4. Secondary analysis

4.1 What proportion of television and radio programming contains representations of people with disabilities?

Beyond the central finding that 8% of programmes in the sample featured representations of a person or persons with a disability, the most striking discovery was the fact that 10% of all television programmes in the sample featured a person with a disability as compared with 6.5% of all the radio programmes looked at. The fact that television programmes are nearly twice as likely as their radio counterparts to feature a person with a disability needs some consideration. That radio and television are characterised by different genres characteristic (see below) may partially account for this. However, one cannot overstress the impact of the greater difficulty involved in identifying an individual with a disability in the absence of visual cues in the case of radio. Although, in most representations of people with disabilities across the sample (73%) the audience's attention was drawn verbally and therefore directly to the fact of an individual's disability, more than a quarter were identified by tacit or implied cues, e.g. a visual cue such as the fact that an individual is in a wheelchair or uses a cane or through an aural cue such as slurred speech. However the aural cues offered by radio do not permit the identification of disabilities which do not affect speech.

The matter is further complicated by separating the television and radio results for representations of people with disabilities. On television, 12 of the 51 representations of people with disabilities (or 24%) were identified through tacit signals. Surprisingly, however, an even higher percentage of radio representations - 41% or 15 of the 37 representations - are coded as tacitly signalled. This appears to undermine the suggestion that on radio only clearly signalled representations of people with disabilities are evident to listeners. However, if one examines closely the 15 tacit representations on radio, a slightly different picture emerges. 6 refer to Cho Seung-Hui, the Virginia Tech shooter. Although later news bulletins explicitly refer to his mental illness, the initial six described him only as "a troubled loner". Since this was not a direct, overt reference to mental illness, the coders correctly opted not to code this as a direct reference. Instead they coded it as an implied, tacit reference.

However, strictly speaking this is not correct either since "tacit" is only intended to be applied to non-linguistic cues signalling impairment. In a similar vein, a further 6 tacit representations of people with disability are accounted for by the NEAR FM disability-focused show "Equality Time". In this case, none of the six individuals are singled out as disabled, so the coders are again correct to refrain from categorising their impairment as being directly signalled. However, since it is made clear in the introduction to the programme that the show is for and about individuals with disabilities, it is reasonable to infer that the participants are disabled. Again, however, coding their disability as tacitly signalled could be regarded as somewhat misleading since it is a direct spoken reference at the outset which allows the listener to infer the status of the participants. Indeed, a close examination of the 15 radio representations of people with disabilities categorised as "tacit" suggests that only one - the person with Down Syndrome presenting a Ros FM music show - was not signalled by semantically overt statements.

All 14 radio representations then (the 15 referred to above minus the Ros Fm case), are examples of a situation which the questionnaire design did not anticipate. However, for the purposes of assessing the extent to which the presence or absence of people with disabilities may be disguised on radio, it would be more accurate to code all 14 instances referred to above as direct references to disability. In effect, this means that only 3% of representations of people with disabilities on radio are tacit in the same way that 24% of television representations are considered tacit. This appears to offer some statistical support for the hypothesis that some incidences of impairment are "invisible" on radio and therefore uncapturable for research purposes.

In this respect, it should also be noted that the research categorised as disabled only those individuals who were represented as such through direct or tacit cues. Thus even if an individual coder knew a particular individual featured in the sample to be in some way impaired, they would not code them as such since to all intents and purposes, that individual was not represented as disabled For example, British Prime Minister Gordon Brown is technically disabled since he has a visual impairment (a false eye). However since this fact is rarely referred to in media coverage he is generally not represented as a person with a disability.

4.2 What proportion of television and radio programming contain spoken references only to disability?

The analysis found that 9% and 11% of all radio and television programming respectively contained spoken references only to disability. That spoken references to disability were more evenly spread across media than actual representations of people with disabilities appears to lend further weight to the thesis that the absence of visual cues may have been a factor in the relatively low proportion of radio programmes featuring people with disabilities.

Although 95 radio and television programmes (or 12% of the total sample) featured spoken references to disability where there was no person with a disability in immediate evidence, there were 108 separate references to disability across those 95 shows. The majority were very brief (usually no more than a sentence or phrase).

