Scottish Government

CHAPTER EIGHT: ATTITUDES TOWARDS SPECIFIC SYMPTOMS OF MENTAL ILL-HEALTH

8.1 Chapter 7 considered attitudes towards mental ill-health in general. This chapter explores attitudes towards specific symptoms of mental ill-health, and in particular people exhibiting symptoms associated with three different conditions: depression, schizophrenia and stress.

Mental ill-health 'scenarios'

8.2 Each respondent in the survey was presented with a scenario depicting a person with symptoms of either depression, schizophrenia or stress. The person in the scenario was either male or female (named Robert or Shona) giving six scenarios in total. The full text of the six scenarios is provided in Annex J. Without being given a diagnosis of the symptoms described, respondents were asked a series of questions about the person in the scenario and his/her symptoms (scenarios were randomly assigned to respondents). The questions focused on the likely cause(s) of the symptoms, possible sources of help, the likelihood of Robert/Shona harming themselves or others, and the extent to which respondents would be willing to interact with them. At the end of the section, respondents were asked to say what condition they thought was being described in the scenario.

8.3 Given that there were six scenarios, each was considered by a relatively small sub sample of respondents. This point should be borne in mind when considering the findings.

8.4 The following section provides an overview of the findings, before focusing specifically on respondents' willingness to interact with the person depicted in the scenarios under a range of circumstances. A detailed question by question analysis of all findings pertaining to the scenarios can be found in Annex J.

Overview of attitudes towards symptoms of depression, schizophrenia and stress

8.5 The most likely causes of the symptoms described in all scenarios were thought to be stressful or disturbing events in Robert's/Shona's life and the circumstances in which he/she lives. Those who were shown the depression or schizophrenia scenarios were more likely than other respondents to associate the symptoms with a chemical imbalance in the brain, while those shown the stress scenario were more likely to mention Robert's/Shona's own character or personality.

8.6 Respondents who were shown the male versions of the depression or schizophrenia scenarios were more likely than those shown the female version to mention the way he was brought up. Furthermore, those who were shown the male depression scenario were also more likely to feel the symptoms were Robert's own fault. Meanwhile, those shown the female version of the stress scenario were more likely than those shown the corresponding male version to associate the symptoms with Shona's upbringing.

8.7 Between 2004 and 2006, there were significant increases across all scenarios in the proportions of respondents who felt that the symptoms may be associated with Robert's/Shona's own character or personality.

8.8 Family doctors, family members and qualified counsellors were seen as the best sources of help for the people in each of the scenarios. However, half of those shown the schizophrenia scenarios mentioned a psychiatrist. Those shown the male versions of the schizophrenia or stress scenarios were more likely to mention family members, while those shown the female version were more likely to mention friends and neighbours.

8.9 Whichever version of the scenario respondents were shown, a majority felt that the best place for Robert/Shona to live was in their own home with support from family members or friends. However, a significant minority of those shown the schizophrenia scenarios thought they should live in special housing with professional support in the community.

8.10 The person in the scenario depicting symptoms of schizophrenia was judged to be more likely to harm him/herself than the person experiencing depression, and the person in the stress scenario was assessed as being least likely to self-harm. These findings are consistent with a recent British survey of non-fatal suicidal behaviour. Meltzer et al. (2002) reported that those individuals who had been diagnosed with schizophrenia were most likely to engage in self-harm (compared with other diagnostic categories), with approximately 50% having self-harmed at some time in their lives. Few people thought Robert/Shona was likely to harm others but, again, it was the person experiencing symptoms of schizophrenia who was felt to be most likely to cause harm to others. Respondents who considered the male version of the scenario were more likely than those who considered the female version to feel the person depicted was likely to harm others. There have been few changes in these results over the three waves of the survey but among those shown the female version of the schizophrenia scenario, the percentage suggesting that Shona might harm others has fallen by 10 percentage points between 2004 and 2006.

8.11 For all scenarios, majorities said that they would be willing to interact with Robert/Shona under a range of circumstances, including doing them a favour, making friends with them, moving next door to, or spending an evening socialising with them. However, smaller proportions were willing to have Robert/Shona marry into the family and fewer than half in each case said that they would allow them to provide childcare for someone in their family.

8.12 Willingness to interact with Robert/Shona was highest among respondents who were shown the stress scenarios and lowest among those who were shown the schizophrenia scenarios. For each of the scenarios, respondents were generally more willing to interact with a female displaying the symptoms depicted than with a male showing the same symptoms, although whether this is a general feature of people's willingness to interact with the two sexes, or something specific to people suffering mental health problems, we cannot say.

