Perceptions of Mental Health Conditions in Criminal Cases: A Survey Study Involving Swedish Lay Judges

Perceptions of mental health conditions influence how individuals with psychiatric diagnoses are treated within the community, in the legal system, and at different institutions. We examined perceptions of mental health conditions among lay judges (N = 643), working at district and appellate courts throughout Sweden. Participants read a web-based survey including a crime vignette in which the person charged with a crime was described as having schizophrenia (n = 186), antisocial personality disorder (ASPD) with psychopathic traits (n = 219), or intellectual disability (n = 238). Participants’ perceptions of schizophrenia were largely in line with Swedish legislation regarding the medicolegal concept of severe mental disturbance (SMD). Findings were more varied for the other two conditions, however. Perceptions of individuals with ASPD with psychopathic traits were not consistent with the Swedish SMD legislation. The results highlight the complexity of legislation addressing mental illness and criminality.

T he general public's understanding of mental illness is reported to have improved over recent decades. However, this understanding does not seem to have been followed by increased social acceptance of individuals with mental illness (Markowitz, 2011;Schomerus et al., 2012). Cross-cultural research has demonstrated that the public holds negative perceptions about individuals with schizophrenia including low social acceptance, distrust, and desire of social distancing (Durand-zaleski et al., 2012;Vilhauer, 2016). In clinical settings, a medical history of schizophrenia is negatively associated with expected treatment adherence and management as well as the ability to understand educational materials from treatment providers (Sullivan et al., 2015). There also seems to be a public perception that schizophrenia is related to violence and dangerousness (Angermeyer & Matschinger, 2004;Durand-zaleski et al., 2012;Jensen et al., 2016;Markowitz, 2011). This perception is partly incongruent with meta-analytical work demonstrating that, although schizophrenia and other psychoses are associated with violent behavior, in particular homicide, this association seems to be largely driven by concurrent substance abuse (Fazel et al., 2009).
Lay people's perceptions of mental illness may influence how individuals with psychiatric disorders are treated within the community, and whether they seek and comply with treatment (Corrigan et al., 2014). The past decade has seen an exponential increase in international research on the stigma of mental health diagnoses in relation to different constructs such as "dangerousness" and incompetency (Corrigan, 2016;Corrigan et al., 2014). Stigma broadly refers to social rejection through preconceptions and discrimination of individuals with mental illness. It can be manifested through self-stigma (i.e., affected individuals' internalization of negative self-images), provider stigma (i.e., health care providers' negative attitudes toward individuals with mental illness) and structural stigma, which can impair availability of various community resources (Sheehan et al., 2016).

PercePtions of Mental illness in a criMinal Justice context
Accumulated evidence of stigma associated with different types of mental health conditions prompts the need to investigate whether such attitudes also influence legal decisionmaking. Research from North America has demonstrated that preconceptions about legal concepts (e.g., insanity defense) can substantially impact jurors' general information processing and sentencing recommendations (Louden & Skeem, 2007). According to the attribution theory, people tend to infer attributions of others' behavior in relation to controllability, which then impacts the degree to which they sympathize with or want to punish the individual (Corrigan et al., 2003). Research with college students and police officers suggests there is a higher tendency of social rejection, coercive treatment, and punitiveness when the criminal behavior is perceived to be associated with a mental illness under the person's control (e.g., due to substance abuse). In contrast, offenses that are perceived to be the result of an organic cause outside the control of the person charged with a crime (e.g., due to head injury), are more likely to evoke sympathy and result in less punitive sentiments (Corrigan et al., 2003;Markowitz & watson, 2015).
In the United States, psychopathy evidence may be presented to the court to address a variety of legal questions (DeMatteo, Edens, Galloway, Cox, Smith, & Formon, 2014;DeMatteo, Edens, Galloway, Cox, Smith, Koller, & Bersoff, 2014;Edens et al., 2015; although see DeMatteo et al., 2020). when introduced, this evidence appears to impact legal decision-making (Edens & Cox, 2012). Drawing on the attribution theory, psychopathy could be used as either a mitigating or aggravating factor in court proceedings (Aspinwall et al., 2012;Edens & Cox, 2012;Remmel et al., 2019). There is a general tendency, however, to suggest harsher sentencing based on a view of the condition as a "moral illness" where the individual is capable of separating right from wrong and has chosen to commit a wrongful act (Berryessa & wohlstetter, 2019;Edens, Clark, et al., 2013). Research has also demonstrated that lay people seem to associate psychopathic traits in people charged with crimes with both semiadaptive (i.e., bold and intelligent) and maladaptive (i.e., dangerous and evil) features (Edens, Davis, et al., 2013). A recent meta-analysis investigated perceptions of the psychopathy label in relation to various punishment outcomes (i.e., dangerousness, treatment amenability, legal sentence/sanction) across 22 studies (Berryessa & wohlstetter, 2019). In studies comparing people charged with crimes that meet criteria for psychopathy and people charged with crimes with no such label, the psychopathy label has been weakly, but significantly associated with stronger support for punitive sanctioning and negative perceptions of treatment potential. In studies comparing people charged with crimes that meet criteria for psychopathy versus those with another psychiatric label (i.e., conduct disorder, antisocial personality disorder [ASPD], paraphilic disorder), the effect sizes have been weak and nonsignificant. Overall therefore, the results have demonstrated a general labeling effect of psychiatric disorders, rather than psychopathy specifically (Berryessa & wohlstetter, 2019). Considering psychopathy assessments are commonly used as part of a larger risk-assessment of life-sentenced prisoners within the Swedish legal system (Sturup et al., 2014), these discrepant findings bolster the need to further study perceptions of the psychopathy label among lay people.
Although psychopathy may be viewed in some instances as a mitigating factor (i.e., when viewed as a mental illness; Aspinwall et al., 2012), research suggests it is generally considered aggravating (Boccaccini et al., 2008;Edens et al., 2004Edens et al., , 2005. A meta-analysis of simulation studies suggests jurors who perceive a person charged with a crime exhibiting psychopathic traits also believe the person to be dangerous and evil (Kelley et al., 2018). Jurors may also be more likely to recommend more punitive sentencing. In a recent study specifically regarding juveniles, where study participants encompassed community members summoned for jury duty (n = 326), the fictitious young person charged with a crime was perceived as particularly evil and dangerous when described to exhibit affective psychopathic traits (e.g., lack of remorse or guilt, shallow affect, callousness, lack of empathy; Edens et al., 2016). Furthermore, population-based surveys have indicated that the general public tends to view individuals with psychopathic traits as crime-prone, yet socially skilled and interpersonally adept (Furnham et al., 2009;Smith et al., 2014). In addition, some lay people may conflate the semantically similar labels "psychopathy" and "psychosis" (Edens et al., 2004;Smith et al., 2014).
Although there are few studies to date which have investigated perceptions about intellectual disability (ID; Ditchman et al., 2013;Scior, 2011;Scior & Furnham, 2016;werner et al., 2012), the available research suggests the general public holds negative views about individuals with ID and the ID population faces discrimination in several settings (e.g., health care, housing, employment; Ditchman et al., 2013;werner et al., 2012). Regarding ID in criminal court specifically, one Swedish study investigated court cases published between 2004 and 2006 concerning alleged child sexual abuse when the victim had a neuropsychiatric disorder (n = 14; Lindblad & Lainpelto, 2011). The results demonstrated that victims with ID (n = 10) were considered more credible and trustworthy because the courts considered their cognitive capacity as precluding them from fabricating facts. The results also suggest courts may draw conclusions about developmental aspects of neuropsychiatric disorders in the absence of any expert testimony.
These data illustrate the complexity at the interface of law and psychiatry and also point toward a need for an improved understanding of mental health evidence in criminal cases. More research is needed to understand how predetermined attitudes of mental health "labels" and associated symptoms influence legal decision-making (e.g., regarding treatment decisions and criminal responsibility; Jung, 2015).

