Public Attitudes, Lay Theories and Mental Health Literacy: The Understanding of Mental Health

There is a large body of research into public conceptions of mental illnesses and disorders going back over 50 years (Star, 1955). This chapter seeks to review the complex literature on this topic scattered over a wide range of disciplines including anthropology, psychology, psychiatry and sociology. The aim is to provide the researcher in social psychiatry and allied disciplines the opportunity to have a comprehensive and critical review. Over the years there are three slightly different but overlapping research traditions with regard to this topic: public attitudes, lay theories and mental health literacy. The first concerns studies of attitudes towards people with mental disorders (Nunnally, 1961), that is, beliefs about what people with mental illness are like and also, how they should be treated. These studies may be about specific mental disorders, such as schizophrenia (Siegler & Osmond, 1966) and depression (Rippere, 1977, 1979) or more generally about mental illnesses. These are nearly always large survey based studies typical of market research or attitudinal studies. These studies are important as they can offer an explanation for negative and stigmatising attitudes towards mental disorder (e.g. Nunnally, 1961; Link, Phelan, Bresnahan, Stueve, & Pescosolido, 1999), and why so few of those diagnosed, seek help (Lin, Goering, Offord, Campbell & Boyle, 1996; Andrews, Hall, Teesson & Henderson, 1999). Secondly, studies relating to lay theories of mental illness have been conducted primarily by Furnham and colleagues (i.e. Furnham & Lowick, 1984; Furnham & Manning, 1997; Furnham & Haraldsen, 1998), focusing specifically on the nature, causes, and treatments of disorders, such as heroin addiction (Furnham & Thompson, 1996), and schizophrenia (Furnham & Rees, 1988). These studies are concerned with the structure of beliefs about aetiology and cure and the relationship between them. They originate in social attribution theory in psychology and are concerned with the extent to which lay people endorse biological, psychological or sociological theories for the causes of various illnesses. They are also concerned with the extent to which there is a clear logical correlation between perceived cause and recommended cure for specific individual illnesses. The third approach is the term ‘mental health literacy’ introduced by Jorm and colleagues (Jorm, Korten, Jacomb, Christensen, Rodgers, et al., 1997b) to refer to public knowledge and more specifically recognition of mental disorders. This encompasses theories of mental


