http://eros.lunarpages.com/~openpo2/cumberlandcountychapter.ssns/

MediaguidelinesreportNSJan04.pdf

 http://eros.lunarpages.com/~openpo2/cumberlandcountychapter.ssns/

MediaguidelinesreportNSJan04.doc

TECHNICAL REPORT OF

A PROVINCE WIDE SURVEY INTO THE PORTRAYAL OF SUICIDE AND MENTAL ILLNESS

The Prevention Promotion and Advocacy (PPA) Network Committee

And Anti-Stigma/Discrimination Working Group

December, 2004 

Summary

Aim:

To investigate how the non-fiction media in Canada portray both suicide and mental health, and test the feasibility of asking the general public to monitor media in the province using standardized checklists.

Method

A 6-month survey was open to all residents of Nova Scotia using two standardized checklists available on the World Wide Web and various locations in the community. Participants were asked to use the relevant checklist to determine how the media (newspaper, radio and television) reported on mental illness or suicide: submissions could identify either positive or negative coverage. In the case of print media, we also asked for a copy of the article to check inter-rater reliability.

Results

We received 414 submissions covering 366 media items. Of these, 304 discussed mental health and 62 suicide. 311 submissions came from print media, the other 55 from electronic media. Ratings showed good agreement. Most media items were of good quality on all but one dimension; details of appropriate help services were only included in 8.5% of items. Items covering suicide were three and a half times as likely to contain inappropriate content as those on mental illness, reaching statistical significance (95% CI=1.5-8.0).  

Conclusions

Media reporting of mental illness was generally of good quality and focused less on themes of crime and violence than may have been expected. However, the portrayal of suicide was significantly less appropriate. Ways of improving media reporting in this area include the dissemination of agreed guidelines and presentations to journalism students. Limitations of the study include the possibility of measurement bias as publicity surrounding the survey could have changed media behaviour, either consciously or unconsciously. The publicity may also have made subsequent interventions with the media more difficult

Recommendations

Members of the public can help monitor how mental health and suicide are portrayed in the media through the use of structured instruments.

Our findings suggest that any intervention should initially focus on the inclusion of details of how to get help, and on the reporting of suicide.

INTRODUCTION

Most studies suggest that the media's depiction of mental illness and suicide is mostly negative, inaccurate and unhelpful (1-13). This is an important finding as media portrayal may be an important influence on community attitudes towards mental health issues (2, 3). However, not all results have been negative. Two recent studies of non-fiction media found that the vast majority of items were of good quality, the only failing being the absence of information on appropriate services (14, 15).

Methodological problems with previous research include the representativeness of the samples, retrospective collection of data, and the use of secondary data, which meant that researchers could not independently evaluate the importance or definition of mental illness to the item (13). A related problem is that few studies have used standardized instruments with explicit criteria (14). Finally, although most have been restricted to non-fictional media, some have been restricted to print media, while others have included electronic media as well. These differences make it difficult to compare studies (2, 3, 14).  There has only been one study from Canada on reporting of mental illness in the last decade and a half, and this was restricted to a single newspaper in Calgary (14). Although there have been studies of the effect of reporting on suicide, there had been little in the way of systematic collection of the extent, type and quality of reporting on the issue (3).  The only Canadian studies to examine the portrayal of suicide by the media were conducted twenty years ago (3):  one was a survey of editors’ policies on the portrayal of suicide that showed that a story was most likely to be printed if it involved a prominent person, occurred in a public place, or was done by an unusual method (16); the other compared the frequency of suicide items in Toronto newspapers before and after suicides in ‘epidemic’ and ‘non -epidemic’ suicide years (ref?). Neither specifically examined the content of actual news items (17).

In response, the Prevention Promotion and Advocacy (PPA) Network Committee and Anti-Stigma/Discrimination Working Group of the Department of Health in Nova Scotia undertook a 6-month survey to monitor all non-fictional media items (newspaper, radio and television) that referred to suicide, mental health and illness. Although both groups report to the mental health programme of the Department of Health, they consist of representatives of the Department of Health, District Health Authorities, the voluntary sector, academia, people who have had mental illness, and their families.

