Jurors’ intuitive conceptions of insanity and their relationship to

verdict selection in insanity defense cases

by

Jennifer L. Skeem, Ph.D. Stephen L. Golding, Ph.D.

Postdoctoral Fellow, Law & Psychiatry Professor of Psychology

Western Psychiatric Institute and Clinic Adjunct Professor of Psychiatry & Law

University of Pittsburgh University of Utah

Contact

Jennifer L. Skeem, Ph.D.

Western Psychiatric Institute and Clinics, Suite 430

3811 O’Hara Street

Pittsburgh, PA 15213

TEL: (412) 624-1822

FAX: (412) 624-2360

E-mail: skeemjl@msx.upmc.edu

Cite as:

Skeem, J., & Golding, S. (1999, October). Jurors’ intuitive conceptions of insanity and their relationship to verdict selection in insanity defense cases. Paper and poster presented at the meeting of the American Psychological Association and the Criminal Justice Section of the American Bar Association, Washington, DC. (Available from the American Psychological Association, 750 First Street, NE, Washington, DC 20002.)

Authors’ note: This paper is part of a dissertation completed by Jennifer L. Skeem. The research was supported by a Graduate Research Fellowship from the University of Utah, a Dissertation Award from the American Academy of Forensic Psychology, and a Grant-in-Aid from the American Psychology-Law Association.

ABSTRACT

The legal system implicitly assumes that jurors are blank slates who apply the law and select verdicts in an evidence-driven fashion. However, research indicates that jurors have experience-based knowledge structures that represent their conceptions of the law and guide their legal decision-making. Substantial evidence suggests that jurors rely upon their own conceptions or prototypes of insanity in determining whether a defendant is insane. However, this study is the first to systematically investigate the nature, variability, and effect of insanity prototypes. First, prototype features were elicited from 80 former prospective jurors. Then, 5 jurors reviewed these features and combined those that they deemed highly similar. Their judgments were used to develop a measure of individual differences in insanity prototypes. Finally, 135 jurors completed this individual differences measure and measures of case-relevant attitudes, and rendered a verdict based on an insanity case vignette.

This research produced three primary findings. First, jurors’ prototypes of insanity are complex and differ from commonly accepted legal standards. Second, despite marked individual differences, there are three identifiable groups of jurors with prototypes that emphasize (1) severe mental disability, (2) a lay conception of “moral insanity,” and (3) characteristics more exclusively related to traditional mental state at the time of the offense concepts. Third, these different prototypes affect the way in which jurors interpret cases and render verdicts. This paper analyzes the implications of these findings for jury selection and the presentation of expert evidence to juries.

INTRODUCTION

As evidenced by the voluminous literature on jury selection techniques (e.g., Abbot, Hall & Linville, 1993; Bennett & Hirschhorn, 1993; Frederick, 1995; Starr & McCormick, 1993), legal professionals have long “known” that jurors’ attitudes and preconceptions influence their verdicts (see Gross, 1986). Nevertheless, the legal system implicitly assumes that jurors are blank slates who are capable of applying the law in an evidence-driven fashion (Finkel, 1995; Smith, 1991). This assumption is only occasionally acknowledged and rarely scrutinized. For example, in Lockhart v. McCree (1986), the Supreme Court held that removing venirepersons with strong objections to the death penalty did not violate a defendant’s constitutional rights by producing a more conviction-prone jury. Speaking for the majority, Justice Rhenquist reasoned that:

Prospective jurors come from many different backgrounds, and have many different attitudes and predispositions. But the Constitution presupposes that a jury selected from a fair cross-section of the community is impartial, regardless of the mix of individual viewpoints actually represented on the jury, so long as the jurors can conscientiously and properly carry out their sworn duty to apply the law to the facts of the particular case.

In essence, the system presumes that jurors’ predispositions will be “sorted out” or neutralized through routine jury selection, instruction, and oath procedures (see Dressler, 1991; Shuman, 1994; Tanford, 1992). This presumption is arguably in error. Research repeatedly demonstrates that attorneys often fail to identify jurors who are biased against their side (Hastie, 1991; MacCoun, 1989; cf Johnson & Haney, 1994), in part because they rely upon “stereotypic rules of thumb” and nonspecific demographic and attitudinal variables to identify bias (Abbot, et al., 1993; Hastie, Penrod & Pennington, 1983; Kassin & Wrightsman, 1994; MacCoun, 1989; Shuman, 1994). Moreover, jurors’ attitudes and preconceptions have been shown to affect their verdicts even when they instructed to lay them aside and swear that they will do so (English & Sales, 1997; Smith, 1993).

Thus, despite legal assumptions and safeguards, “jurors do not so much find reality as construct it” (Finkel, 1995, p. 63). Jurors’ perception of case information and verdicts are guided by their personal theories and experiences. Several investigators have found that mock jurors’ attitudes about case-relevant subjects, their knowledge about everyday events, and their prototypes of offenses and stereotypes of offenders substantially influence their case judgments (Bennett & Feldman, 1981; Moran, Cutler & DeLisa, 1994; Pennington & Hastie, 1986; Roberts & Golding, 1991; Roberts, Golding & Fincham, 1987; Roberts, Sargent & Chan, 1993; Smith, 1991, 1993; Stalans, 1993).

In cases where defendants raise the controversial defense of insanity, the influence of these “knowledge structures” (Schneider, 1991) may particularly jeopardize defendants’ constitutional rights. First, research consistently reveals widespread negative attitudes toward the insanity defense (Hans, 1986; Hans & Slater, 1984; Jeffrey & Pasewark, 1983; Pasewark & Seidenzahl, 1979), and indicates that these attitudes strongly influence mock jurors’ verdicts in insanity defense cases (Bailis, Darley, Waxman & Robinson, 1995; Cutler, Moran & Narby, 1992; Ellsworth, Bukaty, Cowan & Thompson, 1984; Homant & Kennedy, 1987; Roberts & Golding, 1991; Roberts et al., 1987). Second, empirical studies repeatedly indicate that mock jurors fail to apply judicial instruction on legal definitions of insanity in rendering verdicts (Finkel, 1989, 1991; Finkel, Shaw, Bercaw & Kock, 1985; Ogloff, 1991; Ogloff, Schweighofer, Turnbull & Whittemore, 1992; Simon, 1967). The fact that jurors make these determinations without guidance from the law suggests that they rely upon their own conceptions of insanity (Finkel, 1995). Third, jurors draw different inferences about defendants’ cognitive and volitional impairments based on identical case descriptions, and these inferences strongly predict their verdicts (Bailis et al., 1995; Roberts & Golding, 1991; Roberts et al., 1987; Roberts et al., 1993; Simon, 1967). These individual differences in interpreting case evidence may be based on differences in intuitive theories of insanity.

If a juror’s preconceptions preclude genuinely entertaining a defendant’s legal plea of insanity, the defendant will not obtain a fair trial (Cutler et al., 1992). Thus, this series of studies initiated a program of research designed to advance understanding of how jurors determine whether or not a defendant is insane, and to provide methods for identifying and reducing juror bias. The primary purpose of the research was to determine the nature and variability of jurors’ conceptions of insanity and begin to explore their effect on verdicts.

Lay Conceptions of Insanity

Important, but Poorly Defined

As noted above, abundant evidence indicates that jurors fail to apply legal definitions of insanity in rendering verdicts and interpret the meaning of identical case evidence differently. Based on this evidence, several scholars have argued that jurors rely on their own conceptions of insanity to construe case information and determine whether a defendant is insane (Bailis, et al., 1995; Finkel & Handel, 1989; Roberts, et al., 1987). However, this hypothesis has not been directly tested.[1] The closest approximation to work of this nature was conducted by Finkel and Groscup (1997), who explored undergraduate students’ insanity case stories.[2] The authors asked subjects to “use their imagination” to create typical and atypical case narratives about defendants who successfully or unsuccessfully plead insanity in a jury trial. They found that subjects often described a young, male defendant with a psychiatric history who committed a crime against an unknown male based on either grandiose delusions (successful plea) or revenge (unsuccessful plea). Notably, there were no differences between subjects’ “typical” and “atypical” stories. Both story types tended to be extraordinary and to reflect “obvious news media, TV, and Hollywood influences” (p. 216).

These results may well be attributable to the study’s methodology. Subjects were asked to “use their imagination” to create insanity case stories. In doing so, they were encouraged to access and embellish recent media stories to which they had been exposed. In brief, subjects were not asked to access, and arguably did not describe, their personal conceptions of insanity (see Skeem, 1999 for a detailed review).

Conceptions of Insanity as Prototypes

Thus, little is known about the nature of jurors’ conceptions of insanity and still less is known about the process by which they affect verdicts. This study applies an established theory of cognitive categorization, prototype theory, to address these issues (Rosch, 1975; Rosch & Mervis, 1975; see also Hampton, 1993, MacLaury, 1991; Schneider, 1991; Smith & Medin, 1981). According to prototype theory, knowledge about any category is structured around, and represented in memory by, a prototype or quintessential member. Category membership is determined based on a similarity matching process: as the number of characteristics that an object shares with the category’s prototype increases, so does the likelihood that the object will be deemed a member of the category. Because some objects in the category share more critical features with the prototype and are thus more “typical” of the category than others, the category is internally graded rather than uniform and discrete.