Introduction
There is a large body of research into public conceptions of mental illnesses and disorders going back over 50 years (Star, 1955).This chapter seeks to review the complex literature on this topic scattered over a wide range of disciplines including anthropology, psychology, psychiatry and sociology.The aim is to provide the researcher in social psychiatry and allied disciplines the opportunity to have a comprehensive and critical review.Over the years there are three slightly different but overlapping research traditions with regard to this topic: public attitudes, lay theories and mental health literacy.The first concerns studies of attitudes towards people with mental disorders (Nunnally, 1961), that is, beliefs about what people with mental illness are like and also, how they should be treated.These studies may be about specific mental disorders, such as schizophrenia (Siegler & Osmond, 1966) and depression (Rippere, 1977(Rippere, , 1979) ) or more generally about mental illnesses.These are nearly always large survey based studies typical of market research or attitudinal studies.These studies are important as they can offer an explanation for negative and stigmatising attitudes towards mental disorder (e.g.Nunnally, 1961;Link, Phelan, Bresnahan, Stueve, & Pescosolido, 1999), and why so few of those diagnosed, seek help (Lin, Goering, Offord, Campbell & Boyle, 1996;Andrews, Hall, Teesson & Henderson, 1999).Secondly, studies relating to lay theories of mental illness have been conducted primarily by Furnham and colleagues (i.e. Furnham & Lowick, 1984;Furnham & Manning, 1997;Furnham & Haraldsen, 1998), focusing specifically on the nature, causes, and treatments of disorders, such as heroin addiction (Furnham & Thompson, 1996), and schizophrenia (Furnham & Rees, 1988).These studies are concerned with the structure of beliefs about aetiology and cure and the relationship between them.They originate in social attribution theory in psychology and are concerned with the extent to which lay people endorse biological, psychological or sociological theories for the causes of various illnesses.They are also concerned with the extent to which there is a clear logical correlation between perceived cause and recommended cure for specific individual illnesses.The third approach is the term 'mental health literacy' introduced by Jorm and colleagues (Jorm, Korten, Jacomb, Christensen, Rodgers, et al., 1997b) to refer to public knowledge and more specifically recognition of mental disorders.This encompasses theories of mental 5 There are also differences between the content of lay theories and academic theories of mental disorders.The main finding is that lay people place more emphasis on psychological, social, and familial causal factors (Sarbin & Mancuso, 1972;Angermeyer & Matschinger, 1996;Furnham & Thompson, 1996), which can be compared to primarily biological and genetic academic theories.However, lay beliefs about causes vary depending on the disorder.For example Jorm and colleagues (1997a) found that schizophrenia was more likely to be attributed to genetic factors than depression, and autism was more likely to be biological than theories of obsessive-compulsive disorder, which were more likely to be psychodynamic (Furnham & Buck, 2003).It is therefore necessary to investigate the structure and content of lay theories of bipolar disorder, as previous studies show that it is difficult to generalise across disorders.Notably, the finding that lay theories are generally psychosocial, rather than biological, has been frequently replicated and can therefore be used to make predictions about lay theories of bipolar disorder.Lay theories about the treatment of mental disorders show marked differences from current practices in the mental health service, which involve drug treatment for mental disorders and/or psychotherapies such as Cognitive Behavioural Therapy (CBT).It has been found that lay people generally prefer psychotherapy to drug treatment (Angermeyer & Matschinger, 1996;Angermeyer & Dietrich, 2006) due to the perceived side effects (Angermeyer, Daumer, & Matschinger, 1993;Priest, Vize, Roberts, Roberts & Tylee, 1996;Fischer, Goerg, Zbinden, & Guimon, 1999).There is also a common lay belief that 'will power' can effectively facilitate recovery from mental disorders (Knapp & Delprato, 1980), such as agoraphobia and anorexia nervosa (Furnham & Henley, 1988).However, medication is believed to be the most effective treatment for disorders with a higher perceived severity (Furnham & Rees, 1988;Furnham & Bower, 1992), thus showing that lay and academic theories of treatment overlap to an extent.It is unclear how these findings may relate to bipolar disorder, especially since the perceived severity of the disorder is not known.However, it is predicted that psychotherapies will be preferred to drug treatment.Other studies have focused on assessing whether there is a logical relationship between lay theories of cause and treatment.For example, it is expected that if cause is attributed to biological factors, medication should be endorsed as treatment.This has been found in a number of studies which show a strong relationship between similar cause and treatment theories (Furnham & Buck, 2003), and those which are "sensibly" linked (Furnham & Haraldsen, 1998, pp. 696).However these findings are not always replicated.To demonstrate, medication was the preferred treatment for schizophrenia, despite participants attributing the cause to psychosocial factors (Furnham & Bower, 1992;Furnham & Rees, 1988).However, in this case it is predicted that there will be a coherent relationship between theories of causes and treatments of bipolar disorder due to the predominant findings of previous literature.Two general models of lay beliefs have been proposed.The 'medical' model (Rabkin, 1974), which suggests that mental disorders are like any other illness with symptoms caused by an underlying biological pathology and a treatment which addresses this.This has positive implications as it suggests that people with mental disorders should not be viewed differently than those with a physical illness.The second model is the 'psychosocial' model (Sarbin & Mancuso, 1972), which suggests that causes of mental disorder are psychological and environmental.This has positive implications for treatment as it advocates social and community support rather than hospitalisation.However, it has been found that people with beliefs which correspond to this model are less trusting of ex-psychiatric patients than ex-medical patients (Sarbin & Mancuso, 1972).These models can therefore be used to classify lay theories and have wide implications for attitudes towards those with mental illnesses, causal beliefs and treatment preferences.

Determinants of lay theories and mental health literacy
Attempts have been made to determine why particular theories are endorsed more than others.A number of studies have found that lay theories are predicted by demographic variables.Specifically, studies show that both younger and more educated people have more informed beliefs about mental disorders (Shurka, 1983;Hasin & Link, 1988;Yoder, Shute & Tryban, 1990;Fisher & Goldney, 2003).Significant effects of gender (Furnham & Manning, 1997), political persuasion (Furnham & Thompson, 1996), and religiousness (Furnham & Haraldsen, 1998) have also been found.This suggests that demographic variables may have some value in predicting many lay theories.In relation to familiarity with, and knowledge about, mental disorders, it has been found that participants with less knowledge of autism endorse external theories of cause, such as luck and religion, rather than academic theories (Furnham & Buck, 2003), whereas correct recognition of schizophrenia predicts more informed causal beliefs (Jorm et al., 1997a).A large increase in recognition of mental disorders has also been found for mental health professionals compared to the general public (Jorm, Korten, Rodgers, Pollitt, Christensen et al., 1997d).These studies suggests that informed beliefs about the nature, causes and treatments of mental illness come from diverse reading, academic study and/or extensive contact with people affected by mental disorders.Therefore, these variables should have some predictive value for both recognition and theories of mental illness.