The researchers found only ten spoken references which constituted substantial discussions, on the basis of the amount of time/dialogue devoted to the subject and the role played by disability in structuring the programme a whole. These included:

• a substantial discussion around the subject of mental illness on LMFM's news programme in the context of the Virginia Tech shooting
• a discussion of students with special needs on TG4's Nuacht
• a monologue on the question of what constitutes "normality" in the context of a sermon delivered on a Mass broadcast in South East Radio
• a discussion of autism on Cork's 96FM inspired by the deportation of Great Agbonlahore, a Nigerian boy who had sought asylum in Ireland and who was himself autistic
• a recurring theme of faking impairment in an episode of "Seinfeld" on Channel 6
• a discussion of suicide and mental illness on NEAR FM's "Northside Today", and
• a discussion with Finian McGrath TD and his interest in education, disabilities and 'special needs' and a more generalised discussion about alcoholism and drug-taking in Ireland today on "The Wide Angle" on Newstalk 106FM

Of course, this is not to suggest that there were no other substantive discussions of disability in the programme sample. Rather, it highlights the fact that that most of the discussions about disability featured the participation of a person with a disability. The fact that there were so few substantive discussions of disability issues in the absence of people of disabilities points to an apparent preference on the part of broadcasters for ensuring that people with disabilities are invited to participate in more in-depth discussions of the broader subject of disability.

This point is leant further weight by the fact that the 24 references to disability which were categorised as "relevant" (rather than "central") were almost exclusively radio news items in which the disability theme was incidental to more central themes such as industrial action (in the health care sector) and deportation of immigrants (the Great Agbonlahore story).

However, arguably the most significant finding from this question was the extent to which disability is casually referenced in contexts which are otherwise unconnected with discussions around disability. The great majority of references to disability across television and radio (74 or 69% of the total) were considered "incidental". Such incidental references were particularly evident in drama programming (and in particular sitcoms) which used references to mental health, vision and hearing impairments in a colloquial fashion (e.g. "Are you crazy?" "You'd have to be blind not to see..." etc). However such references were also found in contexts as diverse as sports commentary and magazine shows. The fact that such comments are deployed, not merely so casually but also so widely, hints at the extent to which particular meanings are taken to be universally understood as associated with various forms of disability.

For example the sample includes an episode of "Seinfeld" on Channel 6 where one character exclaims "I can't be blind, the blind are courageous."

4.3 What proportion of all people sampled on television and radio programming had a disability?

The fact that only 1.1% of people identified as appearing in the total programme sample were categorised as disabled, indicates the under-representation of people with disabilities relative to their position in the population as a whole. Even if television offers marginally higher figures (1.8% of all people identified being categorised as a person with a disability), this is still far short of the 9.3% figure quoted by the Central Statistics Office as representing the proportion of people with disabilities in the population of the Republic of Ireland.

Fictional characters accounted for 34 of the total 88 appearances by people with a disability during programming examined, with the remainder made up by "real" individuals. All of the fictional characters appeared on television.

The vast majority of those people coded as a person with a disability in programming were physically present but in one or two cases, the researchers included individuals who, though not physically present were extensively discussed and therefore "represented". For example, "Sounds for Sunday", a church sermon broadcast on South East Radio referred at length to Helen Keller, the blind and deaf writer and activist. The discussion of Keller clearly presented her in a particular - saintly - light and thus contributed to the overall representation of people with disability in the programmes sampled. In total, the presence of people with disabilities was distributed as follows:

• 52 programmes featured a single individual
• 7 programmes featured two individuals
• 4 programmes featured three individuals, and
• 1 programme ("Equality Time" on NEAR FM) had 6 individuals with disabilities

Strictly speaking, the 1.1% may even be an overstatement of the proportion of people with disabilities in the sample as a result of the method adopted by the research for counting all individuals appearing in a given programme. In a bid to overcome the difficulties posed by groups or crowds where it would be difficult if not impossible to accurately count the number of individuals researchers only enumerated "individuated" characters, (the definition of which is outlined in the explanatory guide to the questionnaire included in Appendix 2). However this probably means that the figure 7,223 which describes the total number of people appearing within the sample as a whole is an understatement. Thus the 88 appearances form a larger proportion of the total number of people appearing in the programme sample than if a less conservative method of estimating the total number of all appearances was adopted.