Vicki Smith (1991, 1993) has demonstrated that people represent crime categories such as “burglary” with prototypes. These crime prototypes differed substantially from legal definitions, and were more determinative of jurors’ verdicts than judicial instructions on these definitions. Smith’s findings have important implications for jurors’ decision making in insanity defense cases. Just as jurors represent verdict categories in general cases with crime prototypes, so too may they represent verdict categories in insanity defense cases with prototypes of insanity. Determining whether a defendant is insane is a task that is well suited for the process of prototype matching. A juror may make attributions about a defendant’s cognitive and volitional impairments by comparing the defendant’s characteristics to those of her prototype of the insane defendant. The more closely the defendant’s attributes matched those of her prototype, the more likely she would be to judge the defendant a member of the category “insane.” Thus, based on the plausible role of prototypes in representing lay conceptions of insanity (see Skeem, 1999 for a full review), this research applied prototype theory and methodology to determine how jurors construct the meaning of insanity.

Focus on Individual Differences

Traditional prototype research focuses on a “consensus prototype,” or list of characteristics generated most frequently across subjects. In this study, however, individual differences in prototypes of insanity were specifically assessed[3] because “categorization results from a learning process” (Taylor, 1981, p. 87) and individuals with different life experiences have been shown to develop different prototypes that affect their categorization decisions (e.g., Kempton, 1981; MacLaury, 1991). These individual differences are particularly critical, given the jury selection context. As observed by Finkel (1997), aggregate findings do not aid in differentiating prospective jurors with prejudicial prototypes from those with relatively sophisticated prototypes that are in better accord with the law.

Summary and Purpose

Despite legal assumptions and safeguards, jurors rely upon their knowledge and experiences to actively interpret case information and render judgments. Given the controversial nature of the insanity defense, this reliance may particularly jeopardize defendants’ rights in insanity defense cases. Abundant evidence suggests that jurors rely upon their own conceptions of insanity to determine whether a defendant is insane. However, the nature of these conceptions and the process by which they affect verdicts is poorly defined. This research applies prototype theory and methodology to determine the nature and variability of jurors’ prototypes of insanity and explore their relationship to case judgments. This paper outlines these studies for attorneys and expert witnesses, emphasizing their implications for professional practice and legal reform. [1] The research is available in complete form from the primary author (Skeem, 1999).

METHOD, RESULTS AND INTRODUCTORY DISCUSSION

Overview of Design

The research design was comprised of three studies that adapt traditional prototype methodology to measure individual differences. In the first study, the characteristics or features of jurors’ conceptions of insanity were elicited. In the second study, this large group of features was condensed and developed into a measure of individual differences in insanity prototypes. In the third study, this measure was administered to a sample of jurors along with an insanity case vignette to identify individual differences in prototypes and assess their relationship to case judgments.

Participants in all three studies (“jurors”) were former venirepersons, or members of the jury pool, for criminal and civil cases randomly selected from the Third District Court of Utah (which represents Salt Lake County and accounts for 40% of the state’s population). Participants were usually invited to take part in a study by telephone, then mailed the study materials. Across studies, approximately 33% of the venirepersons approached agreed to take part in the study and returned the study materials. However, there were virtually no differences between study participants and study nonparticipants[4] with respect to basic demographic characteristics (e.g., gender, age, race, education), and whether or not individuals were selected for jury service (see Skeem, 1999 for statistics). This section outlines the method and results of each of these studies, focusing on Study 3.

Study 1: Elicitation of Prototype Features

Method

In accord with prototype tradition, prototypes were operationally defined as lists of characteristics or features that subjects generate when asked to describe the “typical” member of a category (Cantor, Mischel & Schwartz, 1982; Reed, 1972; Smith & Medin, 1981). Prototype features were elicited based on a measure adapted from Cantor et al. (1982). The instrument asks participants to spend a few minutes “forming a mental image or representation of the typical person who is not responsible for his criminal actions due to mental illness,” then to candidly describe the characteristics that are common to their conception.

Participants in this study were 81 former venirepersons who were typically White (99%), middle-aged (M=45 years), college-educated adults (M=15 years) who did not serve as jurors on the index case (21% served). Approximately half (54%) were female.

Results: Multifaceted, Irreducible Conceptions of Insanity

To determine the general nature and quality of participants’ whole responses, the data from Study1 were inductively analyzed to identify recurrent patterns based primarily upon the methodology of Miles and Huberman (1994). Participants’ responses were generally thoughtful narratives or feature lists that were absent of the bizarre, dramatic features evoked by the context of Finkel & Groscup’s (1997) study.[5] More critically, the inductive analysis revealed that participants’ prototypes were multifaceted and could not be reduced to legal formulations of insanity or even to single, abstract themes related to the prototype’s mental state at the time of the offense (e.g., themes such as “out of touch with reality;” a delusional or incomprehensible crime). The modal response reflected a combination of two-to-three mental state themes that were enriched by a unique description of the prototype’s human characteristics.

Because lay sources of knowledge are markedly different from those of professionals, the fact that jurors’ representations could not be reduced to psycholegal constructs was not surprising. In essence, these jurors were unimpressed by historically important distinctions among the type and extent of mental state impairment required by particular legal definitions of insanity. In fact, they often combined mental state themes (e.g., “wild beast” and “M’Naughten) that have fueled bitter debates over insanity defense standards.

Study 2: Reduction of Prototype Features

Method

In Study 2, the whole prototype descriptions obtained in Study 1 were condensed and developed into a set of key prototype features to use in a measure of individual differences. There were three primary steps in this process, as depicted in Figure 1.

[Insert Figure 1]

First, experimenters segmented the whole responses into 498 prototype features. Second, this large feature set was reduced by combining highly similar features. Either experimenters or jurors could have been assigned to do so. To preserve the lay meaning of the features, jurors were assigned to decide which features meant the same thing. However, to reduce the load on these jurors, experimenters first placed the features in a manageable order. In specific, experimenters wrote the 498 features on cards, then sorted them into 10 broad, organizational categories (e.g., symptomatology, personality characteristics, mental state at the time of the offense). Next, each of 5 former venirepersons were presented with the features, one organizational category at a time. They were asked to use their judgment to sort the features into as many piles necessary of characteristics that meant essentially the same thing to them. When they were finished sorting the features within each organizational category, they were asked to review their sorts and combine highly similar features across categories.

When at least a majority of the jurors (3 out of 5) agreed that a pair of features meant the same thing, the features were combined. This agreement criterion was conservative, given the diversity of the juror sample[6] and its likely association with different systems of meaning. This process yielded approximately 57 features (or feature groups) that jurors agreed meant the same thing. Third, and finally, experimenters assigned labels to these features to capture their essence, as depicted in Figure 2. These 57 labels became the items of the measure of individual differences in conceptions of insanity.

[Insert Figure 2]

Results: Individual Differences in Conceptions

The number of participants who contribute to each feature may be considered an estimate of “the degree of consensuality in prototype descriptions” (Cantor et al., 1982, p. 50). In this study, approximately one-third (34%) of the features generated by jurors were classified as idiosyncratic. Of the remaining features, few (8%) were listed by more than 10 jurors. This distribution is depicted in Figure 3, and indicates that there are marked individual differences in conceptions of insanity. In fact, there may be as many prototypes of insanity as there are individuals. This is a critical issue to consider when addressing the results of the next study, in which three groups of people with different prototypes of insanity are identified. The three prototypes that will be presented are less representations of the entire group’s prototype than they are prototypes of “idealized individuals” who fall at the group’s centroid and best represent the key prototype dimensions for the group (see Tucker & Messick, 1964).

[Insert Figure 3]

Of course, the data depicted in Figure 3 indicate that the “consensus prototype” approach and aggregate-based research miss the mark. In order to produce meaningful results, individual differences must be assessed. This was the focus of the final study.

Study 3: Individual Prototype Differences and Case Judgments

Method

In the third study, the measure developed in Study 2 was used to investigate the nature and variability of prototypes of insanity and explore their relationship to case judgments. Participants in this study were 133 former venirepersons who were typically White (95%), middle-aged (M=43 years), college-educated adults (M=15 years) who did not serve as jurors on the index case (22% served). Approximately half (53%) were female.

The feature labels developed in Study 2 became the items of the measure of individual differences in conceptions of insanity, the Conception Checklist (CC). The CC asks participants to (a) spend a few minutes forming a clear mental representation of the prototypical person who is not responsible for his criminal actions due to mental illness, then (b) rate the extent to which each of the 57 features is important in defining their prototype, using anchored Likert scales that range from “irrelevant” to “essential.” Study participants also rendered case judgments based on a brief written vignette that depicted a typical insanity defense case (adapted from Roberts et al., 1987), and completed measures of insanity case-relevant attitudes.

Results

Identifying Individual Differences in Conceptions of Insanity

To identify groups of individuals with different prototypes, jurors’ scores on the CC were cluster analyzed. Cluster analyses include a broad range of specific techniques for empirically organizing cases to form relatively homogeneous “clusters” or groups of highly similar cases. Given the context of this paper, these analyses are merely outlined here. A detailed account of the analyses are available elsewhere (Skeem, 1999). Because different clustering methods can produce different solutions, multiple methods that perform well in recovering known group structure were applied (Ward’s, average linkage, and k-means), and various solutions compared. The solution that was the most interpretable and stable across methods and sub-samples of the data was retained. This solution was a 3-group solution obtained by using Ward’s method to define a starting point for the k-means pass method (see Blashfield & Aldenderfer, 1988; Borgen & Barnett, 1987; Lorr, 1994). Because cluster analyses impose structure, it is important to determine whether the groups derived represent “natural” groups. In this study, the cluster solution was validated on external attitudinal and case judgment variables that were conceptually related to the variables used in the cluster solution (e.g., see “Exploring the Impact…” below).

Describing Individual Differences in Conceptions of Insanity

The nature of the groups identified using cluster analyses are typically presented by investigating the mean scores of the groups across the variables that were analyzed. Because it is difficult to describe the group’s similarities and differences across 57 variables, a principal components analysis was completed on the CC to condense these interrelated variables into four major underlying components (again, details are available in Skeem, 1999). After describing these four components, the three groups’ scores on the components will be presented.

