis of prime time commercials that became the most often cited work in disability literature when referring to quantitative estimates of consumer advertising that contained images of the disable. From the entire pool of almost 3,000 television commercials coded, they found only 15 (.5%) contained character roles that met the coding parameters of physical disability for the study (Ganahl & Arbuckle, 2001). At the time even the very small frequency of these roles was considered a positive sign of recognition of the disability community by advertisers. Prior to 1988 the physical imperfections of the disabled made them invisible to an industry that employed a minimum standard for physical appearance (Hahn, 1987).

Since that issue was published studies examining the portrayal of disability in advertising have adopted qualitative content analysis as the preferred methodology. As justification for analyzing these commercial messages at least one respected researcher cited the Ganahl and Arbuckle study as proof of the "…inefficiency of quantitative methods" (Haller, 2010, p.194). It is hard to argue against qualitative content analysis when levels of occurrence are less than 1% as they were in the Ganahl and Arbuckle study. The qualitative approach also has provided significant evidence of positive thematic change and expansion of the types of disability portrayed in advertising messages in the U.S. and the U. K. (Haller & Ralph, 2006; Haller & Ralph, 2001).

But more than 20 years after McDonald's included a scene depicting students signing about going to the burger chain in their television commercial and Target included a model in a wheelchair in it's flyer, isn't it time to see if this particular idea is being adopted by the advertising industry? While the most recent examinations of media depictions of disability give us a guide to how people with disabilities are being portrayed, in terms of societal adoption or even impact of a practice such as the industry's use of disabled character roles in advertising, it is important to place context on the practice. It is fair to ask the question of whether the portrayal of disability in consumer advertising has reached a significant level with regard to frequency in the actual media environment so that the public sees the actor and not the disability. In essence there is value to asking the question, "Are we there yet?"

For this study a quantitative content analysis of disability portrayals in prime time television commercials was conducted. The study provided a much-needed update to the literature and also examined whether events of the past decade have influenced the diffusion of AI advertising within the industry.

Background

The media's contribution to the push for social integration of the disabled became the focus of several communication researchers in the 1990s because one of the positive outcomes from the battle for passage of the ADA was an era of heightened visibility and awareness of disability issues on the part of the media. In the 10 years that followed passage of the ADA, the disability community began to appear in front of the camera in ways that were a departure from the pattern of negative stereotypes that painted the disabled as either evil or a burden on society (Black & Pretes, 2007; Byrd, 1989; Safran, 1998). The previously "invisible minority" (Nelson, 2003 p. 175) enjoyed positive portrayals, albeit in small amounts, in entertainment media, traditional news outlets and even advertising.

If one accepts the power of the media to speed diffusion of an idea through a society, place an item on the public agenda, or serve as a learning tool for social behavior, then the increased visibility of people with disabilities in the media since 1990 should have contributed to significant improvements in the area of social acceptance for the disabled. Unfortunately the latest survey of Americans with disabilities conducted by Louis Harris on behalf of the National Organization on Disability (NOD) reported large gaps between disabled and nondisabled respondents' measures of quality of life (www.2010disabilitysurveys.org). It would be naïve to think that just because the media started portraying people with disabilities in a positive light that all barriers to equality might disappear, but "…while movies entertain, they simultaneously provide viewers with information about disabilities…and representations of how individuals fit into a nation's social and political landscape" (Safran, 2001, p. 223). If the images from films, TV and even advertising portray people with disabilities in negative fashion as objects of pity, then society will see them from that perspective.

Classroom SADP Suggests Additional Concerns

Prior to the decision to go forward with this study, we attempted to get some indication of any trend with regard to attitude toward people with disabilities on the part of the nondisabled by administering the Survey of Attitudes toward Disabled People (SADP) scale to 345 college students across the country as part of a classroom introduction to consumer research. The 24-item summative-rating SADP scale required respondents to rate each of 24 statements about disability on a six-point scale ranging from '-3' (I disagree very much) to '+3' (I agree very much). The range of scores was 0 — 144 with a higher score indicating a more favorable attitude toward people with disabilities. The grand mean for the sample was 100.45. To put that into perspective the mean SADP score during original reliability testing by Antonak was 111 (Antonak & Livneh, 1988). A published study on AI advertising in 1996 reported a mean SADP score of 108 for college students in the southeast (Farnall, 1996). Using a one sample t test with Antonak's original number as the population mean, the 2008 measure of 100 is significantly lower (p < .001) indicating a somewhat counter-intuitive trend between expected media exposure and attitude shift. It is important to note that no measure of media exposure to AI advertising was taken for the 2008 sample therefore no direct correlation can be made.

