Effects of a psychiatric diagnosis vs a clinical formulation on lay attitudes to people with psychosis

ABSTRACT

Background: Limited research has investigated whether replacing psychiatric diagnosis with psychological formulation-based approaches has implications for lay attitudes to mental health. The present study investigates experimentally whether presenting psychosis in terms of a schizophrenia diagnosis vs. formulation narrative affects stigma and treatment attitudes in the general public.

Method: The study employed a between-groups experimental vignette design, with data collected online. 351 participants (64.1% female, aged 18–66,) read a vignette about a person experiencing psychosis, defined with either a diagnosis of schizophrenia or a narrative-based formulation. Participants completed a battery of scales measuring their attitudes to the vignette character (social distance, attribution, recommended treatment options, mental help-seeking attitudes).

Results: Desired social distance was significantly greater in participants exposed to the diagnostic label of schizophrenia. The schizophrenia label led participants to rate medical care as significantly more helpful relative to the formulation condition but did not affect ratings of specialist or community care or mental help-seeking attitudes.

Conclusions: These findings suggest that a psychological formulation approach may slightly lessen stigma-related attitudes, relative to traditional diagnostic systems. Popularisation of formulation models need not compromise general orientations to help-seeking or perceived helpfulness of specialist care but may lead to less medicalised treatment preferences.

KEYWORDS:

• Psychosis
• stigma
• diagnosis
• case formulations
• labelling

Introduction

Recent debates about the classification of mental illness have proposed replacing traditional categorical diagnostic systems with a psychological formulation model. While competing approaches are typically appraised in terms of scientific validity or clinical utility, it is also important to consider their implications for lay attitudes to mental illness. The current study experimentally investigates the impact of framing a case of psychosis using psychiatric diagnosis vs. psychological formulation on stigma and treatment attitudes.

Categorical diagnostic systems

The fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) (American Psychiatric Association, 2013) is the standard classification of mental disorders in the USA and many other countries (Blashfield et al., 2014). The purpose of the DSM-5 and other categorical classification systems is to aid clinicians in diagnosing the disorders patients are suffering with using diagnostic criteria that summarise the characteristic symptoms displayed by the patients (American Psychiatric Association, 2013). The DSM has always courted controversy, for example, in its definition of homosexuality as a mental disorder until 1973 (Silverstein, 2009) and the limited evidence-base supporting some disorders such as Dissociative Identity Disorder (Paris, 2012). The launch of the fifth edition of the DSM (DSM-5) in 2013 sparked strong criticism from prominent voices in American psychiatry, who denounced DSM-5 for sacrificing validity for reliability (Insel, 2013) and for pathologizing typical human experiences (Frances, 2013). Alongside concerns about overdiagnosis (Bandini, 2015), the DSM has also been criticised for high comorbidity between diagnostic categories (Wang et al., 2018) and poor reliability between clinicians (Chmielewski et al., 2015).

An additional limitation of the DSM’s approach to mental health is that by classifying individuals using diagnostic labels, DSM categories may inadvertently reinforce social stigma. Theories of the stigma process position mental illness labels as key triggers of discriminatory social responses (Link et al., 1989; Scheff, 1974). The diagnostic label can be a cue for stereotypes, with the associated diagnostic criteria reinforcing prejudice by validating negative beliefs about people who hold that diagnosis (Corrigan, 2007). A recent systematic review of experimental vignette research exploring the impact of diagnostic labels on social attitudes confirms the existence of such labelling effects but identifies significant variations between diagnostic categories (O’Connor et al., under review). For instance, while revealing that a symptomatic person holds a diagnosis ameliorates stigma in the case of autism spectrum disorder (O’Connor et al., 2019), disclosure of diagnostic labels exacerbates stigmatising responses to people with other conditions, such as attention-deficit/hyperactivity disorder (Thompson & Lefler, 2016) and depression (Abdullah & Brown, 2020). Very limited research has explored whether stigma can be reduced by presenting mental health difficulties using frameworks that do not rely on categorical diagnosis.

