2
comes to the possible processes linking cannabis use with later suicidal behaviours, it was explicitly stated that these are Ôby no means clearÕ (p. 401), but that it seems reasonable to suggest that they either must be related to neurophysiological pathways or social processes. A few studies pointing in both directions were referred. I of course agree with Dr Satya- narayana in that more research is warranted in this area. Willy Pedersen University of Oslo – Sociology and Human Geography, PO Box 1096, Blinderen Oslo 0317, Norway E-mail: [email protected] The need to rely on evidence not ideology in stigma research DOI: 10.1111/j.1600-0447.2009.01362.x Numerous studies have found that both diagnostic labelling and the espousal of bio-genetic causal beliefs by the public are related to increased fear and prejudice towards people with mental health problems including schizophrenia. Two recent reviews (1, 2) pointed out that destigmatization programmes designed to teach the public an illness model are therefore probably increasing rather than reducing stigma. In a recent paper, proponents of educating the public to use diagnostic labels and to adopt illness-type causal beliefs tested the conclusions of these two reviews. Specifically, they Ôexam- ined the hypotheses that stigmatizing attitudes are increased by use of psychiatric labels, by conceptualization of symptoms as a medical illness and by belief in genetic causesÕ (3, p. 315). Using a ÔschizophreniaÕ vignette they found, like many previous studies, that Ôbelief in dangerousness was predicted by medical illness conceptualizations and by genetic causal attri- butionsÕ (p. 315). They then, however, attempt to interpret these findings in a way that does not challenge their basic presuppositions. These presuppositions are reflected in their repeated attribution of Ômental health literacyÕ to that minority of the public who agree with biological psychiatry about what causes ÔschizophreniaÕ (2). Their efforts to avoid concluding that some of their data clearly confirm what they set out to question rests on the argument that ÔBiomedical conceptualizations are not the major causes of stigma; rather it is the behaviour associated with mental illness and the belief that this is because of personal weaknessÕ. They claim to have found that the behaviours in the vignette and a belief in personal weakness as a cause are both Ômore importantÕ factors than illness conceptualizations or genetic causal beliefs. In fact, their study provides no evidence to support either of these claims. Claiming to have demonstrated that the behaviours in the vignettes were Ômore importantÕ is false. There were no analyses comparing the differential effects of the various vignettes with the effects of either causal explanations or illness conceptual- ization. The researchers argue that it is Ôparticularly the contrast between schizophrenia and depressionÕ which supports their claim of the greater importance of behaviours over other factors. Again, however, no relevant analyses were reported. The four schizophrenia vignettes (early–chronic; male–female) were analysed only in comparison with each other, as was the case for the four depression vignettes. Indeed, the only significant differences (within each diagnostic category) in dangerousness beliefs that related to the vignettes were to do with gender not behaviour. Thus, their conclusion about the high importance of the behaviours in the vignettes (which seems to place responsibility for prejudice on the target of the prejudice) has no support in their study. A more balanced interpretation might have men- tioned that, for the schizophrenia vignette, those labelling the person ÔschizophrenicÕ were 1.47 times more likely (P = 0.012) to believe they are dangerous. How can the behaviour be the most ÔimportantÕ determinant of dangerousness belief when even those labelling the person in the depression vignette as ÔschizophrenicÕ were 3.9 times more likely (P = 0.088) than others to believe the person was dangerous? The second factor claimed to be Ômore importantÕ than bio- genetic beliefs and illness conceptualizations was Ôweakness of characterÕ. No analyses were reported to test whether the relationship of dangerousness to Ôweakness of characterÕ was significantly greater than its relationships to illness conceptu- alization or genetic causal explanations. All three were signif- icant at the P < 0.01 level at least. Jorm and Griffiths bemoan the use of the term Ômental illnessÕ in stigma research because it is too ÔdiverseÕ. This misses the point, demonstrated by studies all over the world (including their own), that the ÔillnessÕ framework, with its multiple connotations, such as pessimism about recovery, loss of control, permanent biological abnormalities or deficits (4), etc., is absolutely key to understanding why decades of destigmatization programmes seem to have had little effect. We suggest, therefore, that a more empirically justified conclusion about the data in the Jorm and Griffiths paper would read: ÔThe following factors were strongly related to dangerousness beliefs: belief in Ôweakness of characterÕ as a cause, belief in genetic causes, conceptualization of symptoms as a medical illness, and male gender of the target. All require attention in the design of future destigmatization programmesÕ. While waiting for the actual evidence about illness beliefs to overcome ideology we do all seem able to agree on the importance of increasing contact with the Ôout-groupÕ, directly or via films, etc., when trying to reduce fear and prejudice (1–3, 5). John Read 1 Nick Haslam 2 Emma Davies 3 1 Psychology Department, The University of Auckland, New Zealand 2 Psychology Department, The University of Melbourne, Australia 3 Institute of Public Policy, Auckland University of Technology, New Zealand E-mail: [email protected] Letter to the Editor 412

The need to rely on evidence not ideology in stigma research

Embed Size (px)

Citation preview

Page 1: The need to rely on evidence not ideology in stigma research

comes to the possible processes linking cannabis use with latersuicidal behaviours, it was explicitly stated that these are �byno means clear� (p. 401), but that it seems reasonable tosuggest that they either must be related to neurophysiologicalpathways or social processes. A few studies pointing in bothdirections were referred. I of course agree with Dr Satya-narayana in that more research is warranted in this area.