Is our nation's attitude toward people with disabilities becoming more negative as the classroom SADP experience would seem to indicate? Again the role of the media comes into question. In the pre-ADA period several major studies found the depiction of people with disabilities to be negative, unrealistic and stereotypical (Donaldson, 1981; Longmore, 1985b; Zola, 1985). By the early 1990s studies were reporting improvement in frequency and type of portrayal for the disabled (Byrd 1989; Nelson 2003). According to Farnall and Smith (1999), individuals who viewed positive portrayals of people with disabilities on television shows and in films were less likely to report negative emotions associated with disability images. But just like the 1930 Payne studies on film, Gerbner's Cultural Indicators Project, and other well known research that linked societal attitude to media, step one in addressing the media portrayal-attitude question seems to be content analysis.

Baseline AI Advertising Study

As previously mentioned, the only published quantitative content analysis addressing the frequency and portrayal of people with disabilities in television advertising was published in a special issue of Disability Studies Quarterly in 2001. In that study Ganahl and Arbuckle (2001) collected 2,999 prime time network television commercials that aired during two successive February sweeps periods in 1998 and 1999. Each spot was coded for presence of characters displaying physical impairment. In this study a physical impairment was defined as a visible and non-correctable impairment. If those images were found, coders determined whether the character played a primary or secondary role in the commercial. From the entire pool of commercials they found only 15 (.5%) commercials that met the first condition. Although the total number of AI commercials was less than 1% of the total sample, the study provided at least some kind of estimate for disability advertising. If there was a positive sign from this accounting it was that almost 75% of the roles classified as "physically disabled" were also coded as major character roles (Ganahl & Arbuckle, 2001).

Advertiser Concerns Since Baseline

In the almost 14 years since data were collected for that study the advertising landscape has changed drastically but no new quantitative measure of AI advertising has been published. There are reasons to expect a positive growth for AI advertising, including a 2006 quote in powerhouse trade publication Advertising Age that called the connection between people with disabilities and major corporations, "just good business sense" (Voight, 2006); but recent history has called the value of the association into question.

Negative Indicators

There were three AI advertising-related events that may have negatively impacted the diffusion rate for AI advertising in the U.S., and reduced the number of advertisers willing to incorporate disability roles into their marketing strategies. The most visible was the infamous Christopher Reeves commercial. The second was the cancellation of the only national awards program that recognized companies that created positive roles for people with disabilities in advertising. The third was the series of lawsuits that targeted companies affiliated with the disability cause.

1. Superman bombs. Recent history has shown that writing ads that incorporate disabled characters in general consumer products campaigns can offend more than endear (Cheng, 2002; Sewell, 2005). The most famous example was the Christopher Reeves spot of 2000 for an investment company in which Reeves was portrayed as walking again thanks to funded research. Critics of the ad said that Reeves' refusal to accept his quadriplegia was offensive to those who have learned to live with their disabilities (Nelson, 2000). In addition the computer graphics used to import Reeves' head on the body of an actor resulted in a less than proportionate figure. In particular, people with spinal cord injuries protested that the Reeves' spot damaged marketing efforts on spinal cord research and presented a false picture that mobility was only a matter of time (Sewell, 2004). Reeves actually defended the spot in an interview with Diane Sawyer saying that he thought the quality of the production and the simplicity of the message made it an appropriate message.
2. Advertising agencies loose a chance to shine. Success in advertising has been operationalized in many ways, but the number of awards won has been one standard used by agencies and advertisers alike for decades (Till & Baack, 2005). In fact one estimate placed the number of advertising award shows at more than 500 worldwide (Maxwell & Wanta, 1998). It has been 10 years since Easter Seals cancelled the biggest and longest running AI advertising awards program in the country, the nationally recognized Equality, Dignity and Independence (EDI) awards program. Since 1989 the EDI Awards program had attempted to recognize superior broadcast and print advertising that included positive portrayals of people with disabilities. The EDI program brought advertising agencies and their clients national publicity for "doing the right thing." The show was a major media event in New York, costing several hundred thousand dollars and often featuring major stars as guest hosts. It proved to be a popular event with the industry as entries in the third year of the competition had increased 300% (Farnall, 2000). While it may not be a politically correct statement, the truth is that advertisers and agencies like to win awards and they benefit from the press they receive, especially when the brand is associated with positive community action. Cancellation of this program may have had a negative impact on advertisers' motivation to design and execute AI advertising.
3. ADA lawsuits plague corporations. The volley of AI ads in the mid 90s was followed by a wave of ADA-related lawsuits. American Online, Wal-Mart and Kmart paid out approximately $25 million in ADA related claims (Voight, 2006). These lawsuits made marketers cautious about associating with the disability issue at all and cut down on the number of clients asking to be associated with people with disabilities in any communication (Voight, 2006). No figures were available for total claims paid but it is safe to say that companies do not like paying settlements on lawsuits.