However, the research also identified non-individuated groups of people with disabilities in the sample which were not included in the count of people with disabilities as it was not impossible to establish how many people were in the group. These appeared on:

• "The 'L' Word" on Channel 6
• "The Business" on RTÉ Radio 1 on Autistic Schools
• "Nationwide" on RTÉ 1 segment, also focusing on a school for Autistic children
• "Exorcist: The Beginning" on TV3, a motion picture partially set in an asylum, and
• "Your Call With Brenda Power" on Newstalk 106, in the course of a discussion about Autism

The fact that these individuals were not included in the estimate of the proportion of people with disabilities relative to the total number of individuals in the sample may partially compensate for the conservative estimate of the latter figure.

4.4 What proportion of programming is disability-focussed and what proportion is mainstream?

That only 5 programmes were considered as disability-focused out of a total of 804 is evidence of the peripheral status of disability-focused broadcasting. Arguably even the three editions of News for the Deaf are primarily news-focused in content even if their form is directly geared to the needs of hearing-impaired viewers. Similarly, the fact that the only radio programmes were found on community radio stations (Ros FM and NEAR FM) seems to reflect a reluctance on the part of mainstream broadcasters to provide programming actively targeting audiences with disabilities, in particular during mainstream hours.

It may be the case that general programming may well serve the needs of people with disabilities as much as anyone else's, although it is beyond the scope of this research to establish whether or not this is in fact the case. Furthermore, there are programmes on mainstream broadcasters which are not included in this sample. But their absence from the sample reflects the extent to which such programming tends to be scheduled outside of prime time viewing (and outside the sampling frame). For example, RTÉ's "Mind Matters", "Outside the Box" and "Audioscope" are all broadcast after 7pm on radio. The only exception to this is RTÉ 1 television's "360°" which is usually broadcast in a 7.30pm time slot but again this programme happened to fall outside the sampling period upon which the research was based.

4.5 How many different people/characters (in the case of fictional/dramatic content) contribute to the representation of people with disabilities?

The research found that the 88 representations of people with disabilities were based on 72 individuals. The nature of the sample which, in some cases included 12 hour blocks of consecutive programming ensured that any individual with a disability featuring prominently in news bulletins would make many appearances. The most evident example of this was Cho Seung-Hui, the university student who shot a number of students at Virginia Tech, who appeared 11 times in the sample, accounting for 13% of all instances of appearances of a person with a disability in the sample period. There were also three separate representations of Great Agbonlahore, an autistic boy who, along with his family, was deported to Nigeria. And finally there were also two programmes representing the autistic son of Yvonne O Cuileachain, who sued the Department of Education in relation to her son's education.

However the fact that the same individuals appear several times does not necessarily mean that they are somehow over-represented in the sample. Even the total headcount of the 7,223 individuals appearing in the overall sample includes many references to some high profile individuals. For example, figures like the then Taoiseach Bertie Ahern, may well appear in virtually every news bulletin. ^[8]

4.6 In which genres of programmes are people with disabilities most and least often represented?

The question of programme genre demands extensive consideration. In particular, it is important to note that genre is a key factor accounting for the differences in how radio and television treat disability. Understanding the significance of programme genre is also crucial for any understanding of the distinctions between the extent to which imported and domestic programmes represent disability. Both these points are taken up below and later on in the discourse analysis in the next section of the report.

The research used nine broad programme genre classifications to categorise programmes with an additional "other" category to capture any miscellaneous programmes. The genres used were:

1. Drama
2. News/Current Affairs
3. Young People's
4. Comedy
5. Lifestyle
6. Music
7. Factual
8. Sports
9. Movies/Mini-Series

To introduce clarity into our classifications we outlined sub-categories for "drama", "young people's", "lifestyle" and "factual". Drama included: -

• serious drama ("CSI", "Wire in the Blood", "The Sopranos")
• comedy drama ("Ally McBeal", "Sex in the City", "Desperate Housewives"), and
• soap opera ("Fair City", "Coronation Street")

"Young people's" programming included:

• cartoons ("Rugrats", "Transformers")
• drama ("Sabrina, The Teenaged Witch")
• mixed format programmes such as "The Den" or "Going Live" and
• news programming targeted at youth audiences

"Lifestyle" is a particularly broad category and is increasingly difficult to separate from factual programming. For the purposes of this research, however, "lifestyle" was defined as including:

• chat/talk shows ("The Late, Late Show"/"Oprah"/"Ricki Lake")
• phone-in shows ("Liveline", "Your Call")
• magazine shows ("The Afternoon Show"/"The Tubridy Show");
• quiz shows ("Winning Streak"/"The Price is Right")
• "How-to" shows (diy/gardening/cookery/house buying/house upgrading/makeover - "Beyond the Hall Door"/"Ground Force"/Rachel Allen/"Location, Location, Location"/"Room to Improve"/"Pimp My Ride")
• Infotainment, and
• Travel shows

Thus "factual" programming referred to:

"straight" documentary ("Hidden History"),

• nature programming (David Attenborough)
• reality programming ("You're a Star", "Celebrity Farm")

Moving to the actual results, we begin by considering the apparent contradiction whereby news and current affairs accounts for the largest number of appearances by people with disabilities but is also a genre in which one is less likely to encounter such individuals. This clearly requires some explanation.

If the relative prominence of all the programme genres in the sample as a whole and in that subset of programmes featuring people with disabilities was identical (e.g. if news accounted for 45% of the overall sample and 45% of those programmes actually featuring people with disabilities) then one can say that the likelihood of a person with a disability appearing in that genre is evenly distributed across genres i.e. one is as likely to see a person with a disability in a given hour of news as one is in drama, comedy, lifestyle programming etc.

However if some programme genres are more prevalent in the subset of programmes featuring people with disabilities than in the sample as a whole, then one can say that people with disabilities are more likely to appear in any given example (i.e. programme) of that genre. Similarly, if some programme genres are less prevalent in the subset than in the sample as a whole, then one can say that people with disabilities are less likely to appear in that genre.

Three genres were noticeably more present in the subset of programmes featuring people with disabilities. One was 3.5 times more likely to encounter a person with a disability in a comedy, 2.4 times more likely to encounter them in drama and 1.75 times more likely to encounter them in a lifestyle programmes than if the presence of people with disabilities had been evenly distributed across programme genres.

4.7 In programmes that represent people with disabilities, what is the level of appearance of the person/people with disabilities?

The majority (69%) of representations of people with disabilities were characterised as representing them in minor or incidental roles. This may be accounted for by reference to a combination of factors. Key amongst these are the procedures followed by researchers in assigning individuals to these categories and - related to this - the fact that news programming accounted for 29% of all appearances of people with disabilities. These factors may also account for the fact that, as noted in the previous section, people with disabilities are more likely to be represented in major roles on television than on radio.

Let us first outline how the coders determined which category to assign to given representations. Researchers applied a combination of four criteria relating to:

• Amount of on-screen/on-air time;
• Amount of speech/dialogue;
• Role in developing the narrative/story, and;
• Extent of character development (this mainly applied to fictional representations) plays some role in this.

In effect this meant that for the appearance of a person with a disability to be categorised as "major" in the context of news/current affairs programme, the item would have to be not merely focused on that particular individual but the individual would have to be in a position to play some role in shaping the manner in which the story was told.

For example virtually all the news bulletins about the Virginia Tech killings were focused on a person with a mental health disability. However, the person in question, Cho Seung-Hui, scarcely appeared on-air except in still photos (in television reporting), said nothing and, since he had killed himself by the time, the first bulletins went out, could not play any role in shaping the story. Thus although 24 news stories in the sample featured an individual with a disability those individuals generally played a passive role in the story. On-air interviews with a person with a disability were also almost entirely absent from news programming. Thus only 2 of the 27 appearances coded as "major" were drawn from news/current affairs.

This finding should be related to the point made in section 4.6 regarding the role of editorial decisions in shaping representations of people with disabilities. In contrast to other genres, such as drama, which may be produced at a relatively leisurely pace, news and current affairs production is commonly understood as fast-paced, responding to events as they happen. Even if, in reality, this is something of a caricature, it remains the case that news and current affairs production is governed by deadline pressures that do not apply to other genres. The question is whether this in any way accounts for the absence of major appearances of people with disabilities in news programming? Is there any perception on the part of news editors that finding an interviewee with a disability is too time-consuming and thus impractical?