The survey formed part of a media initiative to develop and disseminate guidelines for the portrayal of suicide and mental illness in the Province in collaboration with the media.  We wished to identify whether there were specific areas that should be the focus of locally developed guidelines. The aim of this study was therefore to establish the extent of any problem in Nova Scotia and test the feasibility of using the general public to monitor media in the province. This approach would have several advantages. The use of media monitors was too costly to be sustained over the several years that were to be covered by the Working Group’s strategy. The potential involvement of people who have had mental illness, and their families, also helped foster a sense of collective activity in addressing a tangible example of discrimination in society. This also meant that the study could include coverage of community newspapers and radio, not just Province or nationwide media.

METHOD

The survey was open to all residents of Nova Scotia using two standardized checklists available on the World Wide Web and various locations in the community. Checklists were also distributed through the Canadian Mental Health Association and the Schizophrenia Society of Nova Scotia.  Participants were asked to use the relevant checklist to determine how the media reported on issues relating to mental illness and suicide. Both checklists were derived from previous research in Australia, particularly Achieving the Balance (1). This document highlighted the importance of media portrayal of suicide and mental illness, and formed part of Australia’s National Mental Health Strategy. Participants made their quality ratings using nine-dimension instruments based on criteria in Achieving the Balance. Each dimension was phrased as a question, to which participants responded ‘yes’, ‘no’, ‘unsure’ or ‘not applicable’. To guide the participants in their decision-making, the form provided a description and examples for each dimension in lay language. The questions used to examine mental illness and suicide are shown in Table 1.  Responses of ‘yes’ and ‘no’ indicated poor and good quality, respectively, except in the case of the ‘help services’ dimension, where the direction was reversed (1). Full descriptions are available at URL: http://www.gov.ns.ca/health/mhs/media_initiative.htm          

Table 1: Questionnaire Items

To improve participation rates, each entry was eligible for a cash prize (up to $2000) in each of the categories of mental illness and suicide. Entries could identify both positive and negative portrayals, give details of where and when the item appeared and, in the case of print media, had to be accompanied by a copy of the article.

We checked inter-rater reliability of participants’ ratings for any item when there was more than one entry. We also checked the consistency of coding by assessing the agreement between ratings for submissions that were accompanied by a copy of the article, with those made independently by one of the authors (JD).

RESULTS

Characteristics of submissions

We received 414 submissions that covered 366 media items. Of these, 304 discussed mental illness, and 62 suicide. Three hundred and eleven submissions came from print media, the other 55 from electronic media (radio and television) (Figure 1).

Figure 1:

Most submissions were of media based in the Halifax area (Figure 2). We used intraclass correlation coefficients (ICC) to assess the agreement between submissions when two or more covered the same item, and between ratings for submissions that were accompanied by a copy of the article, with those of one of the authors. Ratings showed good agreement: the ICC for the mental illness ratings was 0.93 (p=0.001), while that for suicide was 0.96 (p=0.001). 

Figure 2

Less than a third of items had headlines that were inaccurate or inconsistent with the focus of the item, or that were unnecessarily dramatic or sensationalised (Table 2 & Figure 3).  Examples included words such as ‘suffering’, ‘victim’,  ‘disturbed’ and ‘afflicted’;  or phrases like 'demons of mental health problems', ‘punished by the torment in her head’, ‘lived a schizophrenic existence’,  ‘emotional breakdowns’, ‘stress taking its toll’ and ‘woman needs a babysitter who isn't psycho’.

More extreme examples of sensational and dramatic language were less common, but included ‘dangerous schizophrenic’, ‘wacko standoff’, ‘psychotic killers’, and ‘unstable madman’.  Less than a fifth used medical terminology inappropriately such as the use of ‘schizophrenic mix’ when referring to Ottawa or ‘psychotic’ when meaning personality disorder.