The studies
Table 1 summarises the results of two dozen studies on over a dozen mental illnesses.The table shows, in essence, the method and results of the studies which will not be repeated here.All were completed in western developed countries and participants were generally better educated and younger than the population as a whole.This would suggest that they probably have more sophisticated lay theories

Disorder
Authors Study Alcoholism Furnham & Lowick (1984) 265 participants completed a questionnaire in which they rated 30 explanations for their importance in explaining the causes of alcoholism.
• Findings indicated a gender effect: females, more than males, believed alcoholics to be socially inadequate and anxious and held the belief that there is too much social pressure and not enough prohibitions against drinking.• Furthermore, there was an age effect: middle aged rather than younger or older tended to explain alcoholism in terms of poor education, social and cultural pressures and biological and genetic mechanisms.

Disorder
Authors Study Factor analysis revealed six factors: psychological stress, personal and social problems, psychoanalytic theories, sociocultural explanations, biological or genetic explanations and social desirability or pressure.Anorexia Nervosa Furnham & Hume-Wright (1992) 168 participants completed a 105-item questionnaire which explored their beliefs regarding the cause, correlates and cures of anorexia nervosa.
• Findings suggested lay people hold elaborate, consensual and moderately accurate (parallel to clinical theories) beliefs about the description, cause and cures of anorexia nervosa.

•
Sex, personal experience of eating disorders and being acquainted with an anorexic were significant correlates of a number of factors.

•
Factor analysis identified clusters of responses that showed underlying factors of family, stress of change, conflict in contradictory social roles, goals and demands, rebellion and security.Furnham & Manning (1997) 147 participants completed a 108-item questionnaire, based on Furnham & Hume Wright (1992).The four parts of the questionnaire were individually factor analysed and an interpretable factor structure emerged for each.
• Results indicate young people (16-19 year olds) hold moderately accurate beliefs about the causes and cures of both anorexia nervosa and obesity.

•
Participants seemed to see social pressure affecting the development of anorexia as most important; and selfworth as most important for cures.

•
Sex, actual body size, estimated body size and having experience with an eating disorder were found to correlate significantly with a number of factors.

•
Factors of cause and cure were not correlated regarding anorexia but were for obesity.Benveniste, Lecouteur & Hepworth (1999) Lay theories of anorexia nervosa using critical psychology perspective (Discourse Analysis) were investigated through 10 semi-structured interviews with 5 women and 5 men aged 15-25.

•
It is concluded that lay theories of anorexia nervosa were structured through these key discourses which maintain separation between socio-cultural and individual psychology in relation to anorexia nervosa.

•
This reinforces the concept that anorexia is a form of psychopathology.

Disorder
Authors Study Autism and obsessive compulsive disorder Furnham & Buck (2003) A total of 92 participants were involved in the two studies.
In study 1 parental interviews of sufferers were conducted and revealed that, as hypothesised, parents hold predominantly biomedical views regarding autism.
• Participants then completed the questionnaire with varied levels of experience of autism (no experiencerelatives of sufferers) which involved rating a range of theories of aetiology and treatment approaches for each disorder.

•
Statistical analysis confirmed that lay beliefs about autism were primarily biomedical and beliefs about OCD were primarily psychological.

•
Multiple regression analyses indicated that a range of individual difference factors (i.e.religiousness and age) predicted beliefs about the importance of the factors derived from factor analysis of the belief statements.

Depression
Furnham & Kuyken (1991) After a pilot study asking people to list the causes of depression, 201 participants completed a questionnaire which involved rating 32 explanations and 5 current theories of depression.The study reports university students' attributions for the causes of and cures for depression in Turkey.

•
Seven components were found for cures: hobby, sensation seeking, avoidance, professional help, religious practices, esteem and spiritual activities.

•
Men rated religious practices as more useful than women did.

•
No other gender differences were found.Budd, James & Hughes (2008) The study aimed to develop a robust factor structure of lay theories of depression, while more adequately sampling from the full range of hypothesised causes of depression.

•
The reasons rated most important for depression were related to recent bereavement, imbalance in brain chemistry and suffering sexual abuse or assault.

•
The data was best described by a 2-factor solution, the first representing stress and the second depressogenic beliefs.