The question is thrown into relief by the fact that genres with longer lead-in production times such as lifestyle shows (and in particular chat and magazine shows) frequently conduct set-piece interviews with individuals with disabilities as a means of discussing particular impairments. Indeed lifestyle shows accounted for more than a quarter of the appearances that were coded as "major". Similarly, drama which has even longer lead-in times accounted for 22% of all "major" appearances.

Whatever the reason, the absence of "major" representations of people with disabilities in news and current affairs had a cumulative impact on their representation on radio in particular, since news/current affairs programming accounted for more than half (56%) of all radio representations of people with disabilities. Thus the nature of news coverage as outlined here immediately above may contribute to the relative absence of representations of people with disabilities in major roles on radio.

4.8 What types of disability are represented?

Representations of people with mental health issues during the programming sampled were the most frequent (accounting for one-third of all the people represented) while representations of sight and hearing-related disabilities were relatively scarce. Although these representations of mental health included representations of schizophrenia, obsessive-compulsive disorder and anorexia, depression accounted for 19 of the 32 appearances by individuals with a mental illness. The coverage of the Cho Seung-Hui story noted above was a factor here since in 10 of his appearances he is described as suicidal/depressed. However, even without his appearance, depression would account for a majority of the mental health appearances.

The figure of Cho Seung-Hui notwithstanding, the prevalence of representations of mental illness is also accounted for by the incidence of such representations within fictional genres of comedy, drama and movies/mini-series. These genres accounted for nearly a third of all such representations as Table 6 indicates.

Although this is discussed at greater length in Section 5 below, it is evident that mental health conditions such as depression offer a wide range of possibilities to screenwriters. Long-established characters (as in a soap opera) can be reinvigorated by conjuring a storyline around the sudden emergence of a mental heath condition. This is harder to do with a physical disability which is usually signalled as soon as the character is introduced to a given dramatic context. (It is not however, entirely impossible: in addition to the possibility that a character may become physically disabled as a result of physical injury there are a number of progressive physical illnesses such as Multiple Sclerosis, Parkinson's Disease and Arthritis.)

Furthermore, by contrast with a physical disability, a mental health condition is dramatically useful in that it can subsequently be written out (i.e. treated/cured), thus allowing the character to return to "normality". As a condition, addiction, offers similar dramatic possibilities: four of the nine representations of addiction occurred in television dramas or comedies.

The category of people with learning disabilities is similarly dominated by one impairment. 11 of the 15 appearances were by people with Autism although there are two instances featuring people with Down Syndrome. The prevalence of autism was partially related to news coverage of the Great Agbonlahore story, an autistic Nigerian boy deported to Nigeria.

However, news coverage of ongoing disputes between parents of Autistic children and the Department of Education also ensured that autism loomed large (three big stories about autism occurred during the sample period). It is difficult to account for the prevalence of autism by reference to anything other than timing i.e. the fact that several major news stories related to autism broke during the sample period.

The "other" category was also quite large. A single episode of the comedy show "Scrubs" on Channel 6 featured three narcoleptics while there were two individuals with cystic fibrosis on RTÉ 1 television shows "Pobal" (a community-oriented magazine show) and the "Seoige and O'Shea" chat show. Beyond that were a variety of chronic illnesses and unusual conditions such as AIS (Androgen Insensitivity Syndrome which featured in TV3's medical soap opera "All Saints") and Turner's Syndrome, a disabling genetic disease (in police drama "Law and Order: Special Victims Unit" on TV3). 6 individuals included under the "other" category could not be coded. All appeared on "Equality Time" on NEAR FM and although it was inferred that all 6 participants had a disability, the specific nature of their individual disabilities was not disclosed and there were insufficient tacit cues to make a judgement possible.

The distinctions in the range of disability types in radio and television are were partially accounted for by the different characteristics of television and radio. The finding that physical impairments accounted for 22% of all those represented in the television element of the sample, but only 5% of those in radio element, reflects the fact that purely physical disabilities may not be discernible on the radio without a verbal cue.

Similarly, although hearing-related disabilities were solely found on television, this is less surprising when one considers the aural nature of radio.