Willy PedersenUniversity of Oslo – Sociology

and Human Geography,PO Box 1096,

Blinderen Oslo 0317,Norway

E-mail: [email protected]

The need to rely on evidence not ideology instigma research

DOI: 10.1111/j.1600-0447.2009.01362.x

Numerous studies have found that both diagnostic labellingand the espousal of bio-genetic causal beliefs by the public arerelated to increased fear and prejudice towards people withmental health problems including schizophrenia. Two recentreviews (1, 2) pointed out that destigmatization programmesdesigned to teach the public an illness model are thereforeprobably increasing rather than reducing stigma.

In a recent paper, proponents of educating the public to usediagnostic labels and to adopt illness-type causal beliefs testedthe conclusions of these two reviews. Specifically, they �exam-ined the hypotheses that stigmatizing attitudes are increased byuse of psychiatric labels, by conceptualization of symptoms asa medical illness and by belief in genetic causes� (3, p. 315).Using a �schizophrenia� vignette they found, like many

previous studies, that �belief in dangerousness was predicted bymedical illness conceptualizations and by genetic causal attri-butions� (p. 315). They then, however, attempt to interpretthese findings in a way that does not challenge their basicpresuppositions. These presuppositions are reflected in theirrepeated attribution of �mental health literacy� to that minorityof the public who agree with biological psychiatry about whatcauses �schizophrenia� (2).Their efforts to avoid concluding that some of their data

clearly confirm what they set out to question rests on theargument that �Biomedical conceptualizations are not themajor causes of stigma; rather it is the behaviour associatedwith mental illness and the belief that this is because ofpersonal weakness�. They claim to have found that thebehaviours in the vignette and a belief in personal weaknessas a cause are both �more important� factors than illnessconceptualizations or genetic causal beliefs. In fact, their studyprovides no evidence to support either of these claims.Claiming to have demonstrated that the behaviours in the

vignettes were �more important� is false. There were no analysescomparing the differential effects of the various vignettes withthe effects of either causal explanations or illness conceptual-ization. The researchers argue that it is �particularly thecontrast between schizophrenia and depression� which supportstheir claim of the greater importance of behaviours over otherfactors. Again, however, no relevant analyses were reported.The four schizophrenia vignettes (early–chronic; male–female)were analysed only in comparison with each other, as was thecase for the four depression vignettes. Indeed, the onlysignificant differences (within each diagnostic category) indangerousness beliefs that related to the vignettes were to dowith gender not behaviour.

Thus, their conclusion about the high importance of thebehaviours in the vignettes (which seems to place responsibilityfor prejudice on the target of the prejudice) has no support intheir study. A more balanced interpretation might have men-tioned that, for the schizophrenia vignette, those labelling theperson �schizophrenic� were 1.47 times more likely (P = 0.012)to believe they are dangerous. How can the behaviour be themost �important� determinant of dangerousness belief wheneven those labelling the person in the depression vignette as�schizophrenic� were 3.9 times more likely (P = 0.088) thanothers to believe the person was dangerous?The second factor claimed to be �more important� than bio-

genetic beliefs and illness conceptualizations was �weakness ofcharacter�. No analyses were reported to test whether therelationship of dangerousness to �weakness of character� wassignificantly greater than its relationships to illness conceptu-alization or genetic causal explanations. All three were signif-icant at the P < 0.01 level at least.Jorm and Griffiths bemoan the use of the term �mental

illness� in stigma research because it is too �diverse�. This missesthe point, demonstrated by studies all over the world (includingtheir own), that the �illness� framework, with its multipleconnotations, such as pessimism about recovery, loss ofcontrol, permanent biological abnormalities or deficits (4),etc., is absolutely key to understanding why decades ofdestigmatization programmes seem to have had little effect.We suggest, therefore, that a more empirically justified

conclusion about the data in the Jorm and Griffiths paperwould read: �The following factors were strongly related todangerousness beliefs: belief in �weakness of character� as acause, belief in genetic causes, conceptualization of symptomsas a medical illness, and male gender of the target. All requireattention in the design of future destigmatization programmes�.While waiting for the actual evidence about illness beliefs to

overcome ideology we do all seem able to agree on theimportance of increasing contact with the �out-group�, directlyor viafilms, etc.,when trying to reduce fear andprejudice (1–3, 5).

John Read1

Nick Haslam2

Emma Davies31Psychology Department, The University of Auckland,

New Zealand2Psychology Department, The University of Melbourne,

Australia3Institute of Public Policy, Auckland University of Technology,

New ZealandE-mail: [email protected]

Letter to the Editor

412

Page 2: The need to rely on evidence not ideology in stigma research

References

1. Read J, Haslam N, Sayce L, Davies E. Prejudice and schizo-phrenia: a review of the �mental illness is an illness like anyother� approach. Acta Psychiatr Scand 2006;114:303–318.