8.13 Between 2004 and 2006, respondents have become somewhat less willing to interact with the people described in the scenarios. This is particularly evident in terms of: the proportion of those shown the depression scenario who are willing to start working closely with Robert/Shona, or have them marry into the family; the proportion shown the schizophrenia scenario who are willing to make friends with Robert/Shona; and the proportion shown the stress scenario who are willing to move next to Robert, to do him a favour or to make friends with Shona.

8.14 Around half of respondents shown the schizophrenia scenario and a third of those shown the depression scenario felt Robert's/Shona's freedoms and rights might have to be limited because of their illness. Meanwhile, a lower proportion (around one in five) of those considering the stress scenario felt this was likely.

8.15 While majorities of those shown the depression scenarios were able to attribute the symptoms to the condition correctly, this was not the case for either the schizophrenia or stress scenarios. Indeed, almost half of those considering the stress scenarios thought that Robert/Shona was exhibiting symptoms of depression. Likewise, a significant proportion of those shown the schizophrenia scenario gave diagnoses of depression, a nervous breakdown or a personality disorder. These findings were consistent with those from the 2004 survey.

Willingness to interact with the person in the scenarios

8.16 The key measure included in the scenarios section was the item in which respondents were asked how willing they would be to interact with Robert/Shona on a number of different levels, namely:

• Move next door to Robert/Shona
• Spend an evening socialising with him/her
• Make friends with him/her
• Start working closely with him/her
• Have him/her marry into the family
• Do him/her a favour if he/she asked you to
• Have him/her provide childcare for someone in your family (e.g. babysitting, childminding)

8.17 Table 8.1 presents the aggregate results for the six scenarios. It shows that a majority of respondents would be willing to move next door to Robert/Shona, spend an evening socialising with him/her, make friends with him/her, start working closely with him her and do him/her a favour. However, only two in five would be willing to have Robert/Shona marry into the family and only around half this proportion would be willing to have Robert/Shona provide childcare for someone in their family.

Table 8.1: Willingness to interact with person in the scenarios - aggregate results for 2006

Source: Ipsos MORI

8.18 Analyses were undertaken to explore to what extent responses on the 7 interaction measures varied by a) survey wave b) scenario (both in terms of the gender of the person in the scenarios and his/her symptoms), and c) socio-demographic characteristics of the respondent. The principle statistical technique used for the analyses was ANOVA (analysis of variance) which simultaneously compares the mean responses of a number of sub-groups to identify whether these groups respond identically or otherwise.

Variation by survey wave

8.19 The inter-wave analysis aimed to assess the extent to which willingness to interact with the person in the scenario has varied over the 2002, 2004 and 2006 surveys. Table 8.2 shows the mean responses for each of the 7 interaction measures by survey wave. As for all tables present below, the means range from 1 to 5, with 1 indicating a response of 'very unwilling', and 5 a response of 'very willing'.

Table 8.2: Willingness to interact with person in the scenarios - mean responses

Source: Ipsos MORI

8.20 The analysis revealed that there is no consistent pattern of variation or trend across the three waves of the survey; rather, the picture appears to be one of short term fluctuation. More specifically, between 2002 and 2004, willingness to make friends with Robert/Shona, start working closely with him/her and have him/her marry into the family increased. For the remaining interaction measures, there were no differences between the two waves. Between 2004 and 2006, willingness to interact with Robert/Shona, decreased for all 7 measures, with most of the change significant at the 0.01% level. In general, however, the decreases simply cancelled out the increases observed between 2002 and 2004. In other words, there is little variation between the 2002 and 2006 results - the only exception being in the case of willingness to do Robert/Shona a favour which was lower in 2006 than in 2002.

8.21 The decline in willingness to interact with Robert/Shona between 2004 and 2006 is somewhat surprising given the consistency of attitudes towards mental health between 2004 and 2006 highlighted elsewhere throughout this report. It may be that people today understand more about the symptoms of mental health problems but do not yet feel equipped to cope with them. Alternatively, and as noted in Chapter 7, our exploratory examination of the print media press clippings may point to the potential effects of negative representations of mental ill-health on people's behavioural intentions. A third possible explanation is methodological and relates to changes in the ordering of questions in the scenarios section for the 2006 survey. In 2004, the 'willingness to interact' questions were preceded by a question asking whether the people in the scenarios should have the same rights at work as others. This may have primed respondents to think in terms of equal rights and thus to give more sympathetic responses in the willingness to interact questions. For the 2006 survey the 'rights' question was asked after the 'willingness to interact' questions, thereby eliminating this potential order effect. The question was moved in 2006 following a decision to change the wording of the item to: "How likely or unlikely do you think it is that Robert/Shona's freedoms and rights might have to be limited because of their illness?" There were concerns that the negative slant of this question may have encouraged respondents to give more negative answers to the 'willingness to interact' items, had it preceded those items. Consequently, it may be that in 2006 we are seeing the most 'honest' response to these questions.

