Previous Article in Journal
Women Through the COVID-19 Pandemic: Challenges, Consequences, and Resilience
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Encyclopedia
Volume 5
Issue 4
10.3390/encyclopedia5040168
encyclopedia-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Entry

Perceptions of Sexual Offenders (PSO) Scale

by
B. J. Rye
1,* and
Amy G. Tuer
2
1
Department of Psychology, St. Jerome’s University, University of Waterloo, Waterloo, ON N2L 3G3, Canada
2
Department of Psychology, University of Waterloo, Waterloo, ON N2L 3G1, Canada
*
Author to whom correspondence should be addressed.
Encyclopedia 2025, 5(4), 168; https://doi.org/10.3390/encyclopedia5040168
Submission received: 1 August 2025 / Revised: 2 October 2025 / Accepted: 8 October 2025 / Published: 15 October 2025
(This article belongs to the Section Behavioral Sciences)
Versions Notes

Definition

Derived from the Community Attitudes Toward Sex Offenders scale (CATSO), the Perceptions of Sex Offenders (PSO) scale is 20-item instrument designed to be a practical and accurate assessment of perceptions of those who have offended sexually. The PSO was developed by Harper and Hogue through a revision of the CATSO scale, incorporating exploratory and confirmatory factor analyses with new validation samples. A three-factor scale emerged which assesses perceptions related to the sentencing and management of sexual offenders, stereotype endorsement of the person who offends sexually, and perception of the person who has offended sexually as a continued risk. This entry outlines the historical context of the PSO and the social science literature in which the instrument has been used. As well, this entry describes the development and psychometric properties of the PSO as well as the potential uses of the instrument in non-academic settings (e.g., judicial, restorative justice).

1. Introduction

Individuals who have committed sexual offences are often subject to more extreme negative opinions compared to other non-sexually offending criminal groups [1]. Public perceptions of individuals convicted of sexual offences significantly influence their treatment in social, medical, and legal systems [2,3]. Stigma also significantly hampers successful reintegration upon release for these individuals, leading to reduced access to employment, housing, and social services [4]. Furthermore, these obstacles have been linked to a higher risk of reoffending [5,6], further perpetuating cycles of harm and victimization. Therefore, it is essential to assess public attitudes toward people who have committed sexual offences to predict and subsequently address misconceptions. Fortunately, over the last several decades, researchers have established several psychometric tools to measure perceptions toward people who have committed sexual offences. One of these measures is the Perceptions of Sexual Offenders scale (PSO; [7]).
The aim of the present entry is to provide a comprehensive overview of the PSO. Section 2 describes the development of the PSO, including its origins in the Community Attitudes Toward Sexual Offenders scale (CATSO; [8]). Section 2 also reviews how the PSO has been utilized in the research literature, with attention to experimental manipulations, sensitivity to situational and perceiver characteristics (e.g., offender’s gender, perceiver’s gender, victimology, and individual differences), and correlational research. Section 3 evaluates the psychometric properties of the PSO, including reliability, discriminant and convergent validity, and evidence from factor analyses. Section 4 explores some potential practical implications of the scale for risk assessment, community initiatives, interpersonal relationships, and judicial processes. Section 5 concludes by summarizing the PSO’s contributions and outlining directions for future work.

2. Development and Use of the Perceptions of Offenders Scale (PSO)

Harper and Hogue [7] created the PSO by modifying and extending the CATSO [8] in response to replicability and reliability limitations identified in the original instrument. Following the introduction of the CATSO into the academic literature, many researchers faced challenges in replicating its original four-factor structure. Subsequent factor analyses supported a three-factor model [9,10] or even a two-factor model [11,12,13,14]. Two CATSO subscales (i.e., Capacity for Change and Social Isolation) were the most consistent across studies. However, the two other subscales (i.e., Deviancy and Severity/Dangerousness) were less psychometrically robust (e.g., poor internal consistently) across studies, indicating that the CATSO had two stronger and two weaker factors.
In re-envisioning the CATSO to develop the PSO, Harper and Hogue [7] differentiated attitudinal assessments (i.e., affective evaluations) from schematic or stereotype judgements (i.e., knowledge-based evaluations). To distinguish between attitudinal versus stereotype judgements, Harper and Hogue [7] referenced Eagly and Chaiken’s [15] definition of an attitude as an evaluation that reflects a degree of favour or disfavour toward a particular entity. In contrast, they cited Greenwald and Banaji [16] to define a stereotype as the conceptualization of prototypical traits associated with a specific group. Upon review of the CATSO with these distinctions in mind, Harper and Hogue [7] note that the majority of the original CATSO items are stereotypical beliefs (e.g., “most sex offenders are unmarried men”) with some attitude-related items (e.g., “trying to rehabilitate sex offenders is a waste of time”). Furthermore, “…the CATSO was designed specifically to address perceptions and stereotypes of sex offenders” [8] (p. 258), which aligns with Harper and Hogue’s [7] definition of a knowledge-based measure. Despite this purpose, the CATSO has been endorsed, used, and named as an attitudinal measure. In turn, Harper and Hogue [7] argue that the CATSO is not an appropriate replacement for other attitudinal measures. Instead, these authors sought to re-imagine the CATSO and promote its continued use but as an outcome measure as opposed to an attitudinal measure that could be used to detect changes in perceptions (e.g., knowledge, beliefs, and stereotypes) following educational initiatives. The instrument was thus intended to be used complementarily alongside pre-existing attitudinal measures such as the Attitudes Toward Sex Offenders scale (ATS; [17]). Whereas the ATS emphasizes strong affective reactions toward and evaluations of sexual offenders (e.g., unease, disdain, and dehumanization), the CATSO and its successor, the PSO, emphasize cognitive appraisals with less affective content (e.g., how sex offenders should be managed). To reflect the scale’s purpose more accurately and avoid ambiguity, the re-envisioned CATSO was renamed the Perceptions of Sexual Offenders scale (PSO).
Building on the content of the CATSO, Harper and Hogue [7] included eight additional items with statements related to risk and sentencing and excluded four CATSO items (see Table 1) that were previously identified to be problematic [11,13], resulting in a pool of twenty-two items. Harper and Hogue [7] then conducted a principal components analysis (PCA) using a portion of a community sample and identified a three-factor structure. A Confirmatory Factor Analysis was conducted on the remaining half of the sample using the three-factor model identified in the PCA. However, this analysis revealed a poor goodness-of-fit, prompting the removal of two more problematic CATSO items. The revised 20-item scale achieved acceptable goodness-of-fit indices (0.87) and an adjusted goodness-of-fit index (0.84). When pooling the total sample (N = 400), the PSO demonstrated strong internal consistency as a whole (α = 0.92), as well as for the three subscales Sentencing and Management (10 items, α = 0.93), Stereotype Endorsement (5 items, α = 0.85), and Risk Perception (5 items, α = 0.81). Harper and Hogue [7] concluded that the modified 20-item scale could serve as a more reliable outcome measure of perceptions towards people who have committed sexual offences. Table 1 presents the original (CATSO) and revised (PSO) item content.