Research Questions

Clearly there are mixed messages in terms of the history of AI advertising and no real body of evidence that suggests the amount has increased or decreased. One would assume that with the success of major advertisers like Target, McDonalds, Levis, Toys "R" Us, and Cingular Wireless, more and more companies would expand their focus to include people with disabilities. Some marketers have even changed company policy to make sure minorities and people with disabilities are included (Voight, 2006). But issues such as those presented here make the answer to the question of representation unclear leading us to the following questions:

RQ#1a. Has the amount of AI advertising in prime time television programming on both network and cable channels increased from the levels measured one decade ago?

RQ#1b. Is there a difference in the amount of current AI advertising aired on cable versus network television?

RQ#2. Has the composition of AI ADVERTISING portrayals varied since the first Ganahl and Arbuckle (2001) study?

Method

The first content analysis of AI advertising commercials by Ganahl and Arbuckle (2001) provided a good starting point for determining an appropriate sample size and unit of analysis in this study. However, because cable viewership has grown substantially from when the data was collected in 1998 and 1999, this analysis was expanded to include cable networks. It was also determined that cable channels might be more heavily used vehicles for AI advertising because smaller cable audiences mean lower advertising rates and cable provides some narrowly niched programming that would be of interest to this market.

Procedure

Investigators recruited coders from an undergraduate research methods class at a major university. Those students who were interested in receiving the extra credit offered for the task were instructed to attend a training session held outside class. At the training session investigators briefed the students on the project including a discussion of disability, a review of the codebook and instructions and a demonstration of the coding process using several sample AI advertising. All the students who attended the coding sessions decided to continue with the project and were given a set of coding sheets and an instruction sheet that included the channel/time/day to be recorded. So that inter-coder reliability could be calculated, investigators had 20% of the 50 one-hour blocks coded by two different people and their results were compared.

Results

Fifty hours of network and cable television programming were analyzed. The programming contained a total of 1,692 commercial messages. Of those, 21 did not meet the required length standard and were discarded even though they were included by a coder. The final sample contained a total of 1,671 valid commercials that were content analyzed.

Table 1 compares the number of times a disabled character appeared in a commercial based on type of channel. With approximately equal hours of programming, cable channels (52%) carried more AI advertising than network channels (39%). Local independent channels accounted for 10% of the AI ads, but represented only 6% of the total hours programmed. The results are not counter-intuitive but do present an additional area of investigation in the future, namely whether local or spot advertising across channels carries a disproportionate amount of AI advertising.

Discussion

Barbara Kolucki, a consultant on disability and children's media, writing for Rehabilitation International, posed the question, "Has our public education and media efforts made a difference?" (2009). In answering her own question Kolucki points out that there are simply too many variables in the world of attitude formation and change to be able to isolate the contribution of any one. For example, is it advocacy, legislation, public education or media that caused change? Her basic approach to the question was that it does not matter as long as change occurs.

The 200% increase in the amount of prime time TV AI advertising over levels measured in 1998-99 could be interrupted as proof of change on the part of the advertising industry. Haller and Ralph (2006; 2001) established that advertising storylines/themes had changed with their qualitative focus on AI advertising and now this quantitative study confirms that the advertising community is adopting the practice of including images of disability in its executions. The increase in the amount of AI advertising from this content analysis over the 1999 level is a sign that efforts from media and groups like the American Association of People with Disabilities have made a positive contribution to representation of the disability community in advertising.

The 1.7% total for AI advertising in this study also compares favorably with the two most recent content analysis studies that looked at percentage of disabled character portrayal in the entertainment industry. A study of role portrayal in prime time Japanese TV found an identical 1.7 % of roles displaying disability during prime time (Saito & Ishiyama, 2005). Henderson and Heinz-Knowles' (2003) content analysis for prime time U.S. TV during Fall 2000 and Fall 2001 found only .6% of the roles were for disabled actors. Interestingly data for their study was collected about the same time as the baseline AI study by Ganahl (2001) and the total representation for entertainment media was about the same as advertising. If a more recent study of prime time entertainment programming were available for the U.S. it would allow for an interesting examination of trends in disability portrayal amounts across genres.