tHe swedisH legal systeM and lay Judges
Swedish criminal law is unique in the sense that since the enactment of the present criminal code approximately 50 years ago (1965), accountability is not an independent demand for conviction. This means that all people charged with crimes, irrespective of mental status, in principle can be held criminally responsible for their actions. Individuals charged with crimes who suffer from various mental disorders at the time of the crime can therefore be convicted, provided that they meet the "normal" demands for conviction (the actus reus, the required intent, absence of justifying and excusing conditions, etc.). In some cases, such "normal" demands are not met because of the mental disorder (e.g., the disorder produced a lack of intent). Since 1965, issues related to mental disorders instead are given formal room within sentencing, including the choice of sanctions.
According to the Swedish Criminal Code (Chapter 30, Section 6; Chapter 31, Section 3), a person who committed an offense under the influence of a serious mental disturbance (SMD) is preferably sentenced to a sanction other than imprisonment. The court may only impose imprisonment if there are exceptional grounds to do so. The court may not impose a sentence of imprisonment if, as a result of the SMD, the accused lacked the capacity to realize the implications of the act or to adapt their conduct accordingly. However, this does not apply if the accused induced this lack of capacity themselves. If a person who has committed an offense for which a fine is considered an insufficient sanction suffers from SMD, the court may order them to undergo forensic psychiatric care if, in view of their mental state and other personal circumstances, it is called for that they be admitted to a medical institution for psychiatric care, combined with custodial and other coercive measures. The SMD concept is not linked to a particular psychiatric diagnosis, however the majority of individuals with SMD have a psychotic condition (e.g., schizophrenia). Other conditions include severe depression with suicidal intent, severe personality disorder with psychotic episodes, ID with severe compulsive behavior, severe neurocognitive disorder, severe intellectual disability, and severe brain damage.
In the rare cases where there is suspicion that a crime has been committed under the influence of SMD, court-ordered forensic psychiatric evaluations (FPEs) are conducted before sentencing. FPEs are conducted by the National Board of Forensic Medicine (NBFM), a governmental authority subsumed under the Ministry of Justice. According to official NBFM statistics, approximately 500 FPEs are conducted per year. Although the FPE provides a recommendation to the court (i.e., regarding whether the person convicted of a crime should be sentenced to forensic psychiatric care instead of prison), ultimately the court determines sentencing. The court therefore needs to evaluate the extensive information regarding the person's background and psychiatric status encompassed in the FPE.
Given the involvement of lay judges in this process, it is important to understand their knowledge and perceptions of mental health conditions in relation to the SMD concept. In the Swedish legal system lay judges are nominated by political parties and elected in the municipal council or county borough council. Formal qualifications for lay judges include having Swedish citizenship, being 18 years or older, and not having a criminal record. Some professions are also exempt including police officers and individuals working for the courts. when nominated, lay judges commonly serve for 4 years. In contrast to the United States where eligible community members are summoned for jury duty and selected for a specific trial, in Sweden the legal sides have no influence on who is serving as a lay judge in any particular case. Lay judges serve together with professional judges in criminal proceedings to decide on matters of guilt and sentencing. The vote of the lay judge carries the same weight as that of the professional judge. The professional judge, however, has the responsibility to present and explain the legal aspects and relevant issues of the case to the lay judges. Although professional judges are in the majority in appellate courts, in lower courts (i.e., the district courts), lay judges outnumber their professional counterparts.
Although firmly established dating back to the medieval era, the Swedish system of lay judges has been criticized (Alhem, 2012;Schultz, 2011), fueled by cases where disqualifications of single lay judges has caused a retrial of costly court proceedings (wahlberg, 2012). In 2012, the Swedish government commissioned an inquiry that resulted in some changes to the law including more stringent requirements regarding suitability for the appointment as a lay judge and a mandatory introductory training program, according to the Government Offices of Sweden (2014). This mandatory training focuses on conflicts of interest, secrecy, and ethical aspects. However, it does not include any training on mental health evidence.
To our knowledge, only one study has examined perceptions of mental health conditions and the Swedish concept of SMD. Sygel et al. (2017) utilized a vignette-based study to explore the impact of psychiatric diagnosis and gender in people charged with a crime, when forensic evaluators deliver their opinions regarding the SMD designation. Researchers recruited 26 evaluators (i.e., forensic psychiatrists, forensic psychologists, and forensic social workers) conducting FPEs at the NBFM and provided a description of a fictitious person accused of serious assault with varying psychiatric diagnoses (i.e., schizophrenia; borderline personality disorder; substance-induced psychotic disorder; Asperger syndrome; ASPD; intellectual disability). All cases included two versions, varying the sex of the fictitious person accused of assault. All participants considered schizophrenia to be associated with SMD (across genders of the persons accused of assault), and no participants assigned SMD to the condition in which the person was described as having ASPD (across genders). In the ID case, 12.5% of participants considered ID to be associated with SMD. These results imply that forensic evaluators have a clear understanding of differences between psychotic disorder and ASPD in relation to the SMD concept. Data also indicate that opinions vary regarding intellectual disability, where the designation of SMD can depend on various additional factors such as severity of the condition and low psychosocial functioning. To our knowledge however, no study has examined how Swedish lay judges perceive different mental health conditions in relation to legal decision-making.