However, over 40%

of items used outdated, negative or inappropriate language (Table 2 & Figure 3). Examples included: ‘cracked up’, ‘nut job, ‘head case’, ‘basket case, ‘dimwits’,  'deranged',  ‘mad as a hatter’ and ‘drive me nuts’. A similar proportion reinforced stereotypes about mental illness with frequent references to violence, crime and unpredictability. Some items also made generalizations about people with mental illnesses, such as saying that ‘panhandling was the preserve of the mentally ill’

Table 2: Questionnaire results

Figure 3

The majority of items did not disclose that a particular individual had a mental illness (Table 2 & Figure 3). Of the remaining 28%, it was difficult to determine whether the individual’s consent had been given: these were often court reports.

Portrayal of suicide

About a half used inappropriate language (Table 2 & Figure 4). Examples included the use of sensationalist terminology to describe the prevalence of suicide in a community as an ‘epidemic’. A similar proportion used the word ‘suicide’ in the headline  . Forty two percent discussed the method used in the suicide or suicide attempt such as jumping from one of the harbour bridges in Halifax. 

Reporting was more generally appropriate across the other dimensions. Less than a third of items were located on the front page of a newspaper or presented as the leading item during a broadcast; and only a fifth depicted the suicide scene, or photographs of the deceased or family members without permission (Table 2 & Figure 4). Only one in eight items covered celebrity suicide. 

Worryingly, over a quarter included interviews with the bereaved, even though they are at heightened risk of suicide themselves (Table 2 & Figure 4). Also of concern is that about one third of items portrayed suicide as merely a social phenomenon.

Figure 4

  Overall quality

We assessed overall quality in two ways for the combined sample of both mental illness and suicide ratings: the presence of information on obtaining help and the absence of inappropriate content as above. Only 31 items (8.5%) included contact details for help services (e.g. information about telephone help lines, internet sites, community organizations or support groups); 91.5% did not. Items on suicide tended to have more information on obtaining help although this did not reach statistical significance (Figure 5)

Only 28% of items (n=101) were free of any inappropriate content; 72% contained some inappropriate content. Items covering suicide were three and a half times as likely to contain inappropriate content (55/62) as those on mental illness (210/304), this reaching statistical significance (95% CI= 1.5-8.0) (Figure 5).   Print media were twice as likely to contain inappropriate content (232/311) compared to electronic media (33/55) also reaching statistical significance  (95% CI= 1.1-3.6) (Figure 5). This finding was largely explained by the fact that coverage by the Canadian Broadcasting Corporation, an electronic media source, was less than half as likely to have inappropriate content (23/42) than other media (242/324), a statistically significant result (95% CI= 0.2-0.8) (Figure 5).

Figure 5

Response of media

Some print media were defensive and hostile about even the idea of a survey of how mental illness and suicide were portrayed.  An editorial in the National Post, entitled ‘Are these guys nuts?’ described the initiative as a ‘witch hunt for witches that don’t exist instigated by a committee of overwrought activists who have convinced themselves that the lamp of wisdom has been only given to them’ (18).

DISCUSSION

The media are an important source of information on both suicide and mental illness: importantly, reporting can have a direct impact on both community attitudes towards mental illness and suicide attempts, with the strongest evidence for non-fictional media (3, 5).  Our study was therefore designed to examine how suicide and mental health and illness are portrayed in the media in a standardized way across a whole jurisdiction. To our knowledge this is the first time mental illness and suicide have been considered together.

Limitations

This study has several limitations. Firstly, we do not know how representative our submissions were of media items in Nova Scotia during the study. Most of the items were from newspapers rather than the electronic media. Of newspaper articles that were submitted, most came from the Chronicle Herald, which is a provincially distributed newspaper, rather than community newspapers. Secondly, there may have been an information bias through the self-selection of raters.  Participants may have had particularly strong views about mental illness and suicide; and their ratings may not be representative of community views. Another source of bias is the instrument itself. We did try to eliminate the possibility of this by testing the inter-rater reliability between submitted ratings and those of one of the authors (JD). However, we were only able to check the inter- reliability on submissions from print media, and not those from television or radio.