4.9 How does the representation occur (i.e. stereotyped etc.)

4.9.1 Tone

The most striking finding with regard to the tone adopted in representing people with disabilities was the overwhelming incidence of "serious" tones. 76% of all representations were categorised as "serious". This was particularly acute on radio where 97% of representations were considered serious. By contrast on television, although 65% of all representations were categorised as having a serious tone, nearly a third were considered "comedic" or "light". In considering the reasons for these cross-media TV/radio distinctions it is important to bear in mind the type of genres which dominate the respective genres i.e. all of the programmes in the radio sample featuring people with disabilities were non-fiction (primarily news/current affairs and lifestyle programmes). In short, radio representations of people with disabilities occurred in a "real-world" context where ethical constraints (or what might more crudely be termed "political correctness") limited the range of tones within which disability as a subject could be dealt with. In other words, non-fiction representations almost universally adopted the approach that disability is - by definition - a serious subject, not something to be made light of. Arguably, however, this serves to create a somewhat one-dimensional representation of people with disabilities, one which emphasises their perceived "otherness".

By contrast, because television representations occurred across a range of both fictional and non-fiction programme genres, there was greater freedom to adopt less sombre tones.

This is particularly true of fictional representations, where attitudes towards people with disabilities which might be considered beyond the pale in non-fiction, can be expressed by placing them in the mouths of fictional characters. Virtually all of the comedic and light representations of people with disabilities on television occurred in the fictive genres of drama, comedy and movies/mini-series.

4.9.2 Stereotyping

In considering the stereotyping results, a degree of caution is required. Content analysis as a methodology is based on categorising overt programme content (in this case, representations of disability). However, some processes such as stereotyping are not always overt. Stereotyping can be thought of as a social process whereby one group holds ideas about the nature of another group based on the assumption that all members of the latter group share characteristics in common. In the 19^th century for example, the Irish as a "race" were commonly stereotyped as stupid by both English and American writers.

For this research, the coders attempted to identify instances whereby the manner of representing people with disabilities implied that all other people with the same disability shared a number of characteristics unconnected with their disability in common. To facilitate the objective categorisation of instances of stereotyping, coders were asked whether a given representation of a person with a disability matched any one (or more) of 11 well-established stereotypes relating to disability (sinister, pitiable, unable to participate in society etc.) drawn from the existing literature on the subject of media representations of disability.

Thus, for example, coders confronted with the figure of Cho Seung-Hui, the Virginia Tech shooter, categorised his representation as simultaneously "sinister", "aggressive" and an "outcast". Thus one could reasonably argue that the representation of that individual conformed to a number of stereotypes relating to mental health.

However, in interpreting these figures, it is critical to understand that the same representation of a person with a disability could be categorised under many headings. Consider for example, the finding that 35% of all representations of people with disabilities were stereotypically sinister or aggressive. Of the 23 individuals categorised as "sinister", 13 were also given a second categorisation of "aggressive". Indeed, the same 13 individuals were also placed into a third category, that of burden/outcast. These were disproportionately accounted for by, the Virginia Tech shooter, who was consistently represented as sinister, aggressive and an outcast. If one takes into account the potential for multiple categorisations, then the 41 "applications" of the sinister or aggressive category were accounted for by only 28 individuals. This is still a substantial proportion of all the 88 representations (32%), but less dramatic than the initial glance at the chart suggests.

Similarly, the finding that 38% of all incidences of stereotyping noted fell under the categories of "pitiable", "victims" or "burdens" also requires some amplification. This is due to the particular manner in which the "burden/outcast" category was implemented in practice. This category conflates two words which in practice may be understood in quite different ways. It might be assumed that to categorise someone as a burden is to cast them as a victim deserving of pity. However, the fact that Cho Seung-Hui, the Virginia Tech shooter, could account for 13 of the 17 "burden/outcast" categorisations does not suggest that this category was understood as being analogous to "pitiable" or "victim" (quite the opposite in fact). However, the fact that there was virtually no overlap between those categorised as "pitiable" and those categorised as "victims" (there was a single instance of a representation being categorised as both), means that 27 people in total accounted for all applications of the "pitiable" or "victim" categories. Thus it may be safer to conclude that only 31% (rather than 38%) of the total of 88 representations were stereotyped as pitiable or victims.

In sum, the difficulties posed by stereotypical representations for content analysis highlight the importance of the more nuanced interpretation of the sample material which is contained in the discourse analysis in section 5 of this report (below).