tHe current study
Given the stigma that is associated with many mental health diagnoses, the goal of the current study was to evaluate if this stigma translated into differential decision-making among Swedish lay judges. Using a web-based survey with a brief vignette, participants were randomly assigned to three different conditions where the person charged with a crime was described as having (a) schizophrenia, (b) ASPD with psychopathic traits, or (c) ID. 1 The case vignette method is an established method to investigate potential bias in clinical judgment (Aspinwall et al., 2012). These specific conditions were chosen based on their different degree of association with the SMD concept in the Swedish legal system. Schizophrenia is strongly associated with SMD, whereas ASPD with psychopathic traits is unrelated to SMD (Sygel et al., 2017). Furthermore, the association between ID and SMD is generally contingent on degree of severity and disability.
Based on the Swedish legislation regarding SMD, as well as previous research (Sygel et al., 2017), we hypothesized that participants would perceive schizophrenia to be positively associated with SMD and negatively associated with comprehending the meaning of the criminal act. In contrast, we hypothesized that participants would perceive ASPD with psychopathic traits to be negatively associated with SMD, and positively associated with comprehending the meaning of the criminal act. Regarding the ID diagnosis, the investigation was exploratory and no specific hypothesis was made. A secondary aim was to investigate participants' general attitudes and perceptions about individuals with the assigned mental health condition (e.g., regarding perceived prevalence in the community, crime proneness and culpability, punitive sentiments). Such general attitudes and perceptions are informative to delineate aspects of potential stigma.

MetHod ParticiPants
Male and female lay judges in Sweden were invited to participate in this study. Participants were recruited from six district courts and two appellate courts located in different cities across the country. In total, 1,405 individuals were invited to participate and an overall response rate of 51.3% resulted in 721 total participants. Of these, 89.2% passed the manipulation check (i.e., assigning the correct diagnosis to the fictitious person in the vignette, further described below), which resulted in a final sample of 643 participants. Participants' ages ranged from 20 to 84 years (M = 57.3 years, SD = 14.8), and the sample was relatively evenly split between women (53.8%) and men (46.2%).
The majority of participants were either from Stockholm or Gothenburg (the two largest cities in Sweden), and relatively well educated (47.3% had attended university or college for more than 2 years). Regarding work experience, most participants (63.0%) had been appointed as a lay judge for 5 years or less. The majority of participants (77.6%) had been involved in a maximum of five court cases involving a forensic psychiatric evaluation (FPE). About a third (33.4%) had obtained some formal education about mental health, with the majority receiving this education outside of the court system. Furthermore, most participants (58.9%) endorsed the question "Have you/anyone in your family been subjected to a crime?" Among these respondents, 62.3% reported to have been victimized 2 to 5 times, and 47.9% reported that at least one of the crimes was a violent crime. For a detailed description of participant characteristics across conditions, see Table 1. Note. N is based on number of participants passing the manipulation check. ASPD = antisocial personality disorder; ID = intellectual disability; FPE = forensic psychiatric evaluation. a "Have you/anyone in your family been subjected to a crime"?

stiMulus Materials demographic information
Participants completed basic demographic questions (i.e., age, gender, educational level, time appointed as lay judge, city where the court is situated). This section also included two questions about professional experience with mental health cases. Specifically, participants reported the number of court cases in which they have been involved where an FPE was included, and whether they have obtained any formal education regarding mental health. Finally, participants responded to a question about victimization (i.e., "Have you or anyone in your family been subjected to a crime?"). If the participant responded affirmatively, follow-up questions assessed the frequency and severity of this victimization.