Psychological formulation

The critiques and limitations associated with the DSM-5’s approach to understanding mental health (Kinghorn, 2020) have prompted calls for a paradigm shift away from a medical model of mental health (Elkins, 2017). One alternative to psychiatric diagnosis is psychological formulation, the ongoing process of collaboratively constructing a narrative of the reasons behind a person’s difficulties, considered within the context of their relationships, social circumstances and life events (Johnstone, 2018). The Power Threat Meaning (PTM) framework provides one structure for psychological formulations (Johnstone et al., 2018). To collaboratively formulate a narrative with clients, the PTM framework recommends that clinicians consider the operation of power, threat and the meanings the individual derives of their experiences. Power, or what has happened to the person, could represent coercive power, such as bullying, but also reflect the individual’s social situations, like growing up in poverty. Threat, or how the operation of power affected the client, pertains to the effects of disempowerment and the responses it evokes, which manifest in presenting symptoms, such as hearing voices or self-isolation. Meaning, or the sense the person makes of the relationship between the power and threat responses, captures the individual’s understanding or emotional reaction, such as feeling isolated or unworthy (Johnstone et al., 2018).

The proposals and ideals of the PTM framework have been endorsed by many practitioners uncomfortable with medicalised approaches to mental health. Pilgrim (2018), one of the framework’s co-authors, argues that the PTM framework can facilitate a shift towards trauma-informed services, allowing a kinder approach to mental health intervention. Strong (2019) proposes that the framework is well-researched, offers an opportunity to identify meaningful complexities unaddressed by psychiatric diagnoses and is a viable alternative to the DSM-5. However, others have criticised the PTM framework as driven by ideology and difficult to read (Salkovskis, 2018) and unclear in its distinction between medical and psychological challenges (Larkin, 2018). Additionally, the framework’s prioritisation of social causes of human suffering has led to accusations that it ignores evidence of biological influences on mental health (Phillips & Raskin, 2020). Thus far, these debates have been largely based on anecdotal experience and theoretical commitments; little research has empirically evaluated the outcomes of adopting a PTM approach.

Beyond questions of scientific reliability and clinical utility, it is also important to consider the societal implications of various approaches to conceptualising mental illness. As the PTM framework provides an alternative to diagnostic systems (Johnstone et al., 2018), it has been proposed as a way of mitigating the social stigma associated with diagnostic labels (Aherne et al., 2019) by reframing public discussion of mental health (Harper, 2020). However, this has not been directly tested. Previous research has shown the benefits of narratives, an alternative to diagnoses, in reducing self-stigma and improving self-esteem in persons with mental illness (Hansson et al., 2017; Kosyluk et al., 2021), including those with schizophrenia-spectrum disorders (Yanos et al., 2019). As yet, no research has established how the mental health attitudes of the general public (i.e. non-clinical samples) are affected by presenting mental health difficulties using narrative-based approaches.

While some have argued that abandoning categorical labels will promote more positive mental health attitudes, it may also involve some risk. Previous research indicates that diagnostic labels hold some benefits for people who receive them, such as facilitating self-understanding (O’Connor et al., 2018), social validation (O’Connor & McNicholas, 2020) and empowerment (Perkins et al., 2018). Moreover, diagnoses may promote therapeutic engagement by circumventing a key barrier to help-seeking: laypeople’s capacity to distinguish “typical” discontent from clinically significant levels of distress that would benefit from professional intervention (Biddle et al., 2007). For instance, Wright et al. (2012) found that a label of schizophrenia predicted higher perceived helpfulness of specialist treatment options (such as psychiatrists and psychologists) in adolescents. Similarly, Picco et al. (2018) demonstrated that adults who correctly diagnosed cases of depression and schizophrenia were more likely to recommend professional sources of help. Coles and Coleman (2010) found that attributions of symptoms to environmental factors were often related to recommendations against seeking professional help. These findings suggest potential risks of popularising the PTM framework: success in shifting mental health discourse to bypass diagnostic labels, emphasise social influences, and “normalise” mental health difficulties as typical human responses could inhibit potential service-users from seeking specialist support.