•
Results indicated participants were unclear of the causes and cures of GID but these beliefs were logically related.Heroin Addiction Furnham & Thompson (1996) 144 participants completed a questionnaire examining the structure and determinants of lay people's implicit theories of heroin addiction.They had to rate 105 statements about the causes, correlates and cures of heroin addiction.
• Factors seemed similar to explicit academic theories; except beliefs about cure which did not show support for most clinical models.

•
When a higher order factor analysis was performed, four factors emerged: moralistic, psychosocial, sociocultural and drug treatment which reflect coherent views on the nature of heroin addiction.

•
The strongest demographic determinant of lay beliefs in these factors was political beliefs.Right wing voters emphasised moralistic and individualistic theory and left-wing voters supporting the psychological and societal ideas.

Neurosis
Furnham (1984) Three experiments aimed to investigate various determinants of the common-sense conception of neuroticism.In the first experiment subjects completed various standardized psychological tests measuring neuroticism and anxiety, while also estimating the extent of their own anxiety and neuroticism.In the second experiment subjects attempted to detect items measuring neuroticism in a standard personality questionnaire and secondly estimate the extent of their own and the 'average' person's neuroticism.Finally, in the third experiment subjects rated the typicality of various neurotic traits and behaviour which had been supplied by subjects in the previous two experiments.
• Findings demonstrated some similarities in expert explicit theories and lay-person implicit theories, though there appeared systematic biases in subject's perception of their own neuroticism.

Disorder Authors Study Paraphilia
Furnham & Haraldsen (1998) The paper examined four types of Paraphilia: fetishism, paedophilia, sexual sadism, and voyeurism.105 participants completed a four part questionnaire divided into: demographic details, perceptions of etiology, ratings of cure for each Paraphilia and the Eysenck Personality Questionnaire.
• Factor analysis revealed a clear and logical factor structure for etiology and cure items.

•
Further, etiology and cure items correlated strongly with each other but only moderately with demographic and personality differences.Phobia Furnham (1995) 150 people completed a two-part questionnaire that investigates beliefs about the nature and cure of phobia.Five factors emerged from the 23-item attitude section: • The ideas that: certain personality factors related to phobia, there are physical correlates of phobia, observational learning causes phobia, phobias are caused by behavioural pairing, and Freudian ideas of unconscious association.The 13-item treatment section showed four factors:

•
There was a clear and logical relationship between perceptions of the causes and treatment of phobia demonstrating that lay people have coherent theories of the etiology and cure of phobia.Schizophrenia Furnham & Rees (1988) Subjects completed two brief questionnaires, one concerning the description of, and attitudes towards schizophrenia and schizophrenics and the second on the possible cause of schizophrenia.Beliefs about the conceptions of mental illness suggested four factors labelled dangerous, amoral, egocentric and vagrant.The items on the causes factored into five factors labelled stress and pressure, biological, genetic, backward and brain damage.Furnham & Bower (1992) 106 Lay respondents (students, nurses, employed and unemployed) aged 18-60 answered a questionnaire examining five identified main academic theories of schizophrenia (medical, moral-behavioural, social, psychoanalytic and conspirational) along various dimensions (e.g.aetiology, behaviour, treatment).

•
No single model was favoured exclusively but seemed to point to a synthesis of several academic theories.Lay subjects stressed the importance of patient environment in the aetiology of schizophrenia rather than a physiological malfunction, but tended to stress the personal rights of the schizophrenic.

•
Results indicated that for schizophrenia, social environmental factors (day-to-day problems, traumatic events) were often seen as causes which are a contrast to the weak epidemiological evidence for such a role.

•
Genetic factors attracted more attention as a cause of Schizophrenia than Depression.Of notable concern was the popular belief (over half) of respondents that weakness of character was a likely cause of both depression and schizophrenia, implying a negative evaluation of the sufferer as a person.Furnham & Wong (2007) The study investigated 200 (101 female, 99 male) British (100) and Chinese (100) participants' beliefs about the causes, behaviour, manifestations and treatments of schizophrenia.Results confirmed the three hypotheses that: 1. Chinese would hold more superstitious and religious beliefs towards the causation and treatment of schizophrenia and would prefer the use of alternative medicine.2. The British emphasised more on internal (biological and psychological) and external (sociological) beliefs for the causes and treatments.3. Chinese participants held more negative attitudes and beliefs about the behaviour manifestations of schizophrenia than the British.Furnham, Raja & Khan (2008) A total of 305 British, British Pakistani and Native Pakistani medical students completed a questionnaire on general beliefs about people with schizophrenia, causal explanations concerning aetiology and the role of hospitals and society in treating people with schizophrenia.
• There was strong evidence to suggest Pakistanis possessed more negative beliefs and attitudes about people with schizophrenia, but no evidence to suggest Pakistanis believed more in superstitious causal explanations.