2.1. PSO as an Outcome of Manipulations

Since the development of the PSO, the instrument has been used as an outcome measure, as its authors intended. Several studies have utilized the PSO to explore how linguistic influences affect perceptions of individuals who have committed sexual offences (e.g., [18,19]). More specifically, Harper and Hogue [18] administered the PSO to a sample in Britain following a widely publicized sexual offence scandal involving a British broadcaster. They compared the perceptions of tabloid readers—who were exposed to more offensive headline descriptors (e.g., “beast” and “monster”)—to those of broadsheet readers—who encountered less inflammatory language on the matter. They found that the PSO subscales differed as a function of inflammatory language. People who consume media using agitational language regarding sexual offending are more likely to have more punitive preferences for sentencing decisions, hold more negative stereotypes, and perceive people who have committed sexual offences as a greater risk than those that consume less inflammatory media.
A few studies have investigated if ratings on the PSO differed as a function of person-first language (i.e., “people with sexual convictions”) or identity-first language (i.e., “sex offender”). In all investigations, there were no differences between exposure to person-first or identity-first language, indicating that the total score of PSO is not sensitive to these linguistic manipulations [19,20,21]. However, Bernoski [20] did find that language affected the PSO indirectly through empathy, measured as an individual difference. That is, the person/identity language manipulation had a differential impact on empathy, which, in turn, had an impact on PSO scores. Thus, the relationship between language and PSO may be more complex than simply having a direct theoretical effect.
Similar to previous studies on linguistic influences, Harper and colleagues [22] sought to examine how PSO scores might differ based on exposure to an affective narrative versus an informative narrative. Half of participants were presented with a first-person perspective of a self-identified “pedophile” (i.e., an individual with a sexual attraction to prepubescent children) and his personal experiences and consequences of his sexual interests. The other half were exposed to an expert informational account, which discussed research findings on the neural basis of pedophilic sexual interests and support for early intervention. Results showed no significant differences on the Stereotype Endorsement and Risk Perception scales between the informative and affective narrative conditions. However, authors observed less punitive perceptions on the Sentencing and Management scale among participants exposed to an affective narrative compared to an informative narrative.
Studies demonstrate that the PSO is a valuable outcome measure in pre-test/post-test designs. Friestad and colleagues [23] used the PSO to assess first-year prison officer students both before and after their initial year of training. The objective was to determine whether the training influenced the students’ perceptions of individuals who have committed sexual offences as measured by the PSO. After a year of training, prison officer students showed a decrease in their scores on the Sentencing and Management scale, indicating a reduction in punitiveness. The authors also observed no difference in perceptions of risk but an increase in the endorsement of stereotypical views of sex offenders as being socially isolated. Although the training was not designed specifically to address misconceptions about offenders, this study highlights the practical application of the PSO in evaluating the influence of educational initiatives.
In a quasi-experimental design, Wurtele [24] found that students completing a criminal justice sex crime course showed reduced punitiveness (pre-post) compared to students enrolled in an introductory psychology course (unequal control group) when assessed with a four-item instrument conceptually similar to the Sentencing and Management scale but focused exclusively on restrictions on child sexual offenders. Comparing students at the beginning and at the end of an online sex crimes course, Belisle et al. [25] documented less punitiveness on the Sentencing and Management scale and a reduction in Risk Perception but no differences in Stereotype Endorsement. Also using a quasi-experimental design, Harper and Hicks [26] compared a student sample to a sample of forensic professionals (e.g., psychologists, counsellors, social workers, and academics who work with sexual offenders) on the PSO. Forensic professionals reported lower scores across all PSO scales compared to students. This finding indicates that professionals were less punitive, endorsed fewer stereotypes, and perceived less risk overall. Rothwell et al. [27] similarly compared participants studying forensic psychology to a group of general public participants with no such background. Rothwell et al. [27] found that the forensic students had significantly lower scores on the Sentencing and Management, Risk Perception, and total PSO scale relative to the general public sample. The two groups did not differ on Stereotype Endorsement. While forensic expertise might be considered an individual difference, it is a function of education and experience and is a variable that can be adjusted by intervention. These studies suggest that forensic training or criminological education intervention may lead to reduced punitiveness, lower stereotype endorsement, and/or diminished perceived risk, as reflected in PSO scores.
The studies discussed above demonstrate that various manipulations can influence the perceptions of individuals who have committed sexual offences, and these changes in perception can be measured using the PSO. Collectively, these findings support the contention that the PSO is a valid outcome measure. However, linguistic modifications included in the PSO, such as changing the label “sex offender,” had no impact on PSO scores [19,20,21], suggesting that the PSO is not sensitive to some manipulations or needs to be considered in conjunction with individual difference factors [20]. Instead, more direct and effective changes in perceptions are fostered through the dissemination of information, whether through formal education [23], exposure to brief narratives [22], or media consumption [18]. More specifically, educational initiatives can help influence views on sentencing, leading to decreased punitive attitudes (e.g., [22,23]). This inference aligns well with the intended purpose of the PSO, which is to identify perceptions and subsequently influence them meaningfully through educational efforts. However, perceptions of risk and stereotypes did not result in positive changes (i.e., Risk Perception and Stereotype Endorsement scores) in response to learning [22,23]. That said, Friestad and colleagues [23] and Harper and colleagues [22] did not deploy interventions explicitly intended to target stereotypes or educate on risk. Targeted stereotypic and risk-relevant interventions may be more effective in achieving changes in these areas.