Variation by scenario

8.22 A similar analysis was undertaken to explore the extent to which willingness to interact with someone with mental health problems varied depending on the sex of that person and the nature of his/her symptoms.

8.23 As table 8.3 (below) shows, respondents were consistently more willing to interact with a woman displaying symptoms of mental ill-health, than with a man displaying the same symptoms. This variation was significant at the 0.01% level for each of the 7 interaction measures. These results are consistent with the finding, highlighted elsewhere in this report ( Chapter 6), that men who had had a mental health problem were more likely than their female counterparts to have experienced stigma on account of their problem. It may be that men suffering from mental ill-heath are felt to be more unpredictable or to pose more of a threat than women with the same symptoms. Alternatively, the findings may simply be a reflection of people's willingness to interact with strangers of the two sexes, irrespective of whether or not mental health is an issue.

Table 8.3: Willingness to interact with person in the scenarios, by gender of subject - mean responses

Source: Ipsos MORI

8.24 In terms of the symptoms depicted in the scenarios, willingness to interact with Robert/Shona was highest among those shown the stress scenarios, slightly lower among those shown the depression scenarios and lower still among those shown the schizophrenia scenarios, as table 8.4 (below) illustrates.

Table 8.4: Willingness to interact with person in the scenarios, by condition depicted - mean responses

Source: Ipsos MORI

8.25 The analysis also considered whether there was an interaction between the gender of the person in the scenarios and the symptoms depicted for each of the 7 measures - in other words, whether the observed differences in mean response between the male and female versions of the scenarios differed according to symptoms depicted and vice versa. Very little variation was found in this regard, that is, the differences between mean responses for male and female versions of the scenarios were constant, regardless of the symptoms depicted. The only exception was in the case of the "start working closely with Robert/Shona" measure. Here, the gender difference varied for each of the three conditions. Specifically, the difference was most pronounced among those shown the schizophrenia scenarios, slightly smaller among those shown the depression scenario and marginal among those shown the stress scenario. In other words respondents were: much more willing to work with a female showing symptoms of schizophrenia than with a male showing the sample symptoms; somewhat more willing to work with a female with depression than with a male with depression ²¹; and almost equally as willing to work with a female with stress as with a male with stress (see figure 8.1 below).

8.26 Reasons for these differences are unclear. One possible explanation is that, of all the measures of interaction listed, working closely with Robert/Shona involves the closest and most sustained level of contact with him/her. Further, any potential difficulties arising as a consequence of Robert's/Shona's symptoms could be seen to pose more of a problem a work setting than in less formal social setting. These considerations, combined with the hypothesis, set out above, that people are more wary of males with mental health problems than with females with the same problems, may account for the more cautious responses to the male versions of the schizophrenia and depression scenarios

Figure 8.1: Willingness to start working closely with Robert/Shona - interaction between gender and symptoms of person in scenario

Variation by socio-demographic characteristics of the respondent

8.27 This analysis considered to what extent willingness to interact with the person in the scenarios varied according to respondents' gender, age and income.

8.28 Gender was found to be an important discriminator on two of the interaction measures, namely, willingness to spend an evening socialising with Robert/Shona and willingness to make friends with him/her. In both cases, the mean response among women was higher than that among men (table 8.5). These results are consistent with the finding, reported in Chapter 7, that women were less likely than men to say that they would find it hard to talk to someone with a mental health problem.

Table 8.5: Willingness to interact with person in the scenarios, by gender of respondents - mean responses

Source: Ipsos MORI

8.29 For willingness to make friends with Robert/Shona there was also an interaction between gender and income, that is, the observed differences in mean response between male and female respondents on this measure varied according to their income. Among males, willingness to make friends with Robert/Shona decreased as income rose, in other words, men on a lower income expressed a higher level of willingness than those on a higher income. Among women a somewhat different pattern emerged: the lowest and highest income groups were more willing to make friends with Robert/Shona than those on middle-level incomes (see figure 8.2).

Figure 8.2: Willingness to make friends with person in the scenarios -interaction between gender and income of respondents

8.30 Respondents' age was found to be an important discriminator for two of the interaction measures - willingness to have Robert/Shona marry into the family and to provide childcare. Younger respondents, i.e. those aged up to 44 years, were more likely than those aged 45 and over to say they would be willing to have Robert/Shona marry into their family (table 8.6). Similarly, people aged 25 to 35 years expressed a higher level of willingness than those aged 75 and over to have Robert/Shona provide childcare. These findings are consistent with the age-based variation in attitudes towards people with mental health problems, reported in chapter 7, and specifically the finding that people aged 75 and over were among those most likely to agree that the public should be better protected from people with mental health problems and people with mental health problems are often dangerous.

Table 8.6: Willingness to interact with person in the scenarios, by age - mean responses

Source: MORI

8.31 There was no other variation by respondents' socio-demographic characteristics.