Thus, replacing categorical diagnosis with formulation systems may plausibly yield benefits in ameliorating social stigma, yet may also present risks of inhibiting help-seeking and treatment engagement. Without empirical data that compares the competing approaches’ effects on lay mental health attitudes, decisions on which approach to favour remain ill-informed. The current study aims to provides the first such evidence using the test-case of psychosis.

Attitudes to psychosis

Attitudes to psychosis serve a useful platform for exploring the social effects of diagnostic labelling vs. formulations. Schizophrenia is one of the mental illnesses most stigmatised by both mental health professionals (Mittal et al., 2014; Valery & Prouteau, 2020) and the general population (Angermeyer & Matschinger, 2003; Matsunaga & Kitamura, 2016). People with psychosis often face fear, social distance and perceived dangerousness (Angermeyer & Matschinger, 2003). Longitudinal research indicates exposure to stigma exacerbates psychotic symptoms (Rüsch et al., 2015). Evidence regarding the degree this stigma specifically results from the diagnostic label, rather than the symptoms of psychosis, is mixed: while some studies find the schizophrenia label increases stigma (Cheung et al., 2018; Matsunaga & Kitamura, 2016; Mittal et al., 2014), others indicate that it does not affect or decreases stigma (Abdullah & Brown, 2020; Parrish et al., 2019). A belief in the uniquely stigmatising effects of the schizophrenia label is reflected in international proposals to rename schizophrenia (Lasalvia et al., 2015), for example, to “integration disorder” in Japan (Sato, 2006). Research that has tested the effects of such changes finds that alternative labels reduce stigma, but still incite more negative responses than unlabelled symptomatology (Ellison et al., 2015; Yang et al., 2012). Similarly, labels relating to psychosis, such as “paranoid”, are more likely to induce negative attributions in the general public than non-psychiatric labels, like “weird” (Anglin et al., 2014). Destigmatisation may require a more wholesale paradigm shift away from medicalised labels. The PTM framework offers one such alternative approach to conceptualising psychosis by implementing a psychological formulation that uses the person’s own description of their experience of hearing voices, without imposing a particular interpretation (Cooke & Kinderman, 2018). As such, the current study will investigate the following hypothesis:

Hypothesis 1: that describing a case of psychosis using a PTM narrative reduces stigmatising attitudes, relative to using a DSM-5 diagnosis of schizophrenia.

Beyond stigma, the study also investigates implications for lay attitudes to treatment. As early identification of psychosis is associated with improved clinical outcomes (Marshall & Rathbone, 2011; McGorry, 2015), anything that delays or dissuades help-seeking may lead to increased levels of distress. Previous research suggests the schizophrenia label is associated with greater endorsement of professional sources of help and psychological therapies (Wright et al., 2012). Therefore, this study also proposes the following two hypotheses:

Hypothesis 2: that describing a case of psychosis using a PTM narrative is associated with less positive attitudes to help-seeking, relative to using a DSM-5 diagnosis of schizophrenia.

Hypothesis 3: that describing a case of psychosis using a PTM narrative diminishes the perceived helpfulness of treatment options, relative to using a DSM-5 diagnosis of schizophrenia.

The study tests the above three hypotheses using an experimental vignette method. While experimental vignettes have limited ecological validity, they offer a high degree of experimental control and provide a parsimonious initial strategy to explore the differential social implications of diagnostic vs. formulation frameworks. Results will increase understanding of the potential benefits and risks of non-diagnostic formulation approaches at a societal level.

Methodology

Participants

The final sample consisted of 351 participants from the general public. Participants were 225 females, 110 males, nine non-binary people and four who did not disclose gender, with an age range of 18–66 (M = 29.16, SD = 9.69). 132 participants had previously studied psychology and 127 were students (participants were not asked for their university course or education level). Of the participants who were not students, 22% worked in healthcare, 12% worked in retail and food service, 12% worked in IT and engineering, 10% were unemployed, 7% worked in education, 6% in finance and 4% in physical labour. The remaining 23% worked in eclectic roles, varying from poets to attorneys and retirees. Participants mostly resided in Ireland (60.39%), while 18.23% of participants were residents of European countries outside of Ireland (including the United Kingdom) and 18.51% of participants resided outside of the Europe.