Other limitations of this study include the comparatively low number of articles on suicide, and that our sample size for items on mental illness was also much smaller than that of a similar study in Australia (15). However our numbers are  not lower than those of most other studies that have considered media coverage of mental illness (4,14).  Our study could also have been subject to measurement bias in that the publicity surrounding the study may have made the media more sensitive about how they were reporting their stories, whether consciously or unconsciously.

In spite of these shortcomings, it is reassuring to note that our findings were very similar to those from Australia, using the same standardised checklist,  which were rated by a media retrieval agency (15). This suggests that our results may not be solely explained  by observation bias.

We were surprised by the strong negative reaction of some reporters, and a final limitation of this study is that our methodology may have damaged future relationships with the media. This is turn may affect the acceptance of voluntary guidelines on reporting suicide and mental illness.

Strengths

We have shown that it is possible to conduct a media survey using members of the public rather than media monitoring agencies, professional staff or research workers, who are all relatively expensive. This methodology could therefore be more used to measure long-term trends, such as changes following the dissemination of media guidelines. We are also able to make international comparisons as we used standardised checklists derived from Australia.   Up to now, it was unknown whether any differences in the portrayal of mental illness between jurisdictions could solely be explained by the use of different instruments. We have also been able to compare the portrayal of suicide and mental illness that will help to inform a comprehensive strategy that addresses both.

Comparisons with other studies

Our results on the portrayal of mental illness are very similar to those from Australia that used a similar checklist and showed that media reporting on mental illness was predominately positive. These are in contrast to other research where media reporting of mental illness has mainly been negative (2, 4-13). As in the Australian study, media items were of good quality in eight out of the nine dimensions measured. These findings mirror a survey of one of the two local newspapers in Calgary where positive stories outnumbered negative ones by a factor of two to one, even before an intervention to engage reporters and thereby increase positive stories on mental illness (14).

The results for suicide were less encouraging and this is consistent with the existing literature where, in contrast to the mixed findings for mental illness, the reporting of suicide is uniformly poor (3).  By using a similar methodology for both mental illness and suicide, we have shown the items on suicide are three and a half times more likely to have inappropriate content. This finding reached statistical significance in spite of the relatively low number of items covering suicide and is of particular concern given the relationship between reporting on suicide and subsequent attempts.

Another area of concern was the absence of information on where to get help in over 90% of items; this is almost identical to the findings from the Australian study, suggesting that this problem is not restricted to Nova Scotia (15).

Implications

These results will  aide in the development and agreement of media guidelines to cover both suicide and mental illness. Given that it is important to develop guidelines with the people who will use them, we will try to involve editors and opinion makers in the provincial media in their development (19). There is evidence from both Australia and Canada that local media may be easier to influence than national or international media (14, 20).  A longer-term strategy is to influence the education of journalism students. Given our findings, we will particularly focus on the inclusion of details of how to get help, and on suicide. Other areas of concern include outdated, negative or inappropriate language, and the reinforcement of negative stereotypes.

There is no reason why discriminatory and stigmatising language directed at those with mental health problems should be tolerated any more than when it is directed on grounds of race, gender, religion or sexual orientation. We presume journalists or broadcasters would not call a depressed relative or friend with suicidal ideas, a 'fruitcake', or tell them of ways of committing suicide in graphic detail.  Do they not owe the same duty of care to their readers, listeners and viewers? 

REFERENCES

1. Department of Health and Aged Care. Achieving the balance: A resource kit

for Australian media professionals for the reporting and portrayal of suicide and mental illness. Canberra; Commonwealth Department of Health and Aged Care; 1999.

2. Department of Health and Ageing. Reporting suicide and mental illness. Canberra; Commonwealth Department of Health and Ageing; 2002.