In terms of the type of portrayal, this study reported a 2.6% drop in the percentage of major roles dedicated to disabled characters. Even with the drop there are still twice as many major roles for people with disabilities as there are minor roles. Considering that many of the early efforts at integration as demonstrated by television commercials entered in the Equality, Dignity and Independence (EDI) competition in the 1980s made use of the "background kid in a wheelchair" approach, one would have to view the findings of this study in a positive light. Also, the fact that the portrayal of disabled individuals took such varied form in this sample should be viewed as a signal that organizations involved in casting and producing AI advertising for the past 20 years are having a positive impact. These casting organizations such as the Media Access Office, Solution Marketing Group, and WC Duke Associates have been successfully guiding clients through the pitfalls of AI advertising execution for more than a decade (Krossel, 2001).

As Media Exposure Increases…

In the last 10 years many would argue that the media, including advertisers, have done much to project a more positive image of people with disabilities. The correlation between exposure, through personal contact or media portrayal, and positive attitude has been made in the disability literature over and over again (Farnall & Smith, 1999; Nelson, 2000; Shapiro, 1993; Yuker, 1988). Several notable disability advocates have echoed the familiarity theme as a major influence on attitude including Dr. I. King Jordon, Gallaudet University president who said, "The most important barriers to the employment of disabled people are attitudinal barriers. We are often judged unable to do something before we're given an opportunity to try. The worst thing about having a disability is that people meet it before they meet you" (Hopkins & Nestleroth, 1991, p. 74).

So why then did the SADP measure for college students in 2009 show a significant drop over previous studies? The increase in amount and type of AI advertising over the past decade reported in this study should signal a positive shift in the measure of attitudes toward people with disabilities, not the significant decline reported early in this paper. It is this counter-intuitive finding that leads this author to suggest there are two areas of future research that should follow this study. The first is a very complete content analysis of characters in prime-time television broadcasting in the US. The Media Access Office (MAO) has been tracking roles received by their clients since 2000 but many individuals who are SAG members would not be included in the MAO's count. For example Hugh Laurie plays a physically disabled doctor in Fox's House, but he is not a member of the MAO nor does his character get included in their count. Although a large undertaking getting some kind of quantitative count would help paint a clearer picture of what America is watching.

Richard Antonak, the creator of the SADP scale, suggested the other vein of research that should be a follow-up to these content analyses. In an article written for Disability and Rehabilitation Antonak and Livneh (2000) suggested that to obviate validity threats researchers should consider both indirect measures and physiological methods. In particular he suggested that electromyography might provide accurate measures yet no applications of this specific technology for measurement of attitudes could be found. A number of studies have attempted to develop multidimensional attitude scales and implicit association tests as a better means of obtaining true measures of attitude toward the disabled but these attempts have not been widely accepted (Findler, Vilchinsky, & Werner, 2007; Pruett & Chan, 2006; Thomas, Vaughn, & Doyle, 2007). If a reliable and valid measure were adopted that indicated attitude toward the disabled had become more negative in the face of more and more positive media portrayals it would suggest more research is needed in the area of attitude formation and change with regard to the disabled.

Are we There Yet?

A great many government organizations, charities, advocacy groups, consulting firms and media organizations are expending enormous amounts of resources because they all agree with the premise that television is a powerful educator for cultural values and attitudes. Social learning theory, cultivation and media dependency theory all support that premise, as does a body of work in the rehabilitation literature (Elliott & Byrd, 1983; Hopkins & Nestleroth, 1991). In the small world of AI advertising the results of this study suggest that images of people with disabilities in advertising are a bigger part of the overall advertising environment than they were in 2001 despite the difficulties associated with contextual elements in an ad, and the public outrage/civil suits. But if we compare the appearance rate for AI advertising, as part of the total advertising environment, to the percentage of adults classified as disabled in the total U.S. population based, people with disabilities are very much under-represented (1.7% from this study as compared to 12% according to the 2009 Disability Compendium). If equal representation is the goal, we are not there yet. If acceptance of the use of disabled portrayals in general product advertising on the part of the advertising industry is the goal, then much progress has been made.

Perhaps a better question to ask at this point is, "How do we get there?" Partnerships between disability advocates like the American Association of People with Disabilities and corporate giants such as American Airlines (Altitude Awards) and Wal-Mart (Internship funding) should certainly continue. But perhaps it is time to target advertising industry professionals. Instead of measuring public perception and response to AI advertising, it may be time to address agency personnel who are responsible for creating and executing advertising strategy. Is AI advertising even an option in their minds? Are ad professionals more positive or negative in their attitudes about AI advertising? For those agencies that have used AI advertising in the past, was the motivation client-driven as was the case with much of the early AI advertising (Sewall, 2005)? Answers to these questions may shed light on the acceptance of AI advertising and give advocates a clearer picture of just how long it might be before seeing a person with a disability pitch breakfast sausage or shop for family gifts at the local K Mart is just good advertising instead of advocacy advertising.