Materials

The online study was hosted on Qualtrics (Qualtrics [Computer software], 2019). Vignettes describing “Charlie” – a gender-neutral, fictional individual – were developed (see online supplementary material for vignettes). Both conditions received the same brief context for Charlie and then either presented a description of the DSM-5 with Charlie’s diagnosis or a formulation following the PTM framework’s guidelines. The vignettes were similar in terms of structure and length. An attention check was included that verified memory for details in the vignette. Scales were adapted to “Charlie” and any gendered pronouns were removed.

Social Distance Scale (SDS; Link et al., 1987)

The SDS was used to indicate desired social distance. Seven items asked participants to indicate how they felt about associating with Charlie on a seven-point Likert scale. Higher scores indicated less desired social distance. Cronbach’s alpha indicated a high internal consistency of .9.

Attribution Questionnaire (AQ-9; Corrigan, 2008)

The AQ-9 was used for exploring participants’ negative responses to Charlie. On a nine-point Likert scale, participants rated their responses to Charlie on nine items, with higher scores indicating more negative responses. A low internal consistency was found, with a Cronbach’s alpha of .54. Removal of one item improving reliability to an acceptable level (α = .66).

Mental Help Seeking Attitudes Scale (MHSAS; Hammer et al., 2018)

Participants were presented with nine polar items (eg. useful or useless) and asked to consider evaluate the impact of Charlie seeking further help from a mental health professional on a seven-point scale. Higher scores indicate more positive attitudes towards help-seeking. A Test of Reliability found a high internal consistency of .92.

Attitudes Towards Treatment Options Scale (ATTOS; Nolan & O’Connor, 2019, adapted from; Phelan et al., 2006)

Participants rated nine treatment options (e.g. visiting a psychiatrist) for Charlie from a range of “harmful” to “helpful”. Higher scores indicated higher perceived helpfulness of the treatment. A principal axis factor analysis indicated items could be grouped into three factors: “Medical Care” (e.g. going to a hospital, visiting a GP or a family doctor), “Specialist Care” (e.g. visiting a psychiatrist, going to see a psychologist) and “Community Care” (e.g. talking to a family member, joining a social or sports club to meet new people). Specialist Care and Community Care had high internal consistency, with Cronbach’s alpha values of .72 and .74 respectively. However, Medical Care had a Cronbach’s alpha of only .52. As the Medical Care subscale included only three items, none were removed. Therefore, test results using the Medical Care subscale should be interpreted cautiously.

Demographic questions

Finally, participants were presented with a set of demographic questions in order to characterise the sample and to systematically identify any covariates. Participants were asked what gender they identify as, their age, occupation and country of residence. They were also asked if they had ever studied psychology (yes or no).

Procedure

The study was advertised as investigating the influence of mental health information on the general public’s attitudes. Convenience sampling through social media (Facebook, Twitter and Reddit) was used to recruit participants. Advertisements included an anonymous link to the survey, where an information sheet was displayed. Once informed consent was received, the survey programme randomly assigned participants to read either the DSM-5 or the PTM framework vignette. Participants then completed the attention check. Participants then continued to complete the Social Distance Scale, the Attribution Questionnaire, Attitudes Towards Treatment Options Scale and the Mental Help-Seeking Attitudes Scale. Participants were asked a range of demographic questions. On completion, a debrief sheet was shown that explained the study. Ethical approval was obtained from the School of Psychology Undergraduate Research Committee in the university.

Results

Preliminary analysis

Analysis was completed using SPSS. Fully incomplete responses (n = 55) and participants who failed attention checks (n = 19) were not included in the final sample. A listwise approach to missing data in SPSS was used. Therefore, if a participant had not completed a single scale, they were excluded from analyses involving that particular variable, but included in analyses for which they had complete data.