[Insert Table 1]

Component descriptions. The orthogonally rotated component loading matrix is presented in Table 1. To simplify interpretation, the variables are grouped by loading size and loadings under .37 are deleted. The first component was named “Moral Insanity.” The feature labels that loaded strongly on this component reflect characteristics of psychopathy (e.g., unfeeling and manipulative; has no conscience; accepts no responsibility), violent, unpredictable behavior (e.g., violent, angry and hostile; prone to hurting self or others; extreme, erratic), and psychosis (e.g., distorted vision of reality; irrational). The second component was named “Severe, Chronic Disability.” Feature labels that loaded strongly on this component reflect severe functional impairment (e.g., totally debilitated, unable to function in society), mental disability (e.g., mentally retarded; mentally no more than a child; head injured), and psychiatric treatment (e.g., takes medication; has a long, documented history of mental illness [that] cannot be controlled with treatment). The third component was named “Poorly Functioning Social Outcast.” Feature labels that loaded strongly on this component reflect characteristics of marginal functioning (e.g., can’t hold a job; disheveled), social isolation (e.g., withdrawn; detached), and mental illness (e.g., suffers mood swings or chemical imbalance). As revealed by the complex loadings for this component, this was the least “clean” of the four components. The fourth component was named “Mental State-Related Characteristics.” Feature labels that loaded strongly on this component reference the nature and intensity of impaired mental state at the time of the offense (e.g., incapable of choosing between right and wrong; cannot control his thoughts, emotions, actions; is not aware of what he is doing), or expert evidence thereof (e.g., determined by a credible expert to meet the criteria for insanity).

Prototype descriptions. Table 2 begins to describe the nature of the three prototype groups by depicting their average scores across the four components of the CC. The first group (47% of the sample) specifically obtained high scores on the Severe, Chronic Disability component, the second group (32% of the sample) scored highly on the Moral Insanity component, and the third group (21% of the sample) scored highly on the Mental State-Related Characteristics component. As the table indicates, the three prototype groups were labeled “Severe Mental Disability,” Moral Insanity,” and “Mental State-Centered,” respectively.

[Insert Table 2]

Table 2 provides a general description of the nature of each group. To provide a more refined description, ANOVAs were performed on each of the features that at least one group rated as important[7] to their prototype to identify critical similarities and differences among the three groups.[8] Table 3 presents the characteristics that are important to all three prototypes, or overlapping features across groups.[9] The table indicates that symptoms of psychosis and

[Insert Table 3]

impaired mental state at the time of the offense are important to all three prototype groups. In fact, 70% of the features that the groups shared loaded on the Mental State-Related Characteristics component in the principal components analysis. Although the Mental State-Centered prototype group consistently rated these characteristics more strongly than the other groups, all three groups rated them highly. Tables 4, 5, and 6 describe the discriminating characteristics of the Severe Mental Disability, Moral Insanity, and Mental State-Centered prototype groups, respectively. The tables lists features that each group rated as significantly more essential to their prototype than the other groups, in descending order of their importance to the index group. The data corroborate the results depicted in Table 2, and are discussed further below.

[Insert Tables 4, 5 and 6]

Differences in the characteristics of prototype groups. The prototype groups differed from one another with respect to two demographic characteristics. First, there was a significant and meaningful difference across groups with respect to gender (c² [2, N=133] = 9.54, p=.008). Although the SMD group was approximately equally represented by both genders (44% men), the MSC group was composed of more men (71% men), and the MI group was composed of more women (35% men). Second, members of the MSC group had significantly more education than members of the other two groups (F [2, 127] = 4.63, p=.01). Although the MSC group had an average of 17 years of education (SD=3), the SMD and MI groups had averages of 15 years (SD=2, SD=3, respectively).

Exploring the Impact of Prototypes on Case Judgments

Although this study is exploratory, the evidence reviewed in the introduction suggests that differences in individuals’ prototypes of insanity should be related to differences in the way that they construe and judge cases.[10] Thus, to determine the relationship of prototypes to case judgments, jurors read a brief written vignette that depicted a typical insanity defense case (described below). After reading the vignette, they were asked (a) to rate on 7-point Likert scales their construal of the defendant with respect to seven dimensions (e.g., the extent to which he appreciated that his actions were wrong, was capable of rational behavior, was capable of controlling his actions), and (b) to indicate on a 100-point scale the likelihood that they would find the defendant insane and issue a categorical verdict (“guilty” or “not guilty by reason of insanity,” NGRI).

The three prototype groups were compared using MANOVA to determine whether their judgments on the insanity case vignette differed. The groups differed significantly and meaningfully (F [16, 246] = 2.08, p = .009; h²=.23) on the combined construal items and scaled verdict. Post hoc tests using Tukey’s HSD (see Table 7) indicated that the Mental State-Centered (MSC) group differed from the Severe Mental Disability (SMD) and Moral Insanity (MI) groups with respect to 3 of the 7 case construal items. As depicted in Figure 4, the MSC group perceived the defendant as more mentally disordered (MSC M =5.6, SD=1; SMD M=4.6, SD=1; MI M=4.6, SD=2), less capable of controlling his beliefs (MSC M =3.0, SD=1; SMD M=4.1, SD=2; MI M=4.0, SD=2), and less worthy of punishment (MSC M =3.9, SD=1; SMD M=5.0, SD=1; MI M=5.1, SD=2) than the other groups.

[Insert Table 7 and Figure 4]

More importantly, as depicted in Figure 5, the MSC group was more likely to deem the defendant not guilty by reason of insanity than both the SMD and MI groups, although the latter difference did not attain statistical significance (MSC M =57.5, SD=30; SMD M=40.4, SD=28; MI M=42.3, SD=31). As depicted in Figure 6, the groups also differed on the less sensitive but more realistic categorical measure of verdicts (MSC 64% “NGRI,” SMD 30%, MI 35%), but the differences did not reach statistical significance (c² [2, N=133] = 4.46, p=.12).

[Insert Figure 5]

It is important to recall these analyses involve a single vignette that could not be designed to target and test the effect of the three insanity prototype groups. Nevertheless, as explained below, the results of these analyses make sense.

General Discussion

The primary findings of this research may be organized into three points. First, jurors have complex, multifaceted prototypes of insanity that cannot be reduced to legal tests or psychiatric constructs (see Study 1). Second, these prototypes differ markedly across individuals (see Study 2). However, there are three discernible subgroups of jurors, represented by “idealized individuals” who have prototypes that emphasize extreme mental disability, a lay conception of “moral insanity,” and characteristics that are more specifically related to impaired mental state at the time of the offense. Third, these prototypes affect the way in which jurors perceive case information and render verdicts. Thus, jurors bring to insanity defense cases not “blank slates,” but knowledge structures that drive their case judgments. After presenting the general nature and background of each of the prototypes associated with these juror subgroups, this section will more specifically analyze their distinguishing and shared characteristics and their implications for professional practice and legal reform.

Nature and Background of Three Prototypes of Insanity

Given the extent of individual differences in prototypes of insanity (see Figure 3), the three prototypes analyzed below are less representations of the entire group’s prototype than they are prototypes of “idealized individuals” who fall at the group’s centroid and best represent the key prototype dimensions for the group (see Tucker & Messick, 1964). Despite substantial differences in their essential characteristics, these prototypes overlap in their emphasis on psychosis and impaired mental state (see Table 3). However, each of the prototypes resonates with both current research and historical conceptions of insanity.

Prototype of the Majority: Severe Mental Disability

Nearly half (47%) of the jurors had prototypes of insanity that emphasized “Severe Mental Disability” (SMD). As depicted in Tables 2 and 4, the SMD prototype is afflicted with extreme, chronic, uncontrollable mental illness and mental retardation that impairs his[11] ability to function in society. This conception is consistent with prior research which suggest that, in the aggregate, jurors’ conceptions of insanity “remain wed to the wild beast test of 1724” (Perlin, 1997, p. 1379; see Hans & Slater, 1984; Roberts et al., 1987). This is a stringent, narrow test which requires that an insane defendant be “totally deprived of his understanding and memory, and doth not know what he is doing, no more than an infant, a brute, or a wild beast” (Arnold’s Case, 1724, as cited in Walker, 1968, p. 56). However, this research is the first to describe the contours of the lay SMD or “wild beast” prototype. Although the prototype heavily emphasizes features of “furious insanity” (e.g., totally debilitated), it also includes several less extreme mental state features (e.g., inability to discern right from wrong). Thus, as indicated in Study 1, lay prototypes cannot be reduced to legal tests of insanity.

A notable feature of the SMD is his combination of mental disability with psychosis. This combination, which may seem relatively unlikely to mental health professionals, is consistent with prior research. For example, Furnham and Rees (1988) found that a subset of laypeople meld together psychosis and mental retardation in a way that is uniquely relevant to criminal responsibility. This subset who believed that schizophrenia was based on poor intelligence tended to believe that these people were “amoral,” and incapable of “knowing any better” (p. 279). In fact, a defining characteristic of the SMD prototype is that he is “mentally no more than a child.” Thus, the SMD defendant is mentally impaired, psychotic, and deprived of reason to the extent that he is animalistic or childlike, and therefore unworthy of blame. As described by Foucalt (1965), “when the madman has become a beast…man himself is abolished” (p. 72).