crime Vignette and fPe
In all study conditions, participants received the same crime vignette (104 words long; see Appendix) adopted from previous research . The case described a street robbery, where a female pedestrian ("Mrs. K") was approached by a stranger ("Stenberg") who grabbed her handbag and slashed her across the face with a pocket knife. The vignette was followed by a forensic psychiatric statement, which indicated that the court had requested an FPE for Stenberg. The FPE included a description of his current mental health condition followed by a description of his behavior prior to and following the crime (see descriptions below). Participants were randomized into one of three conditions where the fictitious character, Stenberg, was described as having (a) schizophrenia, (b) ASPD with psychopathic traits, or (c) ID. The statement (222-282 words long) described personality traits and behaviors characteristic for the respective conditions. The mental health conditions were created by three authors (JC, MK, KH) who are forensic psychiatrists or psychologists with extensive clinical experience, using criteria outlined in the Diagnostic and Statistical Manual of Mental Disorders (4th ed., text rev.; DSM-IV-TR; American Psychiatric Association [APA], 2000). 2 Furthermore, the vignettes were formulated to include information similar to the FPE statements delivered to the courts in the Swedish legal system. The modeling of real-life scenarios had three important implications: (a) the vignettes included both mental health diagnoses and behavioral effects of these diagnoses, (b) given that all people that are sentenced in the Swedish system (i.e., culpability is not assessed during the FPE) no control condition was needed, and (c) there was a difference in the admission of guilt across conditions, which was deliberate and intended to be consistent with how an individual with each diagnosis typically responds. As mentioned before, in the Swedish judicial system accountability is not an independent demand for conviction. when the individual is referred for a FPE the court has determined whether he or she is guilty of the offense, therefore the admission of guilt is not an issue in the FPE and does not impact the decision about SMD.

schizophrenia
In this condition, Stenberg was described as experiencing symptoms consistent with schizophrenia, paranoid type such as being "confused," "talking about undercover agents," "talking to himself," and "suspicious." Furthermore, he had stopped taking his prescribed antipsychotic medication at the time of the incident. In this condition, he confessed to the crime.

asPd with Psychopathic traits
In this condition, Stenberg demonstrated personality traits and behaviors in line with the ASPD criteria, but also including specifically psychopathic features, loosely based on the Psychopathy Checklist-Revised (PCL-R; Hare, 2003) criteria. These included multiple previous convictions including robbery, assault, and financial fraud. Further descriptors included "no remorse," "stating he is a victim of circumstances," "dominant," "socially bold," "lied repeatedly," "tried to manipulate other detainees and staff," "arrogant," and "excessively friendly." In this condition, he denied guilt.

id
In this condition, descriptors included "never had independent housing," "never held normal employment," "intellectual level comparable to an individual age 7 to 9 years," "had set the house on fire on a previous occasion," and "requires supervision for many daily tasks and reminders about personal hygiene."

Manipulation check
To assess whether participants attended to the crime vignette and instructions, they were asked to identify Stenberg's mental health condition from three options (i.e., schizophrenia, ASPD with psychopathic traits, ID).

Mental Health and Verdict Questionnaire
Following the manipulation check, participants determined whether they thought Stenberg was under the influence of an SMD at the time of the criminal act as well as at the time of the FPE. Participants who endorsed this were also asked whether they thought he "due to the SMD lacked the ability to comprehend the meaning of the act." Participants were also asked what sentence they would recommend: (a) prison or (b) compulsory forensic psychiatric treatment. Participants who recommended the latter were asked if such treatment should include special court supervision. Finally, all participants were asked how prevalent they believe the specific mental health condition to be in the community.

attitudinal statements
Participants then responded to six attitudinal statements about individuals with the given mental health condition. These statements, rated on a Likert-type-scale from 1 (strongly agree) to 7 (strongly disagree), were identical across conditions and assessed perceptions of crime proneness, legal responsibility, and general punitive attitudes.

Procedure
Data collection was completed in May 2016. All participants were informed about the study in writing, with identical study information provided to all participating courts. Administrative contact persons at the respective courts were responsible for sending out study information through e-mail. The study information contained a link to the web-based survey (one separate survey for each condition), hosted through the Google docs platform. Contact persons were instructed to send out each link to one third of the potential participants, to ensure approximately equal numbers in the conditions. Participants received a follow-up e-mail approximately 2 weeks following the initial e-mail. Participation was anonymous and participants were instructed not to discuss the study with colleagues. The study received an advisory statement 3 from the Regional Ethical Review Board of Stockholm (#2015/1786-31/5).  Table 1 includes demographic information for the three separate conditions. No significant differences emerged between conditions. Item endorsement percentages are presented in Table 2. when asked if the fictitious person charged with the crime met criteria for SMD at the time of the crime, a chi-square analysis indicated significant differences between the proportion of participants expected to determine that he meets criteria for SMD and the proportion of participants who actually found him to meet criteria (see Table 3). Follow-up chi-square tests directly compared each condition, however, a stricter alpha was set at .025 to reduce the likelihood of a Type 1 error. More participants in the schizophrenia condition identified him as SMD compared with both the ASPD with psychopathic traits condition and the ID condition (see Table 3). However, there was no significant difference between the ASPD with psychopathic traits and ID conditions (see Table 3). Participants were also asked if Stenberg met criteria for an SMD at the time of the FPE. A similar pattern emerged, with the proportion of participants expected to find him under the influence of an SMD at the time of the evaluation significantly lower than the proportion of participants who made this determination (see Table 3). Again, this difference was driven by the schizophrenia condition, with endorsement rates significantly different at the stricter alpha level of .025, than either the ID condition and the ASPD with psychopathic traits condition (see Table 3). Furthermore, there was no significant difference between the ASPD with psychopathic traits and ID condition (see Table 3).  (109) Note. ASPD = antisocial personality disorder; ID = intellectual disability; SMD = serious mental disturbance; FPE = forensic psychiatric evaluation.
If a participant believed that Stenberg was under the influence of an SMD at the time of the FPE, s/he was also asked if they thought he was unable to comprehend the meaning of the criminal act or had the ability to adjust his actions accordingly. A chi-square analysis indicated differences between groups. Specifically, more participants affirmatively endorsed this item in both the Schizophrenia and ID conditions, compared with the ASPD with psychopathic traits condition (see Table 2). Follow up chi-square analyses with an adjusted significance value of .025 to reduce the likelihood for Type 1 error indicated a significant difference between the schizophrenia and ASPD with psychopathic traits conditions (see Table 3). Furthermore, a significant difference emerged between the ID and ASPD with psychopathic traits conditions. However, there was no difference between the Schizophrenia and ID conditions (see Table 3).
A similar pattern emerged with the sentencing variable in that a chi-square analysis indicated a significant difference between groups. Participants in the ASPD with psychopathic traits condition were significantly more likely to impose a prison sentence compared with participants in the schizophrenia condition and ID condition (see Table 3). Furthermore, Note. Schizophrenia-ASPD = comparison of schizophrenia and antisocial personality disorder conditions; Schizophrenia-ID = comparison of schizophrenia and intellectual disability conditions; ASPD-ID = comparison of antisocial personality disorder and schizophrenia conditions. SMD = serious mental disturbance; ASPD = antisocial personality disorder; ID = intellectual disability; FPE = forensic psychiatric evaluation. *p is significant at the .025 level. **p is significant at the .001 level.
although participants were more likely to recommend institutional compulsory forensic care in both the schizophrenia and ID conditions (see Table 2), there was also a significant difference between these two groups, with participants in the ID condition more likely to recommend a prison sentence (see Table 3). If a participant recommended compulsory forensic psychiatric care, they were also asked to specify if this treatment should include special court supervision (see Table 2). Significant differences in this dependent variable emerged. Specifically, participants in the schizophrenia condition were more likely to recommend special court supervision compared with the ID condition, but not more likely to recommend special supervision compared with the ASPD with psychopathic traits condition (see Table 3). Furthermore, participants in the ASPD with psychopathic traits condition were more likely to recommend special court supervision compared with participants in the ID condition (see Table 3).
Finally, participants were asked about perceived community prevalence rates of individuals with the assigned diagnosis (see Table 4). To facilitate clear comparisons, the response options were dichotomized into the following: 1 out of 100 or more (n = 251) versus 1 out of 1,000 or less (n = 392). There was a significant difference between groups, driven by differences between the Schizophrenia condition and both the ASPD with psychopathic traits and ID conditions (see Table 3). There was no difference between the ASPD with psychopathic traits and the ID conditions (see Table 3).