2. Read J. Why promoting biological ideology increasesprejudice against people labelled ‘‘schizophrenic’’. AustrPsychologist 2007;42:118–128.

3. Jorm A, Griffiths K. The public�s stigmatizing attitudestowards people with mental disorders: how important arebiomedical conceptualizations? Acta Psychiatr Scand2008;118:315–321.

4. Haslam N, Ernst D. Essentialist beliefs about mental dis-orders. J Soc Clin Psychol 2002;21:628–644.

5. Laroi F, van der Linden M. The effects of a documentary filmon reducing stigmatisation about schizophrenia. PsychosisPsychol Soc Integr Approaches 2009;1:61–72.

DOI: 10.1111/j.1600-0447.2009.01361.x

Reply

What the evidence on stigma shows: a reply to Read et al.

We examined a large data set to test a range of hypothesesderived from John Read�s theories of what causes stigma. Mostof these hypotheses were not confirmed. Read et al.�s letterraises several issues with regard to these findings.

Behaviour in vignettes as an important predictor of stigma

We concluded that the behaviour in the vignette was animportant predictor based on the descriptive statistics pre-sented in Table 1 of our paper (1). However, Read et al. arecorrect that we did not carry out any statistical tests on thesedata. We have now done this. Regressions were carried outexamining the effects of type of vignette (depression withsuicidal thoughts, early schizophrenia and chronic schizo-phrenia, with depression as the reference category), gender ofvignette, labels applied to the vignettes (depression, schizo-phrenia, nervous breakdown, mental illness, psychological ⁄mental ⁄ emotional problems, stress, has a problem) andmedical illness belief. Age and gender were covariates in theregressions. This analysis showed that social distance wasincreased by the chronic schizophrenia vignette (B = 1.43,P < 0.001), the early schizophrenia vignette (B = 0.57, P <0.001) and labelling the vignette as �has a problem� (B = 0.49,P = 0.003). It was decreased by labelling the vignette as�depression� (B = )0.62, P < 0.001) and by a medical illnessbelief (B = )0.60, P < 0.001). Belief in dangerousness wasincreased by the early schizophrenia vignette (OR = 1.80,P < 0.001), the depression with suicidal thoughts vignette(OR = 1.73, P < 0.001), and by labelling the vignette as�schizophrenia� (OR = 1.43, P = 0.007). It was decreased bylabelling the vignette as �depression� (OR = 0.70, P = 0.002)and by female gender in the vignette (OR = 0.75, P = 0.001).These findings confirm that the behaviours in vignettesare major determinants of both social distance and beliefin dangerousness. The only support for Read�s hypothesesis that the label of �schizophrenia� is associated withdangerousness, but the odds ratios for the vignettes arehigher. The data do not support the view that medicalconceptualizations or psychiatric labels in general increasestigma; indeed, in the case of the label �depression�, theyreduce it.

Weakness of character as a causal explanation leading tostigma

We examined nine causal explanations of mental disorders.Each of these was examined in four analyses (in relation to twomeasures of stigma across two mental disorders). Weakness of

character was the only causal explanation significant in all fouranalyses. Inherited or genetic was significant in only one of thefour analyses, as was death of someone close. Other studies haveconsistently supported weakness as a correlate of socialdistance, whereas genetic and psychosocial explanations havenot had consistent associations across studies (2).

Contact and stigma

Read et al. conclude that contact is a way to reduce stigma.However, in our data contact was associated with reducedsocial distance, but not with less belief in dangerousness.Contact does not necessarily reduce stigma, as shown bystudies of health professionals, who often report stigmatizingattitudes despite frequent contact (2).

Should campaigns promote illness conceptualizations?

We believe that the public need to be able to recognize mentaldisorders and seek appropriate professional help and that thisknowledge should be promoted in campaigns. The reason isthat, even in developed countries, many people do not getprofessional help or have long delays before receiving help.This lack of help-seeking can have long-term adverse conse-quences in a person�s life. A major reason for this delay is afailure to conceptualize the problem as a treatable mentaldisorder (3). Using a psychiatric label for the problem appearsto activate a schema for appropriate help-seeking (4).

Anthony F JormORYGEN Research Centre, University of Melbourne,

Melbourne, AustraliaKathleen M Griffiths

Centre for Mental Health Research,Australian National University, Canberra, Australia

E-mail: [email protected]

References

1. Jorm A, Griths K. The public�s stigmatizing attitudes to-wards people with mental disorders: how important arebiomedical conceptualizations? Acta Psychiatr Scand 2008;118:315–321.

2. Jorm A, Oh E. Desire for social distance from people withmental disorders: a review. Aust NZ J Psychiatry 2009;43:185–202.

3. Thompson A, Issakidis C, Hunt C. Delay to seek treatment foranxiety and mood disorders in an Australian clinical sam-ple. Behav Change 2008;25:71–84.

4. Wright A, Jorm AF, Harris MG, McGorry PD. What�s in aname? Is accurate recognition and labelling of mental dis-orders by young people associated with better help-seekingand treatment preferences Soc Psychiatry Psychiatr Epi-demiol 2007;42:244–250.

Letter to the Editor

413