The SMD’s extreme combination of characteristics is identifiable not only in research, but also in contemporary media and historical conceptions of insanity. The media, which is laypeople’s primary source of information about mental illness and the insanity defense (see Hans, 1990; Finkel, 1995; Perlin, 1994; Wahl, 1995), often portrays the mentally ill as “bestial.” For example, a recent headline for a New York Post article on the release of a patient from a psychiatric patient read “Village Beast May Go Free” (Wahl, 1995, chap. 3). Of course, this prototype existed long before the media could serve as its vehicle. The image was pervasive throughout the 17^th and 18^th centuries, when the mentally ill were treated as unreasoning, inhuman beings. Although Fink proclaimed that the 19^th century “formed a period which saw the chains struck from the insane and their elevation from beasts to the stature of men” (1938, p. ) the wild beast conception of the severely mentally ill survives today. Although several experts have suggested that all laypersons subscribe to the SMD or wild beast prototype of insanity, this suggestion is based primarily on aggregate-based research (Perlin, 1997; see also Ellsworth et al., 1984; Hans & Slater, 1984; Roberts et al., 1987). This research indicates that, although a large segment of laypersons have prototypes that emphasize SMD characteristics, approximately half do not.

A Lay Conception of Moral Insanity

Approximately one-third (32%) of the jurors had prototypes of insanity that emphasized a lay interpretation of “Moral Insanity” (MI). Many (65%) of these jurors were women. As depicted in Tables 2 and 5, in the MI prototype, characteristics of psychopathy and psychosis are conflated to represent a malevolent, detached, irrational and unpredictably violent offender.

This prototype reflects the evolution of the term “moral insanity” (see Cleckley, 1976; Fink, 1938, chap. 3; Maughs, 1941a, 1941b). In the early 19^th century, moral insanity was defined as a selectively diseased state of will in an otherwise intact individual (see Fink, 1938). Some scholars in medical jurisprudence associated the term with an “irresistible impulse” to commit an offense and argued that moral insanity absolved the offender of culpability (Belkin, 1996). However, critics charged that the will couldn’t be selectively impaired—that moral insanity was nothing more than moral depravity or wickedness (Fullwinder, 1975). In a high profile trial in 1881, moral insanity was rejected as a psychiatric category and legal defense (see Fink, 1938). Before its demise, however, the term moral insanity was applied to almost every form of psychopathology, including dementia praecox and other forms of psychosis. In the 20th century, moral insanity became understood as “moral depravity” and was gradually replaced by the constructs of psychopathy and antisocial personality disorder. The lay prototype of moral insanity reflects this evolution in its blending of psychopathy, psychosis and impaired mental state. Like the early proponents of moral insanity, laypeople with MI prototypes may believe that these people have inherited or acquired moral defects that render them less responsible for their actions.

However, the lay MI prototype is also psychotic: that is, impaired in both will and reason. This blend of psychopathy and psychosis is striking in its resemblance to the most pervasive media stereotype of the mentally ill: the psychotic psychopath (see Day & Page, 1986; Hyler, Gabbard & Schneider, 1991; Matas, el-Guebaly, Harper, Green & Peterkin, 1986; Shain & Phillips, 1991; Signorelli, 1989; Wahl & Roth, 1982). The popularity of this stereotype persists, despite agreement among psycholegal professionals “that the blending of the manipulative, antisocial, but clear-thinking aspects of psychopathic individuals with the disabling pathology of the person with psychosis represents a decidedly unlikely combination (Wahl, 1995, pp. 19-20). These media portrayals are shaped by historical and economic forces. In fact, the history of the term “moral insanity” suggests that versions of the construct existed long ago. Before the early Renaissance, European scholars typically associated psychopathology with demonic possession or with punishment for sin (Carson, Butcher & Mineka, 1998).

More Legally Desirable? Mental-State Centered

Less than one-quarter (21%) of jurors had Mental-State Centered (MSC) prototypes that specifically emphasize characteristics related to impaired mental state at the time of the offense. Most (71%) of these jurors were male and highly educated. Although the MSC prototype shares most of its characteristics with the MI and SMD prototypes, it emphasizes these characteristics more strongly and much more exclusively. As indicated in Tables 2 and 6, the MSC prototype is afflicted almost solely with impairments in his mental state at the time of the offense and these impairments are clearly supported by expert evidence.

The finding that a minority of jurors have tightly focused, legally relevant prototypes of insanity is consistent with the results of several attitudinal studies. These studies indicate that only a minority of laypeople endorse the moral logic and implications of the insanity defense. For example, just over one-third of laypeople believe that it is fundamentally wrong to punish the insane (Hans, 1986). These people who strongly believe in the insanity defense may have MSC prototypes. Two points support this hypothesis. First, the MSC’s characteristics are linked by the premise or opinion that mental illness is associated with reduced capacity for forming intent, making rational decisions and controlling one’s behavior. Second, there was a trend in this study for jurors with MSC prototypes to endorse the fundamental logic of the insanity defense on a measure of insanity defense attitudes more strongly than jurors belonging to the other two prototype groups (see Skeem, 1999).

The MSC prototype has roots that extent “to the earliest recordings of Hebrew law” (Golding & Roesch, 1987, p. 397). It is built upon a broad historical interpretation of mens rea, or fundamental belief that an “act is not legally cognizable as evil, and hence criminally punishable, unless it is committed by a person who has the capacity to cognize the act as evil and then freely chooses to do it” (Morisette v. United States, 1952, as cited in Golding & Roesch, 1987, p. 395). Thus, the MSC includes diverse forms of mental state impairments that span a range of themes associated with legal tests of insanity.

Relationship Between Prototypes and Case Judgments

This research was designed to provide a foundation for testing the extent to which jurors’ prototypes of insanity are related to their case judgments. Future research will do so based on case materials that specifically target the three prototypes identified in this study. Nevertheless, the relationships among prototypes and case judgments were explored in this study by presenting jurors with a single insanity case vignette that portrayed a typical insanity defense case. The results indicate that jurors’ prototypes of insanity are significantly related to the way in which they interpret case information and render verdicts. Jurors with MSC prototypes perceived the defendant as more mentally disordered, less capable of controlling his beliefs and less worthy of punishment than other jurors. More importantly, jurors with MSC prototypes were more likely to deem the defendant not guilty by reason of insanity than those with SMD or MI prototypes, although the latter difference did not attain statistical significance.

These results make sense in light of prototype differences. The case vignette describes a young defendant who is marginally functional (working as a dishwasher), whose adjustment had been deteriorating over recent years, with the emergence of increasing isolation, unclear speech, and delusions. The vignette is organized around the defendant’s persecutory delusions and their relationship to his violent crime. In focusing on his psychosis and impaired mental state, the vignette does not portray the defendant as extremely mentally disabled nor psychopathic. Thus, it makes sense that the MSC group deemed the defendant more mentally disordered, less able to control his beliefs, and more insane than the other groups. The defendant’s characteristics matched the circumscribed essence of the MSC prototype but did not reflect the additional features nor essence of the other prototypes.

Implications for Practice

Are Most Prototypes Conviction Prone?

This research suggests that the “prototypic” defendant in an insanity defense case would typically face a jury with conceptions that would render them highly unlikely to deem him insane. The prototypic defendant who is acquitted by reason of insanity is, like the defendant depicted in the study vignette, a psychotic man with a psychiatric history who commits a violent crime (e.g., Circione, Steadman & McGreevy, 1995; Rice & Harris, 1990). This defendant does not typify most jurors’ conceptions of the insane defendant as a “wild beast” or psychotic psychopath: his characteristics represent a conception of insanity focused on legally relevant mental state impairments that only a small minority of jurors share.[12]

As noted previously, several studies suggest that jurors’ predominantly negative attitudes toward the insanity defense are strongly related to their insanity prototypes and case judgments (Skeem, 1999; Roberts & Golding, 1991; Roberts et al., 1987). Although further research is needed,[13] these data suggest that many jurors’ prototypes may be conviction-prone.

As noted by Finkel (1997), a strong proponent of “common sense justice,” the law should attempt to “curb and correct” such prejudicial prototypes to better insure the fairness and accuracy of jury trials. Even if one does not view the SMD and MI prototypes as prejudicial, however, “lay decisions can and should be brought into better conformity with the law, if for no other reason than to bring greater regularity and predictability to decision making” (Penrod & Heuer, 1997, p. 261). In an effort to aid psycholegal professionals in doing so, speculative guidelines for jury selection and presentation of expert evidence are presented below, based on the results of this research.

Jury Selection

One way of bringing jury decisions into greater accord with the law is to exclude clearly biased jurors. One may attempt to do so by using voir dire to specifically examine the knowledge structures that have been shown to predict jurors’ verdicts in insanity defense cases, using validated measures when available.[14] A prospective juror should be carefully questioned about his or her case-specific attitudes toward the insanity defense, the mentally ill, and the criminal justice system (for a review and a validated measure of insanity defense attitudes, see Skeem, 1999; see also Hans, 1986; Roberts & Golding, 1991).[15] As suggested by the research presented here, a prospective juror’s personal conception of the “typical person who is not responsible for his criminal actions due to mental illness” should also be explored. Assuming that one is unlikely to be successful in removing SMD or MI jurors “for cause,” it is neverthelss important to use the voir dire as a means of laying the groudwork for “counter-prototype arguments later in the trial process (see below).

Several key features distinguish among subgroups of jurors whose primary prototype dimensions correspond to the Severe Mental Disability (SMD), Moral Insanity (MI), and Mental State-Centered prototypes. These features may be especially effective in predicting the way in which jurors will interpret case information and render judgments in insanity defense cases. Important distinguishing characteristics of the commonly-held SMD prototype include (a) “extreme mental illness—TOTALLY debilitated,” (b) “mental retardation,” (c) “has a long, documented history of mental illness,” and (d) “mental illness cannot be controlled with therapy or medication” (see Table 4). According to this research, a juror who conceives of the criminally insane defendant as an obviously psychotic person with extremely low intelligence and sustained, but unsuccessful contact with the mental health care system sets a “high hurdle” for issuing a verdict of insanity.