Disorder Authors Study •
Pakistanis were more likely to consider seeking help from faith healers, but not God, compared with the British Pakistani and British participants.

•
Results confirm cultural (European-Asian) difference in the understanding of the cause, manifestation and cure of schizophrenia.Suicide Knight, Furnham & Lester (2000) Attitudes toward suicide were explored in 150 young people.

•
The strongest correlate of these attitudes was psychoticism scores, with the respondents with higher psychoticism scores viewing suicide more positively than those with lower scores.Walker, Lester & Joe (2006) African Americans' lay beliefs and attributions towards suicide were examined in 251 undergraduate college students using the Attitudes Towards Suicide Scale, Life Ownership Orientation Questionnaire, Stigma Questionnaire and Suicide Ideation Questionnaire.

•
Beliefs about stigma were comparable across ethnic groups.

•
African American students were significantly less likely than European American students to attribute suicide to interpersonal problems and report the individual or government as responsible for life.

•
African American students were significantly more likely to report that God is responsible for life.
Table 1.Previous studies of lay theories of mental disorder Approximately 200 lay people (working adults and students) and over 50 practicing psychologists completed a four-part questionnaire that examined attitude to psychotherapy, beliefs concerning what patients report during psychotherapy, the efficacy of quite different types of psychological treatment, and finally the prognosis for a wide range of psychological problems.

•
The results revealed numerous and consistent differences which showed that, compared to lay people, psychotherapists seemed more skeptical and pessimistic about the efficacy of therapy and the prognosis for various psychological illnesses.

•
Psychotherapists believed that clients in psychotherapy tended to report more positive, favourable reactions than lay adults and students.

•
Psychotherapists tended not to believe that therapists teach specific skills but rather that they provide some sort of social support and help vent fears and other negative emotions.

•
Therapists seem more skeptical in beliefs about the efficacy of different therapies and the prognosis of different problems.Therapists believe that different therapies are suitable for particular problems (and that some therapies are by-and-large fairly useless), while lay people believe therapies are suitable for a wide range of Author(s) "Title" Study psychological problems.

•
Lay people believe the prognosis for a wide range of psychological problems to be better than the therapists.

Wong (1994) "Lay Theories of Psychotherapy and Perceptions of Therapists: A Replication and Extension of Furnham and Wardley"
Two hundred and forty and 43 nonfaculty staff members at the University of Northern Iowa participated in the study.Each participant completed four questionnaires and read a vignette of a part of a psychotherapy session in order to examine the laypersons' perceptions of psychotherapy, the experience of psychotherapy clients, and therapist credibility.

•
The participants appeared to have realistic conceptions about what occurs in therapy and to be quite optimistic about treatment outcomes.
Participants tended to disagree with most of the popular stereotypes about psychotherapy (e.g., most clients lie on a couch; women make better therapists than men) and to agree with the goals and techniques of most types of modern psychotherapy (e.g., therapists teach strategies to reduce conflict or frustration; psychotherapists encourage the expression of emotions).They also indicated that the therapeutic experiences and relationship lead to improvements for a variety of problems and client types.

•
They did endorse some common misconceptions about psychotherapy (e.g., most therapists ask about dreams, believe psychological problems start in childhood, or use personality questionnaires).They also responded neutrally to statements about some important aspects of therapy, such as the client-therapist relationship and the length of therapy.

•
Males, older individuals, and those with more psychological experience were less optimistic, but perhaps more realistic, about the potential benefits of therapy.However, the more experienced participants, surprisingly, did not differ from the less experienced ones in their beliefs about the experience of clients.As expected, when age, experience, and sex were controlled, the student and staff did not differ in their psychological experience or in their beliefs about psychotherapy.

Author(s) "Title" Study
• Neither participant's sex nor the amount of fee charge affected participants' perceptions.Also, treatment modality (client-centered vs. behaviour vs. rational-emotive) significantly affected ratings of overall, and each dimension of, therapist credibility.

•
Only a few beliefs about psychotherapy influenced perceptions of (only limited aspects of) therapist credibility.Two hundred and seventeen participants completed a two-part questionnaire in order to study the structure and determinants of lay beliefs about psychotherapy in general and specifically the effectiveness of various therapies for four different disorders.

•
It was clear from the factor analysis of the ratings (across all four conditions) that participants did not distinguish between a range of talk therapies including psychoanalysis, gestalt and existentialist therapies, on the one hand, and group/marital therapies on the other, as well as more social-behavioural therapies like CBT, assertiveness and thought-stopping.To the participants they all appeared to involve talk which aims to change cognitions and emotions.

•
Participants clearly differentiated between the efficacies of the different therapies.Whether considered to be moderately efficacious (cognitive/talk therapies) or not, participants saw some as being significantly more appropriate than others.
• Cognition therapy was seen as efficacious for depression and delusional disorders.

•
On the other hand, physical therapies were perceived moderately useful with anorexia but not at all for depression.

Limitations
The studies shown in Table 1 and  1 that only a small number of disorders have been investigated.They are however not arbitrary as the topics covered are often the more common mental illnesses (i.e.depression), as well as those most discussed by the public.

Conclusion
It is plausible to draw five conclusions from the scattered literature on lay theories of mental illness.First, as many have noted, knowledge of the various mental illnesses is both patchy and highly varied.Whereas lay people seem relatively well informed about some, like depression, they are surprising ignorant about others, like schizophrenia.This seems to be related to the prevalence of the illness in the population as well as media coverage of the problem, particularly celebrities admitting to being "sufferers".Many hold antiquated and unsubstantiated views on the nature or manifestation of mental illnesses that educators have been trying to correct for years.
Second, when thinking about the cause of mental illness it is common to find five types of explanations: psychological; sociological; biological/genetics; psycho-analytic and moral/behavioural.Most people rate typical psychological explanations seeing the cause to be individual mal-functioning of some system.They are also happy to acknowledge group, societal and structural variables as contributing to illness.To a limited extent, and highly specifically with respect to particular illness, people rate biological (i.e.hormonal, brain damage) and genetic factors as a major causal role in the mental illness aetiology.There is also a surprising number of people who endorse classic Freudian explanations of dysfunctional early relationships with parents or others.For some mental illnesses lay people are happy to suggest the cause is "wickedness" of one sort or another: that is, that the cause is gross moral turpitude.For some people in third world countries the cause is seen to be spiritual: that is the intervention and possible punishment from a "higher force" or "pantheon of gods".The extent to which people appear to endorse one type of theoretical explanation over another is a function of the illness in question, as well as their education and ideological orientation.Third, lay people seem less certain about cure/intervention/management than cause.Once again, various types of cures are seen to be, at least in part, relevant to many problems.For most problems, the favoured cure, in terms of perceived efficacy, is a variety of the "talking cures": that is, psychotherapy of one sort of another.Cures that are acknowledged, but rated as less appropriate and effective, are behavioural, pharmacological(drugs) or surgical.
For some illnesses incarceration or some change in the way society operates is seen as effective.
Fourth, there is often a weak and not always coherent, relationship between perceived cause and the cure of a mental illness.Usually there is a weak positive correlation between psychological causes and cures, though it is recognized that although the cause may be psychological, the effective cure may be drug related.Certainly it does seem that people do not hold very coherent theories for the origin, progression, manifestation and alleviation of mental illnesses.
Fifth, studies that have attempted to identify the demographic and experiential correlates of mental health knowledge or literacy have shown some predictably and explanatory results.Thus, those that are younger, better educated people know more; those with training on psychology/psychiatry know more; and those with personal experience (self/relatives) are better informed.However, the significant effects are very weak.

Acknowledgement
We would like to acknowledge the help of Steven Richards for his work on this project.
Canberra: Commonwealth Department of Health and Aged Care.

Table 2
shows the results of various studies concerned with lay theories of the process of psychotherapy.They are concerned with the perceptions of what occurs in a (typical) therapy session, the efficacy of different cures, the prognosis for different problems, and the differing perceptions of lay people and clinicians.

Table 2 .
Studies on lay theories of Psychotherapy Table 2 suffer various limitations.The first is sampling.Many studies have sampled relatively small, better educated, Western participants, often with disproportionate numbers of students.Far fewer have looked at the ideas of patients, their relatives or psychiatric staff.Further, very few studies have been conducted in third world or developing countries where it is known that people hold very different lay theories.Ideal studies would use large, representative samples.The second is method.Most studies have been questionnaire based after initial testing with interviews.All methods have limitations, and it would be most desirable to use multiple methods to investigate lay theories.It would also be most beneficial to trace theories over time, particularly if some attempts were made to change them via an intervention like mental health education.The third is topic.It is clear from Table