A principal axis factor analysis was conducted for the ATTOS. Three factors were identified and supported by the scree plot. The eigenvalues of each factor exceeded Kaiser’s (1958) criterion of 1 (see Table 1 for the factors identified; see online supplementary material for eigenvalues).

Table 1. Treatment factors identified by principal axis factor analysis

Factor	Title	Eigenvalues	Items
1	Community Care	3.12	Approaching a community support organisation Talking to a family member Joining a social or sports club to meet new people
2	Specialist Care	1.57	Visiting a psychiatrist Going to see a psychologist Going to see a therapist or counsellor
3	Medical Care	1.18	Going to a hospital Visiting a chemist or pharmacist Visiting a GP or a family doctor

Descriptive statistics including the mean and standard deviation for each measure across conditions are included in Table 2.

Table 2. Descriptive statistics for the SDS, MHSAS, AQ-9 and ATTOS subscales

	Diagnosis Condition		Formulation Condition
Measure	M	SD	M	SD
SDS	32.71	8.98	34.57	8.65
MHSAS	54.00	10.15	52.67	11.23
AQ-9	27.91	6.07	26.36	6.57
ATTOS: Community Care	16.68	3.23	16.66	3.57
ATTOS: Specialist Care	19.01	2.50	18.75	2.80
ATTOS: Medical Care	13.86	3.74	13.11	3.21

M signifies Mean and SD = Standard Deviation.

Sociodemographic profiles were closely balanced across experimental conditions. To identify potential covariates for the main experimental analyses, preliminary t-tests explored the associations between demographic and dependent variables. Due to the low number of participants who identified as non-binary (n = 9) or withheld gender information (n = 4), gender was transformed to a binary variable for analysis. Age was dichotomised into people above or below the mean (29.16). Results showed that the SDS was significantly associated with gender (t(328) = −3.47, p = .00), age (t(338) = 4.41, p = .00) and experience studying psychology (t(343) = 4.30, p = .00), the AQ-9 with experience studying psychology (t(344) = −2.49, p = .007), Medical Care and Community Care with gender (t(329) = −2.36, p = .019 and t(329) = −2.97, p = .003, respectively) and Specialist Care with gender (t(331) = −3.91, p < .001) and age (t(340) = 2.30, p = .022). The MHSAS was not significantly associated with any demographic variable.

Stigmatising attitudes

One-way ANCOVAs were used to test the effects of condition on both the SDS and AQ- 9 to test the first hypothesis. Covariates were drawn from preliminarily analyses indicating potential confounding variables.

For AQ-9 scores, no significant differences were found between the DSM-5 condition (M = 27.91, SD = 6.07) and the PTM framework condition (M = 27.36, SD = 6.37) when experience studying psychology was controlled (F(1, 345) = .56, p = .45).

Controlling for gender, age and experience studying psychology, a significant difference was observed in SDS scores (F(1, 325) = 4.53, p = .04). Participants in the DSM-5 condition showed a stronger desire for social distance (M = 32.71, SD = 8.98) than those in the PTM framework condition (M = 34.57, SD = 8.65).

Mental help-seeking attitudes

When testing the second hypothesis, no significant difference was found in mental help-seeking scores between the groups (F(1, 332) = 1.13, p = .26). The mean score for the DSM-5 condition was 53.99 (SD = 10.14) and the mean score for the PTM framework was 52.67 (SD = 10.73), indicating that participants had high positive mental help-seeking attitudes in both conditions.

Perceived helpfulness of treatment options

For the third hypothesis, three one-way ANCOVAs were used to analyse the three subscales, controlling for gender (for all three tests) and age (for Specialist Care). Controlling for these covariates, no significant difference in perceived helpfulness was found between conditions for Community Care (F(1, 330) = .30, p = .58) or Specialist Care (F(1, 328) = 1.28, p = .26). Medical Care was, on average, rated the least helpful treatment option (M = 13.49, SD = 3.50). There was a significant difference in perceived helpfulness between conditions for Medical Care (F(1, 330) = 5.885, p = .02). Investigation of means indicated that participants in the DSM-5 condition perceived Medical Care as slightly more helpful (M = 13.86, SD = 3.74) than those in the PTM framework condition (M = 13.11, SD = 3.21).