4.9.3 Prejudice and discrimination.

62% of representations of people with disabilities were categorised as neutral (i.e. neither prejudiced nor discriminatory). An even higher percentage of references to people with disabilities (72%) were similarly categorised as neutral. How can this be accounted for given that the coders also categorised 75% of all representations of people with disabilities as stereotyped? Is stereotyping not a form of prejudice if prejudice refers to a propensity to pre-judge a particular individual or group unfairly?

There may be less of an apparent contradiction between the large proportion of stereotyped representations and low number of discriminatory representations highlighted in the previous section since it does not necessarily follow that holding an unfair opinion of a person or group will lead an individual to actually treat that person/group unfairly (i.e. to discriminate against them). Nonetheless, the relationship between stereotyping and prejudicial and discriminatory representations needs to be examined.

At the outset, we should state that although the word "prejudice" has generally negative connotations, it is possible to be positively prejudiced towards a person or group. However, this research was primarily concerned with identifying instances of negative prejudice.

It is also important to recall difficulty of implementing workable definitions of prejudice and discrimination for content analysis, an analytical method which seeks to identify and enumerate overt incidences of whatever content type is relevant to the particular piece of research. If, for example, one is assessing the extent to which, smoking is depicted on television, it is relatively straight-forward to identify the number of instances in which a person is depicted smoking. However, the same is not true for the objective identification of prejudiced representation. Accepting the definition of prejudice as referring to "a propensity to pre-judge a particular individual or group unfairly", a prejudiced representation may describe a programme where the researcher can identify a conscious intent to prejudge on the part of the programme maker. However, it may also refer to a programme which, in the opinion of the researcher, and regardless of programme maker intent, is likely to encourage the audience to make prejudgements. In sum, identifying prejudice or discrimination involves an element of subjective opinion on the part of the researcher which is not generally compatible with the notionally objective stance implied in choosing to use content analysis as a research method.

It is for these reasons that the research took the pragmatic decision to define a prejudicial representation as one likely to encourage negative attitudes towards all people with disability. Meanwhile, a discriminatory representation was one which was likely to arouse "active" antipathy towards a particular individual or set of individuals.

Bearing these factors in mind, it was entirely possible for a particular representation to be stereotypical but neither prejudicial nor discriminatory as those terms were defined for this research. For example, an episode of "The Bill" on RTÉ 1 featured a major plot strand around a female character experiencing postnatal psychosis. Her character was coded as "Pitiable and pathetic" and "unable to cope". However, the treatment of the character was overwhelming sympathetic, portraying her as a victim of her condition and did not arouse viewer antipathy. Because of this, her representation was also coded as "neutral".

However, there was a definite relationship between stereotypes and whether a representation was coded as prejudicial or discriminatory. Of the 54 representations coded as neutral, only one was also coded as stereotypically sinister or aggressive (a female character from "Heroes" on Channel 6 who was identified as schizophrenic). Of the remaining 53 neutrally-coded representations, the majority were either stereotype-free or coded as stereotypically pitiable, victims or unable to cope.

4.10 What types of occupational role are fulfilled by a person/people with a disability? To what extent is the experience of disability relevant to the occupation of the person represented?

As noted earlier, the large number of cases where there was insufficient data to allow the coders to identify the occupation held by the individual with a disability (36% of all people with disabilities represented during programmes sampled), makes it difficult to address these questions posed here. However, the fact of so many effectively uncodable cases is suggestive of a significant finding in this regard: in more than one third of the representations of people with disabilities, the representation is so overwhelmingly focused on these individuals as disabled that it could be argued that any further description of them is considered unnecessary by programme makers. Although this maybe a tendentious assertion, it is supported by the finding in section 3.12 to the effect that disability was centrally important in 76% of all representations of people with disability. In other words, for the most part people with disabilities are featured on Irish broadcast media primarily because they are disabled. Based on the sample which informed this research, it is extremely unusual to encounter a person with a disability on television or radio in a context which is entirely unrelated to their disability.

4.11 To what extent is the experience of disability relevant to that person's role as represented?

In making these determinations about the relevance of disability to the individual characters, coders considered what the impact of removing an individual's disability would have on his/her participation in a programme. If the individual's role was essentially unchanged then their disability was coded as "incidental". If their role was substantially reduced by such a removal but not entirely removed then their disability was "relevant". If, finally, their role was essentially reduced to nothing - i.e. there would effectively be no reason left for them to feature in a programme - then their disability was coded as "central".