3. Pirkis J, Warwick Blood R. Suicide and the media: a critical review. Canberra; Commonwealth Department of Health and Aged Care; 2001.

4. Francis C, Pirkis J, Dunt D, Warwick Blood R. Mental health and illness in the media: a review of the literature. Canberra; Commonwealth Department of Health and Aged Care; 2001.

5. Day DM, Page S. Portrayal of mental illness in Canadian newspapers. Can J Psychiatry 1986;31:813-817.

6. Matas M, el-Guebaly N, Harper D, Green M, Peterkin A. Mental illness and the media: part II. Content analysis of press coverage of mental health topics. Can J Psychiatry 1986;31:431-433.

7. Allen R, Nairn RG. Media depiction of mental illness: an analysis of the use of dangerousness. Aust N Z J Psychiatry 1997;31:375–381.

8. Nairn R. Does the use of psychiatrists as sources of information improve media depictions of mental illness? A pilot study. Aust N Z J Psychiatry 1999;33:583–589.

9. Wilson C, Nairn R, Coverdale J, Panapa A. Constructing mental illness as dangerous: a pilot study. Aust N Z J Psychiatry 1999;33:240–247.

10. Thornton J, Wahl O. Impact of a newspaper article on attitudes toward mental illness. Journal of Community Psychology 1996;24:17–25.

11. Steadman H, Cocozza J. Selective reporting and the public’s misconceptions of the criminally insane. Public Opinion Quarterly 1977;41:523–533.

12. Angermeyer MC, Matschinger H. The effect of violent attacks by schizophrenic persons on the attitude of the public towards the mentally ill. Soc Sci Med 1996;43:1721–1728.

13. Coverdale J, Nairn R, Claasen D. Depictions of mental illness in press media: a prospective national sample. Aust N Z J Psychiatry 2002;36:697–700.

14. Stuart H. Stigma and the daily news: evaluation of a newspaper intervention. Can J Psychiatry 2003;48:651-656.

15. Francis C, Pirkis J, Warwick Blood R, Dunt D, Burgess P, Morley B, and others. The portrayal of mental health and illness in Australian non-fiction media. Aust N Z J Psychiatry 2004;38:541-546.

16. Pell, B, and Watters, D Newspaper policies on suicide stories. Canada’s Mental Health 1982;30:8–9

17. Littmann, SK Suicide epidemics and newspaper reporting. Suicide and Life Threatening Behavior 1985;15:43–50.

18. Editorial. Are these guys nuts? National Post, March 4, 2004, A15.

19. Davies J, Freemantle N, Grimshaw J, Hurwitz B, Long A, Russell I, Sheldon T, Wallace S, Watt I. Implementing clinical practice guidelines. Effective Health Care 1994;1(8):1-12

20. Slaven J, Kisely S. The Esperance primary prevention of suicide project Aust N Z J Psychiatry 2002; 36: 617-21

Contributors

Author: Dr. Stephen Kisely, Departments of Psychiatry, Community Health & Epidemiology, Dalhousie University

Co-author: Jennifer Denney, Department of Psychiatry, Dalhousie University

Data entry: Irene Thomas

The Prevention Promotion and Advocacy (PPA) Network Committee

Karen Briand Sandra Nicholas

Joe Bruce Wendy Nickerson

Doug Crossman Tony Prime, Project Officer

Dr Saul Hirsch Shelly Saunders

Dr Stephen Kisely, Chair Bethanna Sullivan

Dr Patrick McGrath Pam Swainson

Dr David Whitehorn Carol Tooton

Tom Freeman Francine Vezine

Sarah Johnson Debbie Wheeler

Duffy McDonald Julian Young

Linda MacDonald

The Anti-Stigma/Discrimination Working Group

John Brazier Roy Muise

Mathew Claener Fran Nunn

Denton Conrad Tony Prime, Project Officer

Myrtle Corkum Kim Rideout

Jack Langley Nicole Thimot-Ginnette

Dr. Stephen Kisely, Chair Sandra Tiller

Duffy McDonald Carol Tooton