Distinguishing characteristics of the MI prototype include (a) “has no conscience,” (b) “grossly distorted vision of reality,” (c) “violent, angry, and hostile to others,” (d) “exhibits extreme, erratic, unpredictable behavior, and (e) “acts without reason, provocation, or rational motive,” and (f) “does not accept responsibility for his actions—feels that his actions are justified or that others are to blame” (see Table 5). A juror who expects the criminally insane defendant to exhibit strong underlying characteristics of psychopathy and periodically manifest irrational, unpredictable, violent behavior is relatively unlikely to deem the typical defendant insane.[16]

Prospective jurors with MSC prototypes may be somewhat more difficult to identify than those with MI or SMD-like prototypes. The MSC prototype is defined by its organization around, and exclusive focus on, characteristics related to impaired mental state at the time of the offense. As noted previously, the MSC prototype shares most of its characteristics with the other two prototype groups, but emphasizes these features more strongly and specifically. Only two characteristics significantly distinguish the MSC prototype from the other two groups: (a) “determined by a credible expert to clearly meet the criteria for insanity,” and (b) “incapable of making a clear choice between right and wrong” (see Table 6). Thus, prospective jurors with MSC prototypes may be identified by their focus on mental state and evidentiary issues to the exclusion of “wild beast” and “psychotic psychopath” characteristics. In voir dire, one might specifically focus on querying jurors with respect to discriminating characteristics of the SMD and MI prototypes.

Presentation of Expert Evidence

When presenting evidence on a defendant’s mental state at the time of the offense, attorneys and experts may maximize their effectiveness by targeting key features of predominant lay conceptions of insanity that predict verdicts. Although they have long been the subject of heated professional debate, legal tests of insanity make little difference to jurors, who produce similar verdict patterns regardless of (a) the specific legal definition of insanity they are instructed to apply, and (b) whether or not they are provided with any legal definition of insanity (see Finkel, 1995; Ogloff et al., 1992; Rice & Harris, 1990 for reviews). For these reasons, focusing on subtleties of the wording and interpretation of an insanity defense statute during a trial is unlikely to be effective. Because there are apparently no data on normative practice in the use of expert evidence in insanity defense cases, the extent to which attorneys assume this focus is unclear. Nevertheless, rather than emphasizing the ways in which a case matches a specific legal test that is of little practical significance, one might specifically address the extent to which a defendant matches the “real” criteria for insanity in effect. There are two options for doing so, based on the extent to which the facts of a given case fit these criteria.

Case Fits Predominant Prototype Dimensions

This research suggests that one should focus on two groups of characteristics in presenting evidence. First, in keeping with prior aggregate-based research (Ellsworth et al., 1984; Hans & Slater, 1984; Roberts et al., 1987), this research suggests that the prototype of insanity that will be best-represented on a jury is the SMD, or “wild beast” prototype. Thus, in presenting expert evidence, one might focus on the critical characteristics of the SMD presented above. In addition to emphasizing the extremity of a defendant’s psychosis and mental disability, one should specifically address the defendant’s involvement with the mental health care system. Most jurors, particularly those with SMD prototypes believe that defendants who refuse effective medication or therapy are responsible for their impaired mental state at the time of the offense, and are therefore blameworthy for their crimes (Finkel, 1995; Skeem, 1999).

The second important group of characteristics on which to focus in presenting expert evidence is the overlapping characteristics important to all three conceptions of insanity, including (a) “unable to discern right from wrong,” (b) “not capable of understanding the harmful consequences of his actions,” (c) “evaluated by a qualified mental health professional and diagnosed as mentally ill,” (d) “impaired mental state is not the result of drug or alcohol abuse,” (e) “cannot control his thoughts, emotions, or actions,” (f) “is not aware of what he is doing,” and (g) “suffers from schizophrenia or psychosis” (see Table 3 for more). Thus, regardless of their distinguishing features, each conception of insanity emphasizes psychosis and a range of mental state impairments.

According to prototype theory and research, as the number of features that the case shares with a category’s prototype increases, so does the likelihood that the case will be deemed a member of the category (e.g., Hampton, 1993; MacLaury, 1991; Rosch, 1975). Smith (1991) found that this principle applies to legal decision-making: as the number of features that a case shares with a person’s crime prototype increases, so does the likelihood that the defendant will be deemed guilty of the crime, regardless of whether the case meets legal criteria for crimes. Although further research is needed to test the extent to which this is true of prototypes of insanity, it is likely that a jury will be more apt to deem a defendant insane as he is shown to exhibit more of the features that jurors’ conceptions of insanity share as well as characteristics of the predominant SMD prototype.

Case Does Not Fit Predominant Prototype Dimensions

Of course, this is not possible if the case facts fundamentally do not match predominant prototype dimensions. In this case, one might directly target the prototype dimensions in voir dire as well as at trial an attempt to modify them. Smith (1993) found that prototypes of crime categories could be brought into greater accord with legal definitions, but only when they were specifically identified, refuted point-by-point, and revised. Strictly reasoning by analogy, one might attempt a similar strategy in insanity defense cases.

For example, consider a functional defendant who delusionally believes that a disc jockey has been terrorizing him for months by stalking him and broadcasting threatening messages about him, and carefully executes a plan to buy a gun and shoot the disc jockey. This defendant does not match some of the primary overlapping characteristics across prototype groups, and clearly does not resonate with the “wild beast” prototype. In presenting this case, an expert may maximize her effectiveness by concentrating her testimony on (a) “prototype” assumptions that are contradicted by the case facts or expert knowledge, and (b) how the case facts comport with a more sophisticated understanding of statutory criteria. The contours of this presentation will depend on how a given case relates to a specific statute. Targeting a key contrast between the SMD prototype and the above case in a jurisdiction with ALI (1962) criteria, however, an expert may focus on the erroneousness of the assumption that insane defendants are completely debilitated by mental illness and retardation—that they are so impaired that they could not possibly plan and carry out a complex course of action. The expert could present theory, research and forensic data on how delusions influence thinking and behavior, focusing principally on the extent to which the defendant was “free” to choose “saner” alternatives. If the defendant had attempted non-criminal ways of removing himself from danger (e.g., by appealing to law enforcement or mental health agencies for help in defending himself), this point would be emphasized. If the defendant had been treated, but ineffectively, the extent to which he was capable of ignoring the “press” of the delusion would be presented. In addition, the expert would specifically emphasize that, according to statute “X,” it is not necessary for a defendant to be totally disabled by a mental illness. In fact, a person can be insane even if he is capable of planning an offense. The statute requires only that a defendant be unable, because of mental illness, “to appreciate the criminality of his conduct” or “to conform his conduct to the requirements of the law” (p. 74). In short, the expert’s evidence would focus on the discrepancies between juror prototypes and case-specific facts when viewed through the lens of expert knowledge.

Caveat

The extent to which these strategies will be effective is unclear. We are in the process of completing studies that (a) directly target the three prototype groups to determine the extent to which jurors decide cases based on a prototype matching process, and (b) determine the extent to which reliance on prejudicial prototypes can be reduced. Nevertheless, the data presented here are an improvement over “flying blind.” Prior research indicated only that jurors seemed to rely upon their own conceptions of insanity to determine verdicts. This research describes what the nature of those conceptions are.

Implications for Legal Reform

This research also has implications for reform of the voir dire process. The courts are increasingly conducting “streamlined” voir dire examinations that limit attorneys’ participation and may be of inappropriate nature and insufficient scope to reveal juror bias (Johnson & Haney, 1994). Judges’ discretion in selecting the topics to be addressed during voir dire is narrowed only in cases that involve interracial violent crimes, capital punishment, or significant pretrial publicity, in which venirepersons’ racial prejudice, attitudes toward the death penalty, or preconceptions about the defendant’s guilt, respectively, must be addressed (Johnson & Haney, 1994; Sklansky, 1996). Although insanity defense cases invoke equally powerful biases, judges are not required to conduct similar inquiries in these cases (112 A.L.R. 531 1996; 28 A.L.R. Fed 25 § 24a, 1996). In fact, the caselaw reflect a pattern in which judges refuse to inquire about bias against the insanity defense or even allow jurors who express biases to be impaneled (see Perlin, 1994, p. 318). Given evidence that attitudinal and conceptual biases are prevalent and strongly affect jurors’ judgments (see Introduction), the voir dire procedure should be reformed such that bias is routinely addressed in order to better insure the fairness and accuracy of insanity defense trials.

Study Limitations and Implications for Future Research

Although the cluster solution in this study was carefully developed and validated on external criteria, it must be replicated on independent samples to determine (a) whether the three prototypes of insanity identified are replicable within the immediate population, and (b) the extent to which prototypes of insanity differ across jurisdictions or cultures. A study that replicates these prototypes within the immediate population is currently underway.

This research was designed to provide the groundwork for testing the “prototypicality” of jurors’ conceptions of insanity. As indicated previously, future research is needed to determine (a) the extent to which jurors render verdicts based on a prototype matching process, (b) test the competitive effects of prototypes and legal definitions of insanity, and (c) the extent to which reliance on prejudicial prototypes can be reduced. To better insure that “knowledge-driven” verdicts are not based partially on the presentation of incomplete evidence that precludes data-driven verdicts, future research should use more realistic trial materials than the written case vignette provided in this study.