2.2. PSO Sensitivity to Situational and Perceiver Characteristics

Other researchers have utilized the PSO to examine how various characteristics of both the sample and the offence scenario described could influence perceptions.

2.2.1. Offender Gender

Offender gender (i.e., male or female) and offender age (i.e., adolescent or adult) generally has not significantly influenced PSO scores [27,28,29]. For example, Osagie and Pica [29] instructed participants to respond to the PSO items with reference to either female or male offenders. There were no significant differences in perception as a function of gender. One exception was Harper and Hicks [26], where participants exhibited more stereotypical views towards male offenders. While this was a significant gender-of-perpetrator effect, it was very weak (partial eta-squared = 0.01).

2.2.2. Perceiver Gender

In terms of the participants, Leon and Rollero [30] found that the gender of the perceiver had effects such that an online sample of Italian female participants reported greater Risk Perception scores than their male counterparts, whereas male participants had increased scores on Stereotype Endorsement, indicating they held more negative stereotypes relative to the females in the sample. However, there were no significant gender-of-perceiver differences for Sentencing and Management ratings. In contrast, though, Harper and Hogue’s original study [7] documented that female participants scored higher than male participants on Sentencing and Management, but there were no significant participant gender differences on Stereotype Endorsement. Like Leon and Rollero [30], Harper and Hogue [7] found females to exceed males in Risk Perception scores. Using a PSO total score, Biteus and Tuiskunen [31] did not find participant gender differences evaluating rape versus sexual harassment offences. In short, there are not consistent findings regarding the perceiver’s gender except that women may perceive offenders as riskier than men do.

2.2.3. Victimology

Characteristics around the victim and offence situation seems to produce some PSO differences. For example, Rothwell et al. [27] found perceptions were generally harsher in situations involving child victims versus adult victims. Despite this total PSO score difference, there were no significant differences between child and adult victim types across the three individual subscales (i.e., Sentencing and Management, Risk Perception, and Stereotype Endorsement). Further, rape versus sexual harassment cases using PSO scales produced differential results, with the former producing more negative ratings [31]. Thus, there is limited evidence that the PSO is sensitive to some victim characteristics. Future research could investigate whether victim type alters PSO scores—for example, whether sexism underlies differences in the ratings of male versus female victims or whether homonegativity or transnegativity affects evaluations of heterosexual versus LGBTQ+ victims.

2.2.4. Perceiver Individual Differences

A number of characteristics of the perceiver have been investigated in relation to the PSO scale. For example, Harper and Bartels [28] examined the role of “entity” implicit theories (i.e., the belief that a behaviour or trait is fixed and cannot change over time) and “incremental” implicit theories (i.e., the belief that a behaviour can change and is subject to variation) using an Implicit Theories About Sexual Offenders measure. They found participants who held higher “entitist” views (i.e., believing that sexual offenders do not change; essentialist beliefs) scored more negatively on the total PSO, exhibiting more punitive sentencing and stereotypes, and they perceived greater risk compared to those who held more “incrementalist” views (i.e., believing that individuals who commit sexual offences can change; constructivist beliefs).
A few other researchers have investigated dispositional types of variables in relation to the PSO. For instance, Boland [21] found that participants with a preference for cognitive processing held significantly less negative perceptions on the PSO than participants who preferred affective processing styles. Boland’s [21] findings are consistent with the original Harper and Hogue [7] investigation, which found significant but very weak correlations of the PSO scales with logical and affective processing styles (rs < 0.23). Bernoski [20] also found that state-level empathy had a relationship with PSO scores but did not describe the nature of that relationship (presumably, more state empathy was related to more favourable perceptions on the PSO). Bernoski [20] also reported that humanization—a pictorial depiction of how evolved a target group is—was related to scores on the PSO, with those viewing sex offenders as more human rating them more favourably on total PSO scores (i.e., reduced punitiveness, lower stereotype endorsement, and/or diminished risk perception). Humanization might reflect an affective orientation toward or evaluation of the group; this finding demonstrates how the overall evaluation of a group is likely reflected in or predictive of perceptions of sex offenders.
Martens and Stewart [32] examined whether participants’ attachment styles (i.e., avoidant, secure, or anxious) predicted PSO scores. Attachment style was measured continuously. While attachment style did not predict the total PSO scores, higher avoidance predicted lower punitiveness on the Sentencing and Management scale, indicating that individuals with greater avoidance held less punitive sentencing and management views. Contrary to theoretical expectations, a higher anxious attachment predicted a lower perceived risk on the Risk Perception scale, suggesting that individuals were less likely to view sexual offenders as dangerous despite typically viewing the world as an unsafe place. Finally, securely attached participants had a weaker endorsement of stereotypes on the Stereotype Endorsement scale, consistent with the relatively optimistic and positive world view characterizing secure attachment.
Snape and Fido [19] investigated two general personality dimensions in relation to the PSO. At a zero-order level, openness did correlate significantly but weakly (r = −0.18) with the PSO. However, neither openness nor neuroticism were predictive of PSO scores. A few demographic characteristics of the perceiver have been investigated in a preliminary fashion. For example, related to children as victims, Osagie and Pica’s [29] research found that those who were parents had harsher perceptions. Harper and Hogue [7] present some evidence that older people and people with more education are less punitive on the PSO scales.
Together, these initial studies regarding individual differences and dispositions suggest that the PSO scores might be influenced by pre-existing cognitive, affective, and interpersonal orientations. Individuals are not “blank slates” regarding the evaluation of sex offenders; rather, prior beliefs, processing modes, attachment styles, emotional orientations, and the like may be brought to bear on PSO responses. Prior individual differences could predispose perceivers to be harsher in their ratings (e.g., more affective processing style) while other dispositions may predispose perceivers to be more lenient (e.g., secure attachment). However, this remains an emerging literature, and further investigation is warranted before definitive conclusions can be drawn about the role of perceiver individual differences as predictors of the PSO.