In section 3.8, it was reported that a third of all representations of individuals with disabilities placed them in major roles. This raised the possibility that their characters might be represented in a rounded fashion (i.e. that aspects of their lives other than their disability would also be stressed). In practice, however, in only 6% of cases was the fact of an individual's disability recorded as incidental to their representation. This points to a somewhat narrow focus on these individuals as disabled. Given this, the absence of any information about the broader lives of so many of the individuals in the sample (such as what they did for a living), noted immediately above, is scarcely surprising.

The extraordinary figures from radio (whereby 97% of cases of disability were considered centrally relevant to the representation) demand some further consideration. It is again important to recall that the absence of visual cues on radio ensured that researchers could only be aware of an individual's disability if attention was explicitly drawn to it. This was particularly significant for radio appearances, since it made it extremely unlikely that any would be categorised as "incidental". However, it could not be determined from the findings the reasons why there were no instances at all of radio appearances categorised as merely "relevant".

4.12 In which types of role are people with disabilities most and least often represented?

The key finding here was a confirmation of the relatively passive roles for the most part played by people with disabilities on-air but also of the singular focus of most representations of people with disabilities in the sample on the disability rather than the person. To fully comprehend this it is important to understand how the categories were implemented in practice by the coders.

"Presenter" and "elected representative" were relatively straight-forward categories. However, since a "person with a disability" might by definition be regarded as an "expert" on their own impairment, for the purposes of coding, "expert" was considered to refer to a person with a disability who is overtly described as an expert in a particular field, i.e. an expertise over and above that acquired purely through lived experience such as academically certified expertise.

Thus "person with a disability" which accounted for 37% of all appearances by a person with a disability was taken as referring to an individual whose role was primarily to explain their everyday, lived experience as a person with a disability but who was permitted to provide their explanation in their own right.

This in turn was distinguished from "subject" of the factual/current affairs item, which described a person with a disability who was designated as a person with a disability by a third party (i.e. the programme makers) but not permitted/able to actively represent themselves. That 61% of the sample of people with disabilities fell into this category offers striking evidence of the extent to which people with disabilities are represented in passive roles.

This treatment of people with disabilities as being the subjects of discussion (also known as 'subjectification') was also pointed to by the fact that, of the small number of non-individuated groups of people with disabilities identified in the sample, only one was regarded as playing a major role.

4.13 Is there a difference between Irish programming content and imported programming content in this regard? And if so, what is the nature of this difference?

81% of the programmes examined were produced in Ireland with the remaining 19% being imported. If this latter figure seems low given the prevalence of imported material on Irish television, it should be recalled that 54% of all programmes examined in the sample came from radio where virtually everything that is heard in Ireland is produced in Ireland. Furthermore, although Irish television is reliant on imported material to various degrees, broadcasters like RTÉ tend to schedule as much domestically-produced material as possible within the primetime hours which are the focus of the programme sample which forms the basis for this research.

The earlier findings noted that both people with disabilities and references to disability were much more likely to be found in imported material. As a consequence, although domestically-produced shows accounted for 81% of the total number of programmes looked at, they accounted for only 58% of programming featuring a person with a disability and just 69% of those programmes referring to disability only (as outlined in the chart below).

Figure 32: Distribution of appearances by people with a disability and references to disability by national origin of programming

This result needs to be considered in conjunction with the programme genre results. We have established that representations of disability are over-represented in comedy and drama and under-represented in news and current affairs programming. Given this, it is important to understand what kinds of programming dominate within imported material.

As the table above shows, drama accounts for 35% of all imported programming in the sample while comedy accounts for a further 27%. Even if lifestyle programming ("Oprah", "Ricki" etc.) only accounts for 5% of imports, the three genres identified earlier as being most likely to feature people with disabilities - drama, comedy and lifestyle - account for 67% of imports. Meanwhile the three genres least likely to feature people with disabilities - music, sports and news/current affairs, scarcely register among imports, accounting for only 2% of all imported programming.

In summary, representations of disability in the sample are over-represented in precisely those genres which are more likely to be imported. Thus the relatively weak showing of representations of disability in Irish made programming examined may be less a reflection of an absence of commitment to disability on the part of Irish broadcasters and more a reflection of the reality that Irish broadcasters rely on imported material for much of "their" drama and comedy output.