Endnotes

REFERENCES

112 A.L.R. 531 (1996).

28 A.L.R. Fed. 26, § 34a (1996).

Abbott, W., Hall, F., & Linville, E. (1993). Jury research: A review and bibliography. Philadelphia, PA: American Bar Association.

American Law Institute. (1962). Model Penal Code §4.01. Philadelphia: Author.

Bailis, D., Darley, J., Waxman, T., & Robinson, P. (1995). Community standards of criminal liability and the insanity defense. Law and Human Behavior, 19, 425-446.

Belkin, G. (1996). Moral insanity, science and religion in nineteenth-century America: The Gray-Ray debate. History of Psychiatry, 7, 591-613.

Bennett, W., & Feldman, M. (1981). Reconstructing reality in the courtroom: Justice and judgment in American culture. New Brunswick, NJ: Rutger’s University Press.

Bennett, C., & Hirschorn, R. (1993). Bennett’s guide to jury selection and trial dynamics in civil and criminal litigation. St. Paul: West.

Blashfield, R., & Aldenderfer, M. (1988). The methods and problems of cluster analysis. In J. Nesselroade and R. Cattell (Eds.), Handbook of multivariate experimental psychology (pp. 447-473). New York: Plenum Press.

Borgen, F., & Barnett, D. (1987). Applying cluster analysis in counseling psychology research. Journal of Counseling Psychology, 34, 456-468.

Cantor, N., Mischel, W., & Schwartz, J. (1982). A prototype analysis of psychological situations. Cognitive Psychology, 14, 45-77.

Carson, R., Butcher, J., & Mineka, S. (1998). Abnormal psychology and modern life (10^th ed.). New York: Addison Wesley Longman.

Circione, C., Steadman, H., & McGreevy, M. (1995). Rates of insanity acquittals and the factors associated with successful insanity pleas. Bulletin of the American Academy of Psychiatry and Law, 23, 399-409.

Cleckley, H. (1976). The mask of sanity (5th ed.). Saint Louis: Mosby.

Cutler, B., Moran, G., & Narby, D. (1992). Jury selection in insanity defense cases. Journal of research in personality, 26, 165-182.

Day, D., & Page, S. (1986). Portrayal of mental illness in Canadian newspapers. Canadian Journal of Psychiatry, 31, 813-816.

Dressler, J. (1991). Understanding criminal procedure. New York: Matthew Bender & Co.

Ellsworth, P., Bukaty, R., Cowan, C., & Thompson, W. (1984). The death-qualified jury and the defense of insanity. Law and Human Behavior, 8, 81-93.

English, P., & Sales, B. (1997). A ceiling or consistency effect for the comprehension of jury instructions. Psychology, Public Policy, and Law, 3, 381-401.

Fink, A. (1938). Causes of crime. Philadelphia: University of Pennsylvania Press.

Finkel, N. (1989). The Insanity Defense Reform Act of 1984: Much ado about nothing. Behavioral Sciences and the Law, 7, 403-419.

Finkel, N. (1991). The insanity defense: A comparison of verdict schemas. Law and Human Behavior, 15, 533-555.

Finkel, N. (1995). Commonsense justice: Jurors’ notions of the law. Cambridge: Harvard University Press.

Finkel, N. (1997). Commonsense justice, psychology and the law: Prototypes that are common, senseful, and not. Psychology, Public Policy and Law, 3, 461-489.

Finkel, N., & Duff, K. (1989). The insanity defense: Giving jurors a third option. Forensic Reports, 2, 235-263.

Finkel, N., & Groscup, J. (1997). Crime prototypes, objective versus subjective culpability, and a commonsense balance. Law and Human Behavior, 21, 209-230.

Finkel, N., & Handel, S. (1988). Jurors and insanity: Do test instructions instruct? Forensic Reports, 1, 65-79.

Finkel, N., & Handel, S. (1989). How jurors construe “insanity.” Law and Human Behavior, 13, 41-59.

Finkel, N., & Sales, B. (1997). Commonsense justice: Old roots, germinant ground, and new shoots. Psychology, Public Policy, and Law, 3, 227-241.

Finkel, N. Shaw, R., Bercaw, S., & Kock, J. (1985). Insanity defenses: From the jurors’ perspective. Law and Psychology Review, 9, 77-92.

Foucault, M. (1965). Madness and civilization: A History of insanity in an age of reason. New York: Pantheon Books.

Frederick, J. (1995). Mastering voir dire and jury selection: Gaining an edge in questioning and selecting a jury. Chicago: American Bar Association.

Fullwinder, S. (1975). Insanity as the loss of self: The moral insanity controversy revisited. Bulletin of the History of Medicine, 49, 87-101.

Furnham, A., & Rees, J. (1988). Lay theories of schizophrenia. The International Journal of Social Psychiatry, 34, 212-220.

Golding, S., & Roesch, R. (1987). The assessment of criminal responsibility: A historical approach to a current controversy. In I. Weiner & A. Hess (Eds.), Handbook of forensic psychology (pp. 395-436). New York: Wiley.

Gross, S. (1986). Overruled: Jury neutrality in capital cases. Stanford Lawyer, 21,11-56.

Hampton, J. (1993). Prototype models of concept representation. In I. Van Mechelen, J. Hampton, R. Michalski, & P. Theuns (Eds), Categories and concepts: Theoretical views and inductive data analysis (pp. 67-95). London, England: Academic Press

Hans, V. (1986). An analysis of public attitudes toward the insanity defense. Criminology, 24, 383-414.

Hans, V. (1990). Law and the media: An overview and introduction. Law and Human Behavior, 14, 399-407.

Hans, V., & Slater, D. (1984). “Plain crazy:” Lay definitions of legal insanity. International Journal of Law & Psychiatry, 7, 105-114.

Hastie, R. (1991). Is attorney-conducted voir dire an effective procedure for the selection of impartial juries? American University Law Review, 40, 703-726.

Hastie, R., Penrod, S., & Pennington, N. (1983). Inside the jury. Cambridge: Harvard University Press.

Homant, R., & Kennedy, D. (1987). Subjective factors in clinicians’ judgments of insanity: Comparison of a hypothetical case and an actual case. Professional Psychology: Research and Practice, 5, 439-336.

Hyler, S., Gabbard, G., & Schneider, I. (1991). Homicidal maniacs and narcissistic parasites: Stigmatization of mentally ill persons in the movies. Hospital and Community Psychiatry, 42, 1044-1048.

Johnson, C., & Haney, C. (1994). Felony voir dire: An exploratory study of its content and effect. Law and Human Behavior, 18, 487-506.

Jeffrey, R., & Pasewark, R. (1984). Altering opinions about the insanity plea. The Journal of Psychiatry and Law,11, 29-40.

Kassin, S., & Wrightsman, L. (1983). The construction and validation of a Juror Bias Scale. Journal of Research in Personality, 17, 423-442.

Kempton, W. (1981). The folk classification of ceramics: A study of cognitive prototypes. New York: Academic Press.

Lockhart v. McCree, 476 U.S. 162 (1986).

Lorr, M. (1994). Cluster analysis: Aims, methods, and problems. In S. Strack & M. Lorr (Eds.), Differentiating normal from abnormal personality (pp. 179-195). New York: Springer.

MacCoun, R. (1989). Experimental research on jury decision-making. Science, 244, 1046-1049.

MacLaury, R. (1991). Prototypes revisited. Annual Review of Anthropology, 20, 55-74.

Matas, M., el-Guebaly, N., Harper, D., Green, M., & Peterkin, A. (1986). Mental illness and the media:, Part II: Content analysis of press coverage of mental health topics. Canadian Journal of Psychiatry, 31, 431-433.

Maughs, S. (1941a). A concept of psychopathy and psychopathic personality: Its evolution and historical development: Part I. Journal of Criminal Psychopathology, 329-356

Maughs, S. (1941b). A concept of psychopathy and psychopathic personality: Its evolution and historical development: Part II. Journal of Criminal Psychopathology, 465-499.

Melton, G., Petrila, J., Poythress, N., & Slobogin, C. (1997). Psychological evaluations for the courts (2^nd ed.). New York: Guilford.

Miles, A., & Huberman (1994). Qualitative data analysis: An expanded sourcebook. Thousand Oaks, California: Sage.

Moran, G., Cutler, B., & DeLisa, A. (1994). Attitudes toward tort reform, scientific jury selection, and juror bias: Verdict inclination in criminal and civil trials. Law and Psychology Review, 18, 309-328.

Ogloff, J. (1991). A comparison of insanity defense standards on juror decision making. Law and Human Behavior, 15, 509-531.

Ogloff, J., Schweighofer, A., Turnbull, S., & Whittemore, K. (1992). Empirical research regarding the insanity defense: How much do we really know? In J. Ogloff (Ed.), Law and psychology: The broadening of the discipline (pp. 171-207). Durham, NC: Carolina Academic Press.

Pasewark, R., & Seidenzahl, D. (1979). Opinions concerning the insanity plea and criminality among patients. Bulletin of the American Academy of Psychiatry and Law, 7, 199-202.

Penrod, S., & Heuer, L. (1997). Tweaking commonsense: Assessing aids to jury decision making. Psychology, Public Policy, and Law, 3, 259-284.

Pennington, N., & Hastie, R. (1992). Explaining the evidence: Tests of the story model for juror decision making. Journal of Personality and Social Psychology, 62, 189-206.

Perlin, M. (1994). The jurisprudence of the insanity defense. Durham, NC: Carolina Academic Press.

Perlin, M. (1997). “The borderline which separated you from me:” The insanity defense, the authoritarian spirit, the fear of faking, and the culture of punishment. Iowa Law Review, 82, 1375-1426.

Reed, S. (1972). Pattern recognition and categorization. Cognitive Psychology, 3, 382-487.