2.2.5. The PSO as a Predictor/Correlate

There have been some predictive findings in relation to the PSO. For example, Martens and Stewart [32] found a significant correlation between PSO scores and a Perceptions of Partners of Sex Offenders measure, potentially exemplify a phenomenon known as attitudinal transference (i.e., from attitudes toward offenders to attitudes toward the offender’s family). Consequently, PSO ratings may reflect a cognitive bias whereby negative views of offenders generalize to those associated with them, known as the reverse-halo or devil effect.
Perplexingly, Martens and Stewart [32] found that the PSO did not correlate to a sex offender feeling thermometer (r = −0.15). Feeling thermometers are one-item measures assessing warmth/coldness toward a target group and are thought to reflect an affective response. This finding is bemusing, as the PSO should be strongly correlated with an overall evaluation of sex offenders (and the PSO was more strongly correlated with the feeling thermometer rating of partners of sex offenders). Indeed, Harper and Hogue [7] present extremely strong correlations between the total PSO score, Sentencing and Management, and Risk Perception with the Attitudes Toward Sex Offenders Scale (rs > 0.78).
Wheeler and colleagues [33] investigated the PSO scales as predictors of health outcomes of 91 UK forensic healthcare workers (e.g., nurses, support workers). Multiple regression analyses yielded counterintuitive results: participants who endorsed harsher sentences (on the Sentencing and Management scale) and perceived greater risk (on the Risk Perception scale), among other predictors, demonstrated lower levels of secondary traumatic stress—defined as symptoms similar to post-traumatic stress disorder but arising from exposure to others’ traumas. Likewise, harsher sentences (on the Sentencing and Management scale) were predictive of lower levels of worker burnout. Worker satisfaction—defined as a sense of making a difference and contributing to society by performing meaningful work—was not predicted by any of the PSO scales. It is possible that the negativity reflected on PSO scales helps protect the workers from the negative psychological consequences of this type of work. This suggests that the relationship between PSO scales and well-being outcomes of forensic professionals is likely not straightforward. Additional investigation into the correlates of the PSO, particularly with a variety of attitudinal and well-being measures, would strengthen both its predictive and construct validity.

2.3. Summary of Uses of the PSO

There are over a dozen studies that have employed the PSO instrument [7,18,19,20,21,22,23,25,26,27,28,29,30,31,32,33], reflecting its growing popularity. These studies—primarily small-scale and involving experimental, quasi-experimental, or correlational designs—offer preliminary support for its effectiveness as an outcome measure. Yet, the PSO also appears sensitive to certain participant characteristics. While the growing body of literature indicates the potential of the PSO, the use of this instrument remains in its early stages in research and applied settings.

3. Psychometric Properties of the PSO

Studies using the PSO have assessed the instrument using different scales and scoring techniques. For example, Harper and colleagues [7,18,22,26,28] measured items on 6-point scales and added the responses within each scale to produce individual scale scores. Sometimes researchers calculate a total score by summing responses across all scale items. However, this total score disproportionately reflects the Sentencing and Management scale, which has 10 items, compared to the Stereotype Endorsement and Risk Perception scales, which each have five items (i.e., 50% vs. 25% each). Many researchers have followed Harper and Hogue’s addition method of scoring the scales (e.g., [19,27]). Others have measured on a 7-point scale [20,32], allowing for a neutral mid-point. In both of these studies, the researchers used a mean-based score rather than an addition calculation, which allows for an easier comparison of ratings between the three PSO scales within and between studies. Friestad et al. [23] also used average scores but employed the 6-point rating scale. Further, some studies report standard errors while others report standard deviations with scale means. While both statistics reflect variability, standard deviations describe dispersion within a sample, whereas standard errors reflect the precision of the mean, making direct comparisons difficult. Finally, some researchers elected to convert scales using a formula (i.e., Biteus & Tuiskunen [31] calculated scale scores ranging from 0 to 1). Regardless of the scoring method, examination of the various means suggests, on an absolute level, that respondents generally disagree with the PSO items. Because most items are phrased negatively—and positively phrased items are reverse-coded—this general disagreement reflects more favourable assessments of offenders. That is, higher agreement scores (usually six or seven, representing “strongly agree”) indicate punitive views. Generally, means fell below the mid-point, with exceptions sometimes occurring for the Risk Perception scale (Appendix A presents PSO scale scores across studies).