Overall, Hypothesis 1 and 2 were supported. The DSM-5 condition (i.e. diagnostic label of schizophrenia) was associated with significantly higher social distance scores, indicating increased stigma. Participants in the DSM-5 condition perceived Medical Care as slightly more helpful than those in the PTM framework. However, there was no significant difference in the perceived helpfulness of Specialist and Community Care. As there was no difference observed in mental help seeking attitudes, Hypothesis 2 was rejected.

Discussion

This study is the first empirical comparison of the mental health attitudes activated by a label of schizophrenia and a psychological formulation. The analysis revealed that presenting a case of psychosis using a diagnosis of schizophrenia was associated with significantly more desired social distance from the fictional character, relative to a formulation account of psychosis. There was no significant difference between participants exposed to the diagnosis and formulation in mental help-seeking attitudes. However, the two conditions did differ in preferred treatment options: the diagnosis predicted significantly higher perceived helpfulness of medical care, though did not affect ratings of specialist or community care.

The study’s findings that the diagnosis condition evoked a significantly stronger desire for social distance than the formulation are consistent with previous research. A similar observation was made by Yang et al. (2012), who also found their vignette with a reference to psychiatric diagnoses increased desired social distance, relative to the control group who viewed no psychiatric label. Formulations provide a structured means of describing clinical cases while avoiding any need for diagnostic labelling, potentially addressing concerns that simply adjusting terminology for a schizophrenia diagnosis would still have negative connotations (Imhoff, 2016). As the stress of stigma can exacerbate mental health difficulties (Rüsch et al., 2015), promoting formulations in place of psychiatric diagnoses may improve service-user welfare by reducing exposure to negative social judgements.

There was no observed difference between the experimental conditions in general attitudes to mental help-seeking. Neither did the study detect effects on perceived helpfulness of specialist (i.e. visiting a psychiatrist or psychologist) or community care (i.e. talking to a family member or joining a social club). This is evidence against the risk that potential service users may be less inclined to seek help after encountering narrative-based framings of mental illness. Participants in both groups rated specialist care as the most helpful option of the three treatment approaches. These findings contrast with Wright et al. (2012) and Picco et al. (2018), who observed that identification of a psychiatric diagnosis was linked with endorsement of specialist treatment options. The current research finds that when the “no diagnosis” condition is structured using a formulation, it does not seem to deter specialist help-seeking relative to typical diagnostic framing. This potentially suggests that a psychological formulation’s normalisation of mental health difficulties need not form a barrier to recognising a need for professional support (Biddle et al., 2007). However, this requires further investigation with clinically at-risk populations.

Medical care (going to a hospital, visiting a GP/family doctor or the chemist/pharmacist) was rated as the least helpful of the three treatment approaches, regardless of condition. This finding could reflect a bias of the sample as 37.6% of participants had academic experience with psychology. Participants exposed to a diagnosis of schizophrenia perceived medical care as a significantly more helpful than those in the formulation condition, although this finding should be viewed with caution due to the suboptimal reliability of this measure. The DSM-5 operates from a medical model of mental health (Deacon, 2013), while the PTM framework aims to move away from the medical “mind-set” (Johnstone & Boyle, 2018). As such, findings that participants in the diagnosis condition perceived medical care as more helpful are consistent with the aims of both the DSM and PTM framework.

These results have a number of implications. The PTM framework aims to provide a formal alternative for clinicians who wish to practice in line with a paradigm shift away from a medical model of mental health (Johnstone & Boyle, 2018). As individuals with schizophrenia have high awareness of social stigma (Pandya et al., 2011), which in turn predicts self-stigma (Bathje & Pryor, 2011) and deterioration of symptoms (Rüsch et al., 2015), clinical use of formulations could help promote potential service-users’ welfare. Using a formulation could also be a method of limiting negative responses during a client’s disclosure to their friends and family. Given that the formulation condition was associated with less stigmatisation, presenting the wider public with narratives of psychosis (e.g. in the media) may help mitigate societal stigma.