Rice, M., & Harris, G. (1990). The predictors of insanity acquittal. International Journal of Psychiatry and the Law, 13, 217-224.

Roberts, C., Golding, S., & Fincham, F. (1987). Implicit theories of criminal responsibility: Decision making and the insanity defense. Law and Human Behavior, 11, 207-232.

Roberts, C., & Golding, S. (1991). The social construction of criminal responsibility and insanity. Law and Human Behavior, 15, 349-376.

Roberts, C., Sargent, E., & Chan, A. (1993). Verdict selection processes in insanity cases: Juror construals and the effects of guilty but mentally ill instructions. Law and Human Behavior, 17, 261-275.

Rosch, E. (1975). Cognitive representations of semantic categories. Journal of Experimental Psychology, 104, 192-232.

Rosch, E., & Mervis, C. (1975). Family resemblances: Studies in the internal structure of categories. Cognitive Psychology, 7, 573-605.

Schneider, D. (1991). Social cognition. Annual Review of Psychology, 42, 527-561.

Shain, R., & Phillips, J. (1991). The stigma of mental illness: Labeling and stereotyping in the news. In L. Wilkins & P. Patterson (Eds.), Risky business: Communicating issues of science, risk, and public policy (pp. 61-74). Westport, CT: Greenwood Press.

Shuman, D. (1994). Psychiatric and psychological evidence. Colorado Springs: Shepard’s/McGraw Hill.

Signorelli, N. (1989). The stigma of mental illness on television. Journal of Broadcasting and Electronic Media, 33, 325-331.

Simon, R. (1967). The jury and the defense of insanity. Boston: Little, Brown & Co.

Skeem, J. (1999). Understanding juror bias and decision making in insanity defense cases: The role of lay conceptions and case-relevant attitudes. Unpublished doctoral dissertation. University of Utah, Salt Lake City.

Sklansky, J. (1996). Right to a jury trial. [Special Issue: 25th Annual Review of Criminal Procedure]. The Georgetown Law Journal, 84, 1139-1160.

Smith, E., & Medin, D. (1981). Categories and concepts. Cambridge MA: Harvard University Press.

Smith, V. (1991). Prototypes in the courtroom: Lay representations of legal concepts. Journal of Personality and Social Psychology, 76, 220-228.

Smith, V. (1993). When prior knowledge and law collide: Helping jurors use the law. Law and Human Behavior, 17(5), 507-536.

Stalans, L. (1993). Citizens’ crime stereotypes, biased recall and punishment preferences in abstract cases: The educative role of interpersonal sources. Law and Human Behavior, 17, 451-470.

Starr, V., & McCormick, M. (1993). Jury selection: An attorney’s guide to jury law and methods. Boston: Little, Brown and Co.

Tanford, J. (1992). The law and psychology of jury instructions. In J. Ogloff (Ed), Law and psychology: The broadening of the discipline (pp.305-329). Durham: Carolina Academic Press.

Taylor, S. (1981). A categorization approach to stereotyping. In D. Hamilton (Ed.), Cognitive processes in stereotyping and intergroup behavior (pp. 93-114). Hillsdale, NJ: Lawrence Erlbaum.

Tucker, L., & Messick, S., (1994). An individual differences model for multidimensional scaling. Psychometrika, 28, 333-367.

Wahl, O. (1995). Media madness: Public images of mental illness. New Brunswick, NJ: Rutgers University Press.

Wahl, O., & Roth, R. (1982). Television images of mental illness: Results of a Metropolitan Washington media watch. Journal of Broadcasting, 28, 599-605.

Walker, N. (1968). Crime and insanity in England, Vol. I: The historical perspective. Edinburgh: University of Edinburgh Press.

Way, E. (1997). Connectionism and conceptual structure. American Behavioral Scientist, 40, 729-753.

Table 1

Rotated Component Matrix for the Conception Checklist

Item	Moral Insanity	Severe, Chronic Disability	Poorly Func-tioning Social Outcast	Mental State- Related Charac-teristics
Unfeeling, manipulative person who takes advantage of others.
.75
Violent, angry, and hostile to others.
.72
Has no conscience.
.71
Control freak who needs to feel power over others.
.66
Prone to hurting himself or others.
.63
Had behavioral trouble during childhood and adolescence.
.60
Does not accept responsibility for his actions—feels that his actions are justified or feels others are to blame.
.60
Has a poor self-image.
.59		.44
Exhibits extreme, erratic, unpredictable behavior.
.58
Intoxicated on alcohol or drugs at the time of the crime.
Grossly distorted vision of reality.
.54			.46
Irrational.
.52
Acts on voices in his head that tell him what to do.
.51
Exhibits unusual, inappropriate behavior.
.47		.44
Has a certain, "off" look in his eyes.
.47		.44
Paranoid—has an unfounded fear of others and perceives everything as a threat.
.46
Impulsive—acts suddenly, without thought or planning.
.42
Delusional—suffers from false beliefs.
.41
Mentally retarded.
	.76
Mentally no more than a child.
	.74
Is extremely mentally ill—totally debilitated.
	.73
Suffered a severe head injury and brain damage.
	.69
Takes medication for his mental condition.
	.67
Hallucinates—hears voices or sees visions that others don’t perceive.
.40	.63
Has a long, documented history of mental illness.
	.59
Suffers from acute delirium.
	.56
Mental illness cannot be controlled with therapy or medication.
	.56
Incoherent—doesn’t make sense when he talks.
	.54	.44
Suffers from mania, bipolar disorder, or severe depression.
	.52

Table 1, Continued

	Moral Insanity	Severe, Chronic Dis-ability	Poorly Func-tioning Social Outcast	Mental State- Related Charac-teristics
Unable to function in society: needs help dealing with the daily problems and decisions in life. May require supervision.
	.50
Incapable of learning
	.49
Does not remember the crime.
.43	.43
Suffers from schizophrenia or psychosis.
	.42
Unstable employment history: can't hold a job.
.39		.64
Withdrawn introvert who has poor social skills and no friends.
.40		.64
Disheveled—unclean and sloppy in appearance.
.37		.59
Face is blank, lacks expression.
.45		.58
As a child, was unloved and uncared for by his mother.
.53		.55
Acts normal most of the time.
		.54
Is detached from his surroundings (i.e., the environment, people, and events taking place around him).
		.52	.37
Looks normal.
		.51
Suffers from a chemical imbalance or related medical problem.
		.50
Suffers from severe mood swings
.49		.50
Unable to understand court proceedings or assist in his defense.
		.46	.39
Impaired mental state is not the result of drug or alcohol abuse.
		.43
Suffered severe, extended abuse or trauma (physical, psychological, or sexual).
.38		.39
Not capable of understanding the harmful consequences of his actions.
			.72
Incapable of making a clear choice between right and wrong.
			.65
Is not aware of what he is doing.
			.64
Acts without reason, provocation, or rational motive.
.48			.61
Determined by a credible expert to clearly meet the criteria for insanity.
			.60
Unable to discern right from wrong.
			.57
Cannot control his thoughts, emotions, or actions.
.38			.51
Is more than “temporarily” insane.
			.47
Evaluated by a qualified mental health professional and diagnosed as mentally ill.
			.43
Confused—cannot make decisions or focus his attention.
	.37		.38
Is more than impassioned or enraged at the time of the crime.
			.37

Table 2

Prototype Clusters’ Average Scores Across Conception Checklist Components

Conception Checklist Component		Cluster 1: Severe Mental Disability	Cluster 2: Moral Insanity	Cluster 3: Mental State-Centered Characteristics
Moral Insanity	Mean	-0.4	1.0		-0.6
Severe, Chronic Mental Disability	Mean	0.6	-0.4		-0.8
Poorly Functioning Social Outcast	Mean	-0.2	0.3		0.1
Mental State-Related Characteristics	Mean	-0.2	-0.3		0.8

Table 3

Overlapping Characteristics Across Groups*

Abbreviated Item Label (and scale on which item loaded)	Cluster 1: Severe Mental Disability	Cluster 2: Moral Insanity	Cluster 3: Mental State-Centered	Mean Score Across Clusters
Unable to discern right from wrong (MSC)
6.1	5.9	6.6	6.2
Can't understand harm/consequences (MSC)
5.7	5.8	6.1	5.9
Evaluated and diagnosed as mentally ill (MSC)
5.9	5.8	6.1	5.9
Impaired MSO isn't result of drug use (MSC)
5.2	4.9	5.6	5.6
Can't control thoughts, emotions, or actions (MSC)
5.1	5.5	5.6	5.4
Isn't aware of what he's doing (MSC)
5.2	5.3	5.4	5.3
Suffers from schizophrenia or psychosis (SMD)
5.4	5.3	4.8	5.2
More than temporarily insane (MSC)
5.4	5.7	6.1	5.2
Delusional—false beliefs (MI)
5.1	5.3	5.0	5.1
Acts on voices in head (MI)
5.1	5.3	4.7	5.0
Unable to function in society (SMD)
5.2	4.6	4.4	4.7
Suffers chemical imbalance/medical problem (MSC)
4.6	4.9	4.6	4.7

* As indicated by a lack of significant (p<.05) group differences detected by ANOVA, given the mean score across groups > 4.6

Table 4

Discriminating Characteristics of the Severe Mental Disability Prototype

Abbreviated Item Label For Discriminating Characteristic^a	Cluster 1: Severe Mental Disability	Cluster 2: Moral Insanity	Cluster 3: Mental State-Centered Characteristics
TOTALLY debilitated***
6.4	5.2	5.1
Mentally retarded***
5.6	3.4	2.3
Illness cannot be controlled with treatment***
5.6	4.6	4.1
Long, documented history of illness*
5.6	5.0	4.8
Hallucinates—voices or visions**
5.4	5.1	4.1
Mentally no more than a child***
5.2	4.0	2.6
Suffers acute delirium**
5.2	5.1	4.1
Suffered injury/brain damage***
5.1	3.7	2.9
Takes medication for illness***
4.9	4.4	3.0
Suffers mania, bpd or depression*
4.7	4.7	3.9