3.1. Reliability and Cultural Context

The PSO is a relatively new instrument but has already been applied in diverse international contexts, predominantly within Europe (see Table 2). It has been administered to public, student, and professional samples. It has been translated into Farsi/Persian [34], Greek [35], and Swedish [31]. The PSO was initially developed and tested using a British sample [7], and many of the investigations thereafter have also used UK samples.
The PSO scales demonstrate good internal consistency across studies (see Table 2). In particular, Sentencing and Management is strongly consistent (i.e., α range of 0.81 to 0.93; median = 0.89), followed by Stereotype Endorsement with an alpha range of 0.61 to 0.91 and a median of 0.83. The Risk Perception scale has much lower and inconsistent reliability statistics reported in the literature (i.e., α range of 0.22 to 0.87; median = 0.68). No study has reported omega reliability coefficients [36].
It is possible that the reliability of the Risk Perception scale is sensitive to language modification. For example, Snape and Fido [19] adjusted the language of the PSO from “sex offender” to “people with sexual convictions” for half of their sample as part of their experimental design. However, the linguistic adjustment seems an unlikely driving factor for the drastically reduced internal consistency (α = 0.22) reported in Snape and Fido’s [19] study, as only half the sample was exposed to the modified version. Biteus and Tuiskunen [31] also changed the items from “sex offender” to ”sexual harasser” or “rapist”, as well as “sex offence” to ”sexual harassment” or “rape.” Further, the internal consistency seems to not be systematically affected by language translation [31,34,35]. While wording adjustments might be a viable explanation for instability, it is unclear why such modification—either through wording changes or translation—would disproportionately affect the Risk Perception scale compared to the other scales.
Comparing sample composition, those with students as participants sometimes demonstrate lower internal consistency metrics for Risk Perception (e.g., [22,23,31,32]), although this finding did not hold true for all student samples (e.g., [26,27]). Examination of the differences in reliability observed between public samples, student samples, and mixed samples suggest that the reduction in Risk Perception reliability could be proportional to the number of students included in the sample. Possible explanations include greater homogeneity in student responses or shared experiences that differentiate student samples from other samples [23]. The latter is a more viable explanation, as reliability changes were evidenced in the same sample longitudinally (i.e., αbefore training = 0.52 to αafter a year of training = 0.64; [23]). As well, research indicates experiential learning opportunities significantly impact attitudes toward offending populations [37,38], which may influence the stability of the Risk Perception scale as a function of an increased understanding of forensic risk assessment and risk level. However, when Friestad and colleagues [23] compared mean scores on the PSO scales between prison officer students who had experience working with specialized units containing sexual offenders to those without such experience, there was no significant difference in scores, although the test was nearly significant (i.e., p = 0.055) for the Risk Perception scale. Thus, more research is required to determine if increased exposure to sexual offenders could create greater variability in risk perceptions.
Another explanation for lower internal consistency in the Risk Perception scale is cultural cohort effects. When the PSO was used with a Greek community sample, the Risk Perception scale was excluded by the researchers because it did not emerge as a unique factor [35]. Conversely, an Italian community sample reported adequate internal consistency (α = 0.70) [30], while a U.S. public sample showed below-adequate internal consistency (α = 0.65) [29]. Potential cultural reasons exist for Risk Perception instability, for example, Greek samples may lack risk assessment understanding [35], perhaps because of a dearth of impact of psychological theory and practice into their social systems [39]. As well, Greek perceptions of crime tend to be attributed to systemic issues rather than individual offenders [40]. The Greek situation is contrasted with the Italian experience where Italy promotes treatment and support for the reintegration of individuals who have committed sexual offences with greater trust in psychological sciences and stronger beliefs in rehabilitation [41,42]. In the United States, there is greater awareness of sexual offending due to increased media coverage; however, the public often subscribes to popular myths, perceiving the risk of reoffending as higher than what empirical evidence suggests, while underestimating the effectiveness of treatment [2]. Thus, the decreased stability of Risk Perception may be a function of cultural considerations.

3.2. Discriminant Validity

Intercorrelations between subscales within an instrument can be low, indicating that the scales are measuring different constructs. Or, it is possible that scale intercorrelations may be modest, indicating the subscales are related [43]. For the latter case, there should be a good conceptual reason for the overlap. Regardless of strength of the relationships, examination of subscale intercorrelations speaks to instrument discriminant validity. For the PSO, it is expected that the three subscales are measuring unique, distinct constructs and, therefore, do not overlap substantially. A few studies [23,26,30,34,35] report interscale correlations (see Table 3). Examination of Table 3 suggests that relationships between the scales are relatively weak [44]. For example, Friestad and colleagues [23] assessed six possible PSO scale intercorrelations (i.e., three at each of the two time points). Of these six potential interscale correlations, only two significant correlations were reported (i.e., Sentencing and Management with Stereotype Endorsement and with Risk Perception). The other four correlations were lower than 0.30, hence, they were weak or poor [44]. Overall, these weak-to-modest correlations among the PSO scales are consistent with the explanation that the PSO scales are measuring relatively distinct constructs. Considering this data is limited to only a few studies, the conclusion must be viewed as tentative.

3.3. Convergent Validity

Numerous studies have compared the PSO to other tools designed to assess attitudes toward individuals who have committed sexual offences (see Table 4). The most commonly used measure for this purpose is the Attitudes Toward Sex Offenders scale (ATS; [17]). Strong negative correlations have been found between total scale scores of the PSO and ATS [7,28]. An examination of Table 4 indicates that Harper and colleagues have found significant correlations between each of the PSO scales and the ATS. The overlap between the Sentencing and Management scale and attitudes toward sex offenders as well as general punitiveness is somewhat concerning. That is, the Sentencing and Management overlap is rather high, and given that Harper and Hogue [7] have argued that the PSO is a distinctly different construct (i.e., more cognitively based than affectively attitudinally based), these correlations are suggestive of a lack of discriminant validity rather than concurrent validity. However, the use of the ATS as a covariate in analyses helps address this concern (e.g., [22]).
The General Punitiveness Scale (GPS; [45]) assesses respondents’ perceptions of how harshly any offender should be punished. The Sentencing and Management scale correlated most strongly with this punitiveness measure [7,30], thereby suggesting convergent validity. Togas and colleagues [35] also developed their own measures for sentencing and management and stereotype endorsement to assess the construct validity of the PSO in relation to the Sentencing and Management and Stereotype Endorsement scales, respectively. Sentencing and Management correlated strongly with the crafted sentencing and management index, providing convergent validity; however, the rather strong correlation with the stereotyping index (r = 0.61) is also perplexing, especially given that Stereotype Endorsement correlated less strongly with the stereotyping index (r = 0.36). Similarly, Harper and Hicks [26] developed a risk judgement index that correlated modestly, as would be expected, with the Risk Perception scale. Yet, the Sentencing and Management scale correlated as strongly with the risk judgement index. Greater assessment of the convergent validity of the PSO scales is needed.