attitudinal stateMents
Participants were also asked to respond to six attitudinal statements about the diagnosis for their respective condition (see Table 5). One-way analyses of variance (ANOVAs) were conducted to examine potential differences between subgroups, with Bonferroni adjusted alpha levels of .0167 (.05/3) per test. For the statement regarding crime propensity, the levels of endorsement significantly differed across all three subgroups. This difference was driven by the ASPD with psychopathic traits and ID conditions (see Table 6). Regarding responsibility (i.e., specifically concerning whether individuals with the assigned diagnosis who commit crimes can be fully blamed for their acts), the levels of endorsement also significantly differed across all three subgroups. Again, the largest difference occurred between the ASPD with psychopathic traits and ID conditions (see Table 6). Regarding the statement whether individuals can understand the difference between right and wrong, there was a significant difference between subgroups. Specifically, participants in the ASPD with psychopathic traits condition provided higher ratings than either the schizophrenia or ID conditions (see Table 6). Regarding sentencing (i.e., in particular whether individuals with the assigned diagnosis should be classified as SMD and sentenced to compulsory forensic Note. ASPD = antisocial personality disorder; ID = intellectual disability. psychiatric treatment), levels of endorsement also significantly differed between groups, with the largest difference occurring between the schizophrenia and ID conditions. Participants in the ASPD condition also supported the statement that individuals with ASPD should be treated more harshly by the criminal justice system, compared with participants in the ID condition (see Table 6). Finally, there was also a significant difference between groups on participant ratings of the statement "an individual with _____ should be locked up regardless of whether he or she has committed a crime." Participants in the schizophrenia condition more strongly supported this statement than participants in the ID condition (see Table 6).