^a As indicated by significant univariate differences among groups, given that one group’s mean feature score is > 4.6

*p <.05; ** p < .01; *** p <.001

Table 5

Discriminating Characteristics of the Moral Insanity Prototype

Abbreviated Item Label	Cluster 1: Severe Mental Disability	Cluster 2: Moral Insanity	Cluster 3: Mental State-Centered Characteristics
Has no conscience***
3.7	6.2	4.0
Grossly distorted vision of reality***
4.6	5.8	5.1
Violent, angry, hostile to others***
3.5	5.7	3.6
Extreme, erratic, unpredictable behavior**
5.0	5.7	4.6
Acts without reason, provocation** (MSC)
4.6	5.7	5.1
Feels actions are justified/blames others***
4.2	5.6	3.0
Prone to hurting himself/others***
4.3	5.5	4.4
Irrational*
4.6	5.4	4.6
Unfeeling, manipulative—takes advantage***
2.5	5.3	2.7
Suffered extended abuse/trauma***
4.8	5.1	3.6
Paranoid—perceives everything as a threat**
4.1	5.0	4.2
Detached from surroundings**
3.9	5.0	4.4
Unusual, inappropriate behavior***
3.2	4.9	3.5
Control freak who needs to feel power***
2.8	4.8	2.1
Suffers severe mood swings**
3.7	4.8	3.7
Behavioral trouble during childhood***
3.4	4.7	2.9
Impulsive—acts without planning*
3.6	4.6	4.2

^a As indicated by significant univariate differences among groups, if one group’s mean feature score is > 4.6

*p <.05; ** p < .01; *** p <.001

Table 6

Discriminating Characteristics of the Mental State-Centered Prototype

Abbreviated Item Label	Cluster 1: Severe Mental Disability	Cluster 2: Moral Insanity	Cluster 3: Mental State-Centered Characteristics
Credible expert opines that he is insane*
5.7	5.5	6.3
Incapable of choosing right or wrong**
4.9	5.6	6.0
More than merely impassioned or enraged**
4.3	5.4	5.5
Cannot assist counsel with his defense*
3.9	3.8	4.9

^a As indicated by significant univariate differences among groups, if one group’s mean feature score is > 4.6

*p <.05; ** p < .01; *** p <.001

Table 7

Post Hoc Comparisons of Insanity Prototype Clusters

Across Case Judgment Variables

Dependent Variable	Group I^b	Group J	Mean Differ-ence (I-J)^a	Stan-dard Error	Signi-fican-ce
Defendant suffers from mental disorder	MSC SMD	SMD MI MI	1.01** .96* .05	.33 .35 .29	.006 .018 .98
Defendant appreciated that what he was doing was wrong before he acted	MSC SMD	SMD MI MI	.80 .22 .58	.42 .45 .37	.14 .88 .25
Defendant was capable of perceiving reasonable alternatives to what he did	MSC SMD	SMD MI MI	-.73 -.54 -.19	.40 .43 .35	.16 .42 .85
Defendant was capable of acting differently than he did	MSC SMD	SMD MI MI	.02 .23 -.20	.41 .44 .36	.99 .87 .84
Defendant couldn’t help believing the things he believed	MSC SMD	SMD MI MI	-1.03* .99* -.04	.38 .40 .33	.02 .04 .99
Defendant was reasonably capable of rational behavior	MSC SMD	SMD MI MI	-.43 -.37 -.05	.36 .38 .31	.46 .60 .98
Defendant should be punished	MSC SMD	SMD MI MI	-1.18** -1.21** .04	.35 .37 .30	.002 .003 .99
Rating of likelihood of personal finding that defendant is not guilty by reason of insanity	MSC SMD	SMD MI MI	17.14* 15.17 1.97	6.79 7.23 5.91	.03 .09 .94

^a The construal items are 7-point scales; the likelihood item is a 100-point scale

^b MSC= Mental State-Centered Characteristics; SMD=Severe Mental Disorder; MI= Moral Insanity

* p<.05, **p<.01

N	Hallucinates—hears voices or sees visions that others do not perceive
4	hears voices
3	Hallucinates
Feature 1	sees things that are not present
1	suffers from hallucinations, especially auditory
1	actively hallucinating
1	sees and feels things that are not there

Figure 2. Example of Feature Label Development

Figure 3. Frequency Distribution of Participants Contributing to Features

Figure 4. Differences Among Prototype Groups in Case Perception

Figure 5. Differences Among Prototype Groups in Scaled Verdicts

Figure 6. Differences Among Prototype Groups in Categorical Verdicts (ns)

Endnotes

---

[1] Investigators have examined mock jurors’ judgments or subjective explanations for their verdicts based on insanity case vignettes and have inferred that jurors’ conceptions of insanity conform to “wild beast,” ALI, or multiple legal standards (Bailis et al., 1995; Finkel & Handel, 1989; Roberts et al., 1987; Robinson & Darley, 1995; see also Hans & Slater, 1984). Although one study (Finkel & Groscup, 1997) has begun to investigate “insanity case stories,” there have been no direct investigations of the nature of jurors’ conceptions or prototypes of insanity and their relationship to verdicts.

[2] Although the authors refer to these stories as prototypes, this assertion is arguably in error. First, the authors found no significant differences between subjects’ “typical” and “atypical” case narratives. However, research consistently demonstrates that prototypes manifest themselves in typicality effects: people perceive some members of a category as more typical members than others (McLaury, 1991; Way, 1997). If these stories were prototypes, there should have been differences between typical cases (sharing many features with the prototype) and atypical cases (sharing fewer features with the prototype). Second, the authors did not use prototype methodology to elicit prototypes.

[3] Finkel (1997; Finkel & Sales, 1997) has underscored the need to assess and understand individual differences in prototypes. Finkel and Groscup (1997) attempted to address this issue by defining single or multiple case stories based on any significant differences detected on their coded independent variables. However, (a) this placed an artificial limit on the number and nature of stories that could emerge for each case type (e.g., only two insanity case stories, “successful” and “unsuccessful,” could have emerged), (b) the insanity case story that emerged from their study was a consensus narrative composed of features that subjects most often described, and (c) there was wide variability in the type of crime and crime precipitants that subjects produced, suggesting that important individual differences were overlooked by this approach.

[4] Or estimates of the characteristics of nonparticipants, based on adult census statistics for the county.

[5] For example, only 5% of responses referenced sensational cases presented in the media. These cases, however, were typically listed as noncategory exemplars, or cases that would not qualify as insane based on the participant’s conception of insanity (e.g., “Ted Bundy knew exactly what he was doing.”).

[6] Jurors’ occupations were; counselor, homemaker, teacher, construction worker, and college student. Their characteristics were as follows: 60% male, 80% White, M age=44 years (Range=22-68), M educational level=14 years. Jurors were selected from cases that did not go to trial, so none of them served as jurors on the index case.

[7] This was operationalized as an average score of 4.6 on a 7-point scale.

[8] These analyses were completed only to aid in describing the prototype groups. Because they fail to account for correlated dependent variables and could not be corrected for the family-wise error rate, they should be interpreted with caution.

[9] This overlap among groups is consistent with prototype theory. Because prototypes are constructs with indefinite boundaries, groups of individuals with different prototypes of insanity will arguably overlap.

[10] As noted above, the validity of a cluster solution can be tested by determining whether resulting clusters of individuals differ on important external variables that are conceptually related to the variables that were used to cluster them (e.g., Blashfield & Aldenderfer, 1988; Morey et al., 1983). Thus, the analyses described in this section were intended not only to determine the nature of group differences in case judgments, but also to validate the cluster solution.

[11] For reading ease, masculine pronouns are used to describe prototypes and defendants, given the overrepresentation of males among criminal defendants.

[12] The discrepancy between the characteristics of the typical insanity defense acquittee and popular prototypes of insanity may be partially attributable to the fact that most (70%) acquittees are deemed NGRI through plea agreements rather than “full-fledged jury trials” (Melton et al., 1998, p. 188). Nationally, roughly 1% of defendants raise the insanity defense and only 25% are acquitted after a trial, and it seems “that the number of ‘valid’ acquittals far outweighs the number of ‘invalid’ ones” (Melton et al., 1998, p. 188; see text for a review).

[13] Further research must be conducted to test this hypothesis, however. This includes research which determines whether jurors with SMD and MI prototypes are more likely than those with MSC prototypes to convict defendants who clearly meet common statutory legal criteria (e.g., M’Naghten) for insanity.

[14] Although this study specifically focused on inductively developing the Conception Checklist, the generalizability of the measure must be tested before it can be recommended as a tool for use at voir dire.

[15] Skeem (1999) found that jurors’ opinions about the extent to which (a) mental illness is irrelevant to criminal responsibility, (b) the mentally ill have control over, and are responsible for their disorder, (c) the insanity defense is frequently abused and endangers public safety, and (d) crime control is more important than due process, are moderately-strongly predictive of their verdicts.

[16] The extent to which the MI prototype represents a conception of the typical person who raises an insanity defense or the typical person who is actually insane remains to be determined. Future research will investigate the extent to which jurors with MI prototypes are willing to deem psychotic, psychopathic defendants NGRI. Nevertheless, this research suggests that those with MI prototypes are relatively unlikely to deem the typical defendant insane.