3.4. Factorial Validity

Factorial or structural validity is a form of construct validity [46]. One issue addressed by PSO factorial validity analyses is whether the three factors should be considered separately within research or if it is worthwhile to use a total PSO score. Aside from the study in which it was developed [7], three published studies have investigated the factor structure of the PSO. Using Confirmatory Factor Analysis (CFA), Togas et al. [35] did not replicate the three-factor structure of Harper and Hogue [7] but rather found a modified structure fit their data better. In short, Togas et al.’s [35] two-factor model replicated the Sentencing and Management and Stereotype Endorsement scales; three of the items from the Risk Perception scale loaded on the Sentencing and Management scale instead and two items were dropped (items 4 and 9, see Table 1 for item content).
Bakhtiari et al. [34] conducted a CFA on the PSO, as well. They found strong psychometric support for the validity of the PSO. Friestad and colleagues [23] conducted principal component analyses (PCA) both before and after a year of training for prison officers. It should be noted that the PCA procedure is an exploratory—not a confirmatory—analytic technique. For both time points, the authors successfully replicated Harper and Hogue’s [7] original three factors by constraining the PCA. However, there was variation in the loadings—somewhat akin to Togas et al.’s findings [35]—for individual items across these three factors. Additionally, the total variance explained by the three factors was lower prior to training than it was at the end of the training.
When Friestad et al. [23] did not constrain the PCA to three factors, a five-factor structure emerged at both time points. However, the item distribution across these five factors varied between assessments, with some items loading equally highly on different factors. Furthermore, the post-training PCA produced two factors represented by only one item each. When the authors explored a pre-defined four-factor structure, the results were described as “equally incoherent.” In short, there is support for the structural validity of two and sometimes three factors, but further investigation of the PSO’s structure is needed.

3.5. Summary of Psychometric Properties

There is preliminary evidence supporting the PSO’s factor structure, particularly for the Sentencing and Management and Stereotype Endorsement scales, which demonstrate a relatively stable structure, strong internal consistency, and good concurrent validity. In contrast, the Risk Perception scale is the weakest, with a questionable structure and lower reliability. Although the evidence base remains limited, there is support for the PSO’s convergent and discriminant validity. Use of a total PSO score is not recommended due to conceptual and statistical inconsistencies across scales. The Sentencing and Management scale appears the most robust, and while early findings suggest overall psychometric promise, further analyses are warranted.

4. Practical Implications

4.1. PSO as an Augmentation to Risk Assessments

Risk assessments are widely used to estimate recidivism risk and guide legal, clinical, and supervisory decisions [47]. While structured actuarial tools are more objective than unstructured clinical judgement [48], they are not immune to bias. Subjective overrides—where assessors adjust scores based on professional judgement—can inflate risk levels, especially for individuals convicted of sexual offences, undermining the predictive validity of risk-assessment instruments [49]. Such biases may also reflect adversarial allegiance, with assessors’ affiliations—such as having a defence or prosecution perspective—influencing reported risk [50]. The PSO could be used to help assessors recognize and manage these biases, improving fairness in risk evaluation. Alternatively, it could be used to identify individuals in the decision-making process who hold overly punitive viewpoints.
Future research could explore how PSO scores relate to reported risk levels to understand the role of perceptions in the assessment process. For example, to examine cultural disparities in clinical forensic risk assessments [51,52], future studies could use vignettes describing sexual offences committed by individuals from diverse cultural backgrounds to assess perceptional differences using the PSO. Such research could help determine whether disparities arise from sociocultural biases embedded within the risk assessment tools themselves or from cross-cultural rater bias. Depending on the findings, such outcomes may support the need for revisions to risk measures or the development of specialized training to enhance cultural and racial fairness in forensic assessments.

4.2. Community Initiatives

Previous studies have highlighted the negative consequences that stigmatization and social rejection have on individuals who have committed sexual offences, such as an aversion to social participation [53] and reduced willingness to seek treatment [54]. In turn, such consequences may increase criminogenic risk factors associated with sexual reoffending, such as feelings of loneliness [55]. Ultimately, the sequela of stigmatization poses a risk to the community and perpetuates a harmful cycle of offending and victimization. Consequently, many researchers have emphasized the effectiveness of adopting a restorative approach and the increased need for support for these individuals as they reintegrate into the community [56].
There are several effective restorative community initiatives geared to support the reintegration of people who have committed sexual offences (e.g., [57,58]). Restorative justice is an approach to addressing and healing from crime that emphasizes repairing harm, promoting accountability, and involving all affected parties, including victims, offenders, their families, and the community. The Circles of Support and Accountability (CoSA) is one of the most renowned initiatives, first established in Canada and has since been implemented in the United Kingdom [57,59]. CoSA is a community-based organization that utilizes informal networks of support founded on principles of restorative justice. The circle supporting the core member (i.e., the person who committed the sexual offence) involves a CoSA coordinator and 2–4 community volunteers whose goal it is to help the core person reintegrate into the community and hold them accountable for both their past behaviour and commitment to desistance [59]. Given the vital support volunteers provide and considering the informal and unstructured nature of these circles, it is essential to evaluate their beliefs to ensure the core member’s emotional safety. The PSO may help CoSA coordinators and other similar community-based initiatives identify volunteers with the most facilitative beliefs.
Another community-based initiative is the restorative conference. The conference involves face-to-face dialogue between the perpetrator, victim, and stakeholders such as friends, family, and members of the affected community to address the harm caused by the sexual offence and develop a plan to repair it [60]. The PSO could be used to gather perceptions of conference attendees toward individuals who have committed sexual offences, helping the coordinator and offender prepare for potential inclinations and areas of challenge that may arise during the conference. Future research could also examine how these conferences may influence perceptions toward people who have committed sexual offences using the PSO as an outcome measure pre- and post-conference.

4.3. Interpersonal Relationships

Personal pro-social networks and close intimate relationships are crucial for maintaining desistance among individuals who have committed sexual offences [61]. However, those closely associated with an offender often face significant collateral challenges that can negatively interfere with the relationship. These challenges include stress, feelings of shame, social rejection, stigmatization, and increased financial difficulties [62]. Families may also experience adverse treatment from media outlets and justice officials, which can lead to feelings of self-blame and secrecy [63]. As discussed earlier, Martens and Stewart [32] also found that perceptions toward people who commit sexual offences, measured using the PSO, are highly correlated with perceptions toward their partners. These findings further illustrate that partners of people who have committed sexual offences are judged by association, adding to social reintegration challenges. The PSO could be used or adapted to explore attitudinal transference further.
The PSO can help determine how the offender’s friends, family, and partners perceive those who have committed sexual offences. More importantly, the PSO can assist in identifying any harmful and inaccurate stereotypes and prejudices they may hold in relation to their loved one. Identifying these perceptions can subsequently inform rectification strategies—such as psychoeducation—to dismantle inaccurate beliefs and ultimately support the offender and their loved ones as they navigate their reality together. Furthermore, future studies could use the PSO in clinical settings to investigate if the perceptions held by the offender’s loved ones moderate the severity of their psychological hardships (e.g., well-being, stress, depression, anxiety, loneliness, and perceived rejection among the offender’s loved ones). In this way, the PSO could assist in customizing clinical treatment plans for the offender and their interpersonal support network.

4.4. Judicial Processes

There is a history of widespread beliefs, propagated by myths in the media, that individuals who have committed sexual offences will indefinitely reoffend [64]. However, Hanson and colleagues [61] conducted a longitudinal study that showed that the actual risk of sexual reoffending is significantly lower than commonly assumed. For instance, low-risk offenders consistently demonstrate recidivism rates between 1% and 5%, while the highest-risk offenders have a recidivism rate of 22% within the first 5 years after release, which drops to 4.2% after they have been offence-free for 10 years in the community. If those with decision-making power—such as those working in the justice system—hold these unfounded beliefs, it could potentially lead to more punitive outcomes than are justified.
Various measures have been used to gauge beliefs among justice-related professionals to identify biases and predict decisions. While several studies employed the CATSO among criminal justice actors [10,11,14,65], most of these reported poor internal consistency across the four factors. The ATS scale has also been employed to assess attitudes toward sexual offenders among various professional groups working in the justice system but has been critiqued for not explicitly addressing stereotypical beliefs about sexual offenders [66]. Only two studies have used the PSO among forensic professionals [23,26]. Harper and Hicks [26] found forensic professionals (i.e., psychologists, facilitators, and social workers) reported less punitive attitudes, fewer stereotypes, and less risk perception compared to students; further, their judgements were not swayed by the perpetrators age or gender. These are ideal findings, as they indicate that the professionals have perceptions that better align with empirical risk estimates and hold fewer negative perceptions that are harmful to the clientele with whom they are working. However, Friestad and colleagues [23] found a decrease in punitive views on the Sentencing and Management scale, unchanged scores on Risk Perception, and increased Stereotype Endorsement among first-year prison officer students after their initial year of training. While the training was not explicitly intended to counteract biases, this study demonstrates the maintenance and even escalation of harmful perceptions of people working in justice settings, highlighting the need for ongoing identification of beliefs and counterbalancing initiatives in this field. Future studies may utilize the PSO to explore the perceptions of justice professionals, particularly those with significant decision-making authority, such as judges and jury members. This approach can also serve as a metric to assess the efficacy of educational initiatives aimed to align perceptions with evidence-based risk assessments and track subsequent decision-making trends.

5. Conclusions

The PSO holds considerable promise as a resource in both academic and professional contexts, extending beyond its research origins. Its use has been methodologically diverse, serving as an outcome measure, a correlate, and a predictor in scholarly work on attitudes toward individuals who have committed sexual offences. However, its psychometric properties, particularly at the subscale level, require further validation, and investigations to date have been largely atheoretical, which limits its broader implementation (e.g., the PSO could be investigated in relation to social cognition models of bias or public policy decision paradigms). Existing evidence is not yet sufficient to support its formal use or its application as a stand-alone metric in critical decision-making processes. Cultural variability in scores also remains insufficiently explored. Nonetheless, careful application of the PSO in practical settings may enhance its psychometric validity and inform ongoing refinements. Continued use in research contexts, with appropriate caveats, can help clarify its strengths and limitations while contributing to theory development. It may be particularly valuable for identifying and targeting educational needs within academic, clinical, judicial, and community-based settings. Ongoing research—especially in applied contexts—is essential to strengthen the PSO’s credibility, reliability, and theoretical foundation. With continued refinement and cross-cultural validation, the PSO could become a credible instrument for both research and practice.

Author Contributions

Conceptualization, B.J.R. and A.G.T.; formal analysis, B.J.R. and A.G.T.; writing—original draft preparation, B.J.R. and A.G.T.; writing—review and editing, B.J.R. and A.G.T.; funding acquisition, B.J.R. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding but was funded by St. Jerome’s University internal faculty research grant, grant number 430.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

No new data were created or analyzed in this study. Data sharing is not applicable to this article.

Conflicts of Interest

The authors declare no conflicts of interest.

Abbreviations

The following abbreviations are used in this manuscript:
CATSOCommunity Attitudes Toward Sex Offenders
CFAConfirmatory Factor Analysis
PCAPrincipal Components Analysis
PSOPerceptions of Sex Offenders

Appendix A

Table A1. Summary of descriptive statistics of studies employing the PSO.
Table A1. Summary of descriptive statistics of studies employing the PSO.
AuthorsRegion of StudySample TypeNScaleExperimental
Condition/
Demographic		Mean (Standard Deviation/Standard Error)
SMSERPTotal
Harper & Hogue [7]UKGeneral
Public		4001–618–27 Years 39.97 (16.09)
28–46 Years 35.18 (18.44)
Over 46 Years 35.26 (15.61)
<Undergraduate Degree 44.41 (21.12)
Undergraduate Degree 35.37 (14.85)
>Undergraduate Degree 31.78 (12.26)
Tabloids Only 45.83 (18.07)
Tabloids and Broadsheets 36.37 (15.49)
Broadsheets Only 32.64 (16.23)
No Newspapers 38.44 (16.81)
Biteus & Tuiskunen [31]SwedenStudents 1861–5Sexual Harassment0.66 (0.14)0.53 (0.22)0.44 (0.18)
Rape0.50 (0.19)0.47 (0.19)0.39 (0.24)
Harper & Hogue [18]UKGeneral
Public		591–6Tabloids17.96 (0.84)8.45 (0.52)16.00 (0.44)48.41 (1.17)
156Broadsheets15.08 (0.51)7.37 (0.32)15.33 (0.27)37.78 (0.71)
105Both16.57 (0.627.40 (0.39)16.43 (0.33)40.40 (0.86)
201Neither16.21 (0.45)8.23 (0.28)15.77 (0.24)40.21 (0.63)
Harper & Bartels [28]UKGeneral
Public		2521–6Adult Male—Entitist 58.06 (14.21)
Adult Female—Entitist 53.09 (14.68)
Juvenile—Entitist 54.00 (16.36)
No Vignette—Entitist 56.18 (14.86)
Adult Male—Incrementalist 22.86 (12.87)
Adult Female—Incrementalist 29.56 (12.14)
Juvenile—Incrementalist 28.94 (10.50)
No Vignette—Incrementalist 32.35 (13.53)
Harper et al. [22]UKStudents 1001–6Narrative9.52 (0.98)12.42 (0.64)13.88 (0.53)
Informative14.55 (0.99)10.93 (0.65)15.25 (0.54)
Martens & Stewart [32]UKStudents1061–7No Condition3.32 (1.04)3.67 (0.86)4.83 (0.97)3.79 (0.70)
Leon & Rollero [30]ItalyGeneral
Public		7681–7Men3.14 (1.19)2.78 (0.98)4.74 (0.95)
Women3.09 (1.09)2.36 (0.98)5.22 (0.70)
Rothwell et al. [27]UKForensic
Students		611–6Adult Victim 21.46 (14.38)10.02 (5.07)18.43 (4.04)50.09 (18.55)
General
Public		40Child Victim22.48 (14.32)10.31 (5.26)19.08 (3.56)51.86 (18.7)
Snape & Fido [19]UKGeneral
Public		1861–6Label-First 48.02 (12.38)
Person-First 47.24 (11.96)
Harper & Hicks [26]UKStudents 3411–6Students—Male16.86 (9.49)12.07 (4.39)45.87 (3.92)
Students—Female17.70 (8.35)10.60 (4.17)16.05 (3.74)
Students—Juvenile15.59 (8.25)10.32 (4.25)16.16 (3.65)
Forensic
Professionals		186Professionals—Male7.15 (6.41)9.45 (4.77)14.19 (4.86)
Professionals—Female6.33 (4.16)8.47 (4.07)12.43 (3.93)
Professionals—Juvenile6.90 (5.42)9.36 (4.09)13.34 (4.52)
Togas et al. [35]GreeceGeneral
Public		4521–6No Condition35.29 (11.78)8.89 (4.33)
Bakhtiari et al. [34]TehranConvenience (mostly students, young professionals)
Boland [21]USGeneral
Public 		8471–6Cognitive 57.82 (0.68)
Affective 50.87 (0.61)
Person-First 54.65 (0.64)
Label-First 54.04 (0.66)
Osagie & Pica [29]USGeneral
Public		226 Male Offenders 66.10 (11.03)
Female Offenders 58.08 (11.02)
Friestad et al. [23]NorwayPrison Officer
Students		1811–6Pre-Training 2.41 (0.68)2.92 (0.77)4.44 (0.65)
110Post-Training2.01 (0.72)3.30 (0.85)4.47 (0.72)
Bernoski [20]USStudents—2 Conditions1171–6Person-First 52.83 (12.32)
120Label-First54.51 (12.79)
Belise et al. [25]UKStudents Enrolled in Online Sex Crimes Course641–6Pre-Course 27.26 (7.83)9.17 (4.19)17.67 (3.76)
Post-Course22.66 (8.59)9.94 (4.87)16.16 (4.14)
Table Notes. SM = Sentencing and Management; SE = Stereotype Endorsement; RP = Risk Perception; and Total = Whole Scale. UK = United Kingdom; US = United States.

References

  1. Rogers, D.L.; Ferguson, C.J. Punishment and rehabilitation attitudes toward sex offenders versus nonsexual offenders. J. Aggress. Maltreatment Trauma 2011, 20, 395–414. [Google Scholar] [CrossRef]
  2. Fortney, T.; Levenson, J.; Brannon, Y.; Baker, J