discussion
In the present study, we choose to investigate the judicial concept SMD in the Swedish Criminal Code, Chapter 30, Section 6. This is a complex, multiprofessional and multidisciplinary concept, applied in Sweden since 1992. It is not related to specific psychiatric diagnoses but to the character of the mental disorder and the effects on psychosocial functioning. This means that there will be many variables that will differentiate between those with and without an SMD. Our aim was to learn more about the perceptions of a person charged with a crime and subject to the judgment of SMD.
This study was the first investigation of legal decision-making and perceptions of mental illness in a large and representative sample of Swedish lay judges. This is a unique study group, which may contribute to improved understanding of perceptions of mental illness within a criminal justice context. Results suggest participants viewed the three mental health conditions differently within the context of Swedish SMD legislation. Specifically, Note. Attitudinal items are coded from 1 (strongly disagree) to 7 (strongly agree). ASPD = antisocial personality disorder; ID = intellectual disability; SMD = serious mental disturbance.
perceptions of schizophrenia were largely consistent with the Swedish SMD legislation. However, findings were more varied for ASPD with psychopathic traits and ID.
In the schizophrenia condition, the large majority of participants perceived the fictitious person charged with the crime to be under the influence of an SMD at the time of the crime, at the time of the forensic evaluation and, due to SMD at the time of the crime, unable to comprehend the meaning of the criminal act or adjust his behavior accordingly. Participants were also very likely to recommend compulsory forensic psychiatric care with special court supervision. This was consistent with hypotheses and corresponds with Swedish legislation regarding SMD. Note. Schizophrenia-ASPD = comparison of schizophrenia and antisocial personality disorder conditions; schizophrenia-ID = comparison of schizophrenia and intellectual disability conditions; ASPD-ID = comparison of antisocial personality disorder and schizophrenia conditions. ASPD = antisocial personality disorder; ID = intellectual disability; SMD = serious mental disturbance. a Indicates difference is significant at a .0167 level (Bonferroni correction).
Findings from the ASPD with psychopathic traits condition were somewhat unexpected and inconsistent with hypotheses. Specifically, almost half of the participants perceived the fictitious person charged with the crime to be under the influence of an SMD at the time of the crime and at the time of the evaluation. Furthermore, 51.6% of participants recommended compulsory forensic psychiatric care instead of prison. These findings suggest that participants may have an unclear conceptual understanding of ASPD with psychopathic traits, and its association with the SMD concept. It is possible participants placed a strong emphasis on the "personality disorder" part of the term, which may give the connotation of a mental illness 4 and lead one to be more supportive of a forensic psychiatric care sentence. Research indicates people think of personality disorders in general as very debilitating and significantly impacting the individual's daily functioning (Sheehan et al., 2016). As such, these data suggest the participants in this study may have considered the ASPD with psychopathic traits symptoms as similarly impacting the life of the person charged with a crime, without the pejorative effects that have emerged in other legal decision-making studies (i.e., Boccaccini et al., 2008;Edens et al., 2005). However, in Swedish legislation and practice, SMD encompasses psychotic disorders. Personality disorders with antisocial and psychopathic traits should not be included unless they co-occur with psychotic symptoms, severe mental illness, or psychotic-based impulsivity according to the Swedish Criminal Code (Chapter 30, Section 6; Chapter 31, Section 3). Of course, severe dysfunctional social behavior would likely impact the FPE assessors' judgment.
Of note, data from the current study are inconsistent with a recent survey of evaluators (n = 26) at the NBFM in Sweden which demonstrated that all participants considered schizophrenia to be associated with SMD, while no participants assigned SMD to the ASPD condition (Sygel et al., 2017). This suggests a disconnect between forensic evaluators and lay judges in determining the appropriateness of an ASPD diagnosis for an SMD designation. Specific reasons for this disconnect, such as different education and training backgrounds, misapplication or misunderstanding of the legal statute, or perceptions of mental health diagnoses, should be explored in future research.
There is no consistent association between ID and the Swedish SMD psycholegal concept. ID can range in severity and impact of symptoms on everyday functioning. In some severe cases (e.g., when occurring in combination with psychotic symptoms), it might be consistent with the SMD designation. However, it may also constitute a condition with limited disabilities and, therefore, be inconsistent with the legal definition of SMD. As such, our investigation in this condition was more exploratory. Approximately half of the participants in the ID condition perceived the fictitious person charged of the crime to be under the influence of an SMD at the time of the crime and the evaluation. A vast majority of participants (88.3%) perceived him to be unable to comprehend the meaning of the criminal act or adjust his behavior accordingly and recommended compulsory forensic psychiatric care (with a relatively even number of participants suggesting with or without special court supervision). The heterogeneity in the perceptions of ID provides a strong reason to supply courts with psychiatric evaluations in which specific characteristics of individuals with this condition are explained.
Some general trends emerged regarding the attitudinal statements. Overall, participants viewed individuals with ASPD and psychopathic traits as crime prone, better able to understand the difference between right and wrong, and more culpable for their actions. Regarding punitive sentiments, across conditions there was no strong tendency to vote for harsh sentencing. Although participants in the ASPD with psychopathic traits condition more heavily endorsed the statement that "individuals with the assigned diagnosis should be treated more harshly by the criminal justice system," the endorsements were low overall. This may be a contextual factor with a Swedish tradition of psychiatric care for individuals with mental health conditions. In contrast to the punishment-oriented culture in the United States, there is a strong tradition within the Swedish society-and subsequently the criminal justice system-to rehabilitate individuals who have committed crimes. Therefore, in the Swedish context it is reasonable to expect that participants perceive compulsory forensic psychiatric care as a more lenient "punishment" than prison.
Consistent with the strongly held, although largely debunked, lay belief that severe mental illness is correlated with violence (Appelbaum et al., 2000;Fazel & Grann, 2006), participants in this study tended to perceive individuals diagnosed with schizophrenia as crime prone. Participants in the schizophrenia condition also provided slightly higher ratings, compared with participants in the other conditions, for the statement "individuals with the assigned diagnosis should be locked up to protect society regardless of whether they have committed a crime." Although the average rating for this item was low (M = 2.16, SD = 1.65), this marginal difference might reflect a general public perception of individuals with schizophrenia as "dangerous." Despite this, participants did not believe individuals with schizophrenia should be fully blamed for their acts or understand the difference between right and wrong.
Results across conditions indicate that there are some misconceptions concerning the Swedish SMD statue in relation to different mental health conditions. In Sweden, the court can summon the retained forensic psychiatrist to clarify their statements in the FPE, however, this is seldom used in practice. In the vast majority of cases the court follows the sentencing recommendations outlined in the FPE. However, given that professional and lay judges independently evaluate the FPE and render the final decisions, a misconception concerning implications of different mental health conditions may be problematic. Although lay judges commonly vote in line with the professional judges, in some cases they might dissent. Moreover, the FPE commonly includes extensive information on the psychiatric status and diagnostics of the person convicted of a crime, which the court must evaluate. This information can be valuable to the courts in several important ways including decisions on credibility and sentencing recommendations. Potential misconceptions among judges regarding psychiatric diagnoses and mental illness constitutes an impending risk that the information is disregarded or misinterpreted. For example, previous research has demonstrated that a victim's mental illness may serve to increase or decrease their credibility (Lainpelto et al., 2016;Lindblad & Lainpelto, 2011).
It is a fundamental problem when the court draws conclusions about mental health conditions that are not empirically founded. Lay judges' preconceptions of mental health conditions might ultimately impact sentencing recommendations or lead to differential treatment of people convicted of crimes. One way to address this problem could be to provide the professional judges with better tools to interpret the FPEs (and additional mental health evidence) so that they are better able to explain the important aspects and consequences to the lay judges. This brings forward a more general question: should professional and lay judges receive specialized training in mental health etiology, behavior, and treatment? The majority of participants (>60%, across conditions) had no formal education about mental health and had only participated in a maximum of five court cases involving an FPE (>70%, across conditions). This is reasonable, given that cases where an SMD is evaluated in court are rare. However, the disconnect between the reality of mental illness and perceptions of mental illness among lay judges may be cause for concern however. Across the Unites States and other western countries (e.g., United Kingdom, Canada, Australia) there are specialized mental health courts that were established with the overall goal to decrease recidivism for people charged with crimes that have co-occurring mental illness (Edgely, 2014;Loong et al., 2016). Mental health courts attempt to end the cyclical pattern of individuals with mental illness in the criminal justice system by referring them to community-based psychological and psychiatric treatment programs under prolonged court supervision (Erickson et al., 2006). Across courts, individuals with mental illness must voluntarily decide to enter into mental health court in exchange for court mandated and monitored mental health treatment (Honegger, 2015). A fundamental aspect of the process at these courts is that the person convicted of a crime appears before a judge several times and that options of treatment and interventions are continuously evaluated by the court. Judges in mental health courts are extensively involved in the planning of rehabilitation and community interventions (e.g., regarding job, housing) for the person convicted of a crime (Edgely, 2014). In mental health courts, it is typical for the judge and the legal sides (i.e., prosecutor and defense attorney) to have received specialized training in mental health disorders (watson et al., 2001), presumably improving their ability to work effectively with people charged with crimes that have co-occurring mental illness. Research examining the appropriateness of such training for Swedish lay judges, as well as the effectiveness of this training in increasing knowledge of mental health disorders, is necessary.

iMPlications and future researcH
The interface between law and psychiatry is complex and multifaceted. In addition to cases that involve an FPE, there are more common situations where the court has to evaluate consequences of mental health conditions in a person charged of a crime. For example, in all criminal cases the court must evaluate the credibility of a person charged with a crime and there may also be a need to consider potential alternatives to prison sentences. To bridge the gap between the two fields, one potential area concerns training in interpreting the written statements and oral communications from either discipline. At present, there is no real forum for sharing knowledge and discussing the different aspects of common matters. It is possible that Swedish courts draw from the examples posed by U.S. mental health courts in developing better pathways of communication between courtroom players. Further, Swedish courts may consider a broader involvement of the psychiatry field and a more dynamic sentencing system of individuals with mental illness.
Future research may consider how lay and professional judges go about the decisionmaking process (Robbennolt, 2005). Neurobiological research has demonstrated that decisions are influenced by both fast (i.e., unconscious) and slow (i.e., conscious and influenced by culture and education) aspects (Frith & Singer, 2008). Previous research has also demonstrated that political affiliation might impact legal decision-making . Social and cognitive psychology is ripe with theories and evidence regarding reasoning and decision-making processes, however, the application of these theories to the study of professional judges and lay judges is lacking. Future studies on judges could unpack what elements come into play when interpreting mental health evidence (Robbennolt, 2005).
Furthermore, statistical analyses employed in this study did not allow for researchers to make causal determinations regarding the relationship between mental health label and lay judge decision-making. Future research should consider if the SMD designation mediates the relationship between mental health diagnosis and legal decisions such as sentencing recommendations. A mediation model may also consider how lay judges' individual differences (e.g., mental health education, professional experience) may further explain the relationship between mental health label and legal decisions (i.e., SMD designation, sentencing recommendation). A more complete understanding of any causal mechanisms may provide insight into necessary treatment and interventions.

strengtHs and liMitations
This study has several strengths including a large sample size of real world decisionmakers recruited from district and appellate courts across Sweden. The questions in the protocol regarding SMD were phrased similarly to corresponding sections in the FPE statement delivered to courts, which adds to the ecological validity of the study. Furthermore, participants' age range and gender distribution were largely consistent with official statistics according to the Swedish National Courts Administration (2016). Regarding experience of court proceedings, the vast majority of participants (75% or more, across conditions) had only been involved in a maximum of five court cases involving an FPE. That is reasonable, given that FPEs are highly resource-intensive and conducted only under rare circumstances. It is worth noting that around 27% to 33% of participants across conditions had been appointed as lay judge for more than 5 years. This is lower than the corresponding figure (i.e., 47%) in official statistics according to the Swedish National Courts Administration (2016). Despite this, overall, our study group can be considered highly representative of the larger population of Swedish lay judges.
This study is also marked by several limitations that must be considered in the overall interpretation and application of the data. The differences in the admission of guilt between the three experimental groups could represent a confounding variable given that admission of guilt may impact conviction rates. As noted above, criminal culpability is not a factor in the FPE evaluation, thus lay judges should not consider the admission when making their SMD determination. Despite this, we recognize this difference between conditions may have inadvertently impacted lay judge perceptions of the case and the fictitious person charged with a crime.
Participants were responding based on diagnostic labels and associated symptoms that go along with these labels. Given that no control group (i.e., where the fictitious person charged with a crime is not described to have any mental health condition) was included, it is not possible to conclude whether there was a "general labeling effect" of mental health conditions among our participants (Boccaccini et al., 2008). Moreover, it was not possible to tease apart the potential effects of labels and symptoms. The aim of this study was to model real-life scenarios where the court is presented with both diagnostic labels and symptoms. The goal, therefore, was to examine how participants apply legal statutes given their perceptions and understanding of mental health conditions (i.e., diagnoses and adjacent symptoms), rather than diagnostic labels per se. Furthermore, there were certain differences between the vignettes (e.g., psychiatric medication, criminal history). These differences were deliberately included to increase the ecological validity of the study, however, we acknowledge that such differences may explain some of the variance in lay judges' decisions. Moreover, the vignette on ASPD with psychopathic traits mainly included interpersonal-affective psychopathic traits, with one descriptor reflecting antisocial traits (i.e., criminal versatility). It is possible that this affected the perception of the fictitious character and future similar studies should tease apart potential differences in perceptions of affective/interpersonal versus lifestyle/antisocial psychopathic traits in persons charged with crimes.
Methodologically, the response rate (51.3%), could imply a selection bias. Furthermore, it was not possible to fully randomize the invitation for each condition at the different courts and it is possible that lay judges from a particular court are overrepresented in a specific condition. Therefore, participation in some conditions might be biased from individual courts. Given the equal numbers of participants in each condition however, the risk of systematic errors in the inclusion process is considered low. Finally, given the context-specific stimulus material, our findings are not easily generalized to other legal systems.

conclusion
This study demonstrated that Swedish lay judges serving at district and appellate courts view legal aspects of schizophrenia largely in line with the Swedish SMD legislation. Their perceptions of ASPD with psychopathic traits, however, were inconsistent with the Swedish SMD legislation. Overall, participants did not clearly distinguish between behavioral-based diagnoses (i.e., ASPD) and psychiatric disorders (i.e., schizophrenia). Training regarding mental health diagnoses may be necessary for professional and lay judges. Furthermore, an ongoing discussion between different disciplines (i.e., law, forensic psychiatry) regarding implications of different mental health conditions, is warranted. Such ongoing discussion may give the professional judges better tools to explain implications of mental health conditions to the lay judges.

aPPendix criMe Vignette
On the evening of 25 October, Mrs. K left her apartment to go shopping. On her way to the subway station, she was approached by a man, Stenberg, who asked for directions. Stenberg grabbed Mrs. K's handbag and threw her to the ground. He slashed her across the face with a pocket knife and then ran into the subway station with her handbag. A passerby witnessed the crime from across the street and came to offer Mrs. K aid. The police were called, and 20 min later, Stenberg was arrested. when arrested, Stenberg was carrying Mrs. K's wallet along with a pocket knife with dried blood on it.

condition a
The court ordered a forensic psychiatric evaluation, completed by court-ordered evaluators. It suggested that Stenberg has antisocial personality disorder (ASPD) with psychopathic traits. Stenberg had been released from prison a couple of weeks before the incident. He had 25 previous convictions including robbery and assault, as well as tax fraud and other financial frauds. Stenberg was a construction worker and had his own company. Stenberg showed no remorse for his action and several times during the evaluation he kept stating that he was the victim of bad circumstances and that he constantly had been duped by others. Throughout the entire investigation, he claimed he was innocent and that Mrs. K had mistaken him for another person. At the forensic psychiatric unit, Stenberg was perceived as dominant and socially bold. He repeatedly lied and tried to manipulate other detainees and staff. when he first arrived to the unit, Stenberg tried to smuggle narcotics. The social evaluation revealed that Stenberg had four children with four different women. At the time of the evaluation, he had no apartment of his own but lived in his past girlfriend's home. During interview sessions, Stenberg acted arrogant and superior, but could also change to suddenly being flattering and excessively friendly. In the police investigation, there was evidence indicating that Stenberg might have tried to sell Mrs. K's purse to another person.

condition B
The court ordered a forensic psychiatric evaluation, completed by court-ordered evaluators. It suggested that Stenberg has schizophrenia, paranoid type. At the time of the incident, Stenberg had been attending outpatient psychiatric treatment for several years. At his outpatient clinic, he was given intramuscular injection with long-term antipsychotics every third week. Almost every second year, Stenberg had recurrent episodes of acute psychosis requiring psychiatric hospitalization after determination that he was a danger to himself or others, due to medication noncompliance. At the time of the current incident, Stenberg had not shown up at the clinic for his two previous injections, and had therefore been unmedicated at the time of the crime. Clinical staff associated with his AOT order had tried to contact Stenberg on several occasions, and a house visit was scheduled the same week the crime occurred. At the time of his arrest, Stenberg was acting in a disorganized manner. He was observed to be talking to himself and he talked about undercover agents. He was immediately taken to the psychiatric emergency unit. At the emergency unit, he appeared disheveled, laughed at inappropriate times and said he believed the room was bugged. He was treated with medication over objection, and deemed to meet criteria for involuntary admission. After arrangements, he was later admitted to the psychiatric jail unit. On the forensic psychiatric unit, Stenberg continued to act in a psychotic manner: he talked to himself and was very suspicious toward the evaluation team and staff. At one instance, he required an injection of antipsychotic medication. Following a few days on antipsychotic medication, the severity of Stenberg's symptoms lessened and he was able to collaborate with the evaluators. Stenberg confessed to the crime. He stated that he believed Mrs. K was an individual sent out by the psychiatric services with the purpose of hospitalizing him. This had made him worried, and he stated that he had acted in self-defense. Stenberg also stated that, during certain periods when he feels paranoid, he usually carries a knife.

condition c
The court ordered a forensic psychiatric evaluation, completed by court-ordered evaluators. It suggested that Stenberg has moderate ID. Stenberg had been going to special needs school throughout his schooling. He pertained the circle of people 1 § in the "law for support and service to handicapped" (LSS in Swedish) and never had independent housing. He never held normal employment, however was involved in daily work in the communal regime. Psychological evaluations demonstrated that his intellectual level was comparable to an individual aged 7 to 9 years. At the time of the incident, Stenberg lived in a group home with staff available 24/7. Staff at the group home reported that Stenberg was able to perform some tasks on his own (such as toilet visits, get dressed, and go to familiar places in town) but that he required supervision for many daily tasks and reminders about personal hygiene (e.g., brushing his teeth, washing his clothes, taking showers, and using the oven when cooking). The day of the incident, he had run away from the group home and staff was out searching for him. when Stenberg underwent the forensic psychiatric evaluation, he admitted to the crime however it as unclear whether he understood the meaning of the crime. It also appeared that he had set the group home on fire on a previous occasion. This was detected early however, and did not lead to any legal consequences. At the forensic psychiatric unit, the staff at several occasions had to explain to Stenberg why he was there, and also they helped him cleaning his room and washing up.