Results provided little evidence to support concerns that moving away from diagnosis will impede help-seeking and treatment engagement. The two conditions produced no differences in general attitudes to help-seeking or endorsement of support from specialist mental health professionals. However, the results suggest that promoting narrative-based formulations of mental health difficulties may undermine confidence in medicalised care. This may have most serious implications within health systems where general practitioners or family doctors are the first point of contact for mental health help-seeking. Additionally, negative attitudes towards medication have been related to poor medication adherence (Levin et al., 2014). Possible interactions between formulation narratives and symptoms of psychosis, attitudes to medication and therapeutic engagement require further empirical elaboration.

Some methodological limitations of the study should be considered when interpreting the results. The vignettes were purposely developed for the study and while their development was informed by the DSM-5 and PTM framework, the text provided may have not been fully reflective of “real-world” clinical formulations. It is also relevant to note that the language in the diagnosis vignette was more technical and medical, referring to symptomatology of schizophrenia and a brief explanation of the DSM-5. While this may represent a confound in the experimental comparison, it also holds ecological validity: informal feedback from implementing PTM framework narratives suggested that the language in the framework is very different from what people usually encounter in mental health discourse (O’Toole, 2019). Future research could investigate if findings are replicated if the DSM-5 diagnostic process is described in more collaborative terms. Vignettes were purposefully gender-neutral due to evidence schizophrenia stigma differs across target gender (Holzinger et al., 2012). Leaving “Charlie” ungendered may have compromised some external validity but limited the risk of confounds due to interactions between mental illness and gender stereotypes (Wirth & Bodenhausen, 2009). However, participants were not asked what gender they imagined Charlie was and therefore gender differences could not be controlled for. Finally, vignettes were limited to a single fictional case of psychosis, so cannot be extrapolated to all manifestations of psychosis or to other mental health conditions. Future studies could investigate whether findings are replicated with other diagnoses that have fewer negative connotations, such as mood or anxiety disorders (Angermeyer & Dietrich, 2006).

The significant findings identified should be viewed with caution due to the multiple dependent measures used. Although appropriate for an exploratory study, further more defined research is necessary to confirm effects. Like most studies of mental health attitudes, the study relied on self-report measures, which may have biased results. The social distance scale queried more concrete behavioural intentions, which allowed more sensitive indicators of responses to individuals with mental health difficulties. Direct measures of behaviour would considerably benefit future research.

The final sample size was large and well-powered. Convenience recruitment strategies led to some demographic imbalances, for instance, a disproportionate number of women. However, gender and other socio-demographic attributes were statistically controlled in relevant analyses. The sample also had a relatively large number of people with experience studying psychology, whose mental health attitudes may not be representative of the wider public. Moving beyond further general population studies to quantitatively and qualitatively comparing current service-users’ experiences of diagnostic and formulation approaches would enrich the literature. Measuring effects on clinical outcomes and patient engagement is a priority for clarifying clinical implications of alternate approaches to conceptualising mental health difficulties.

The PTM framework is a guideline for developing collaborative psychological formulations with clients. The present study offered the first empirical comparison of the mental health attitudes activated by a diagnosis of schizophrenia and a psychological formulation. Findings suggest that desire for social distance from a person with symptoms of schizophrenia was significantly reduced in the formulation condition. While the formulation approach did not affect attitudes to general mental help-seeking, it did undermine confidence in medical care relative to the labelled diagnosis. However, belief in the perceived helpfulness of specialist and community care remained high in both conditions. Results suggest the psychological formulations may offer clinicians an alternative approach that avoids the public stigma associated with the diagnostic labelling of schizophrenia, with minimal risks for mental help-seeking attitudes.

Disclosure statement

No potential conflict of interest was reported by the authors.

Supplementary material

Supplemental data for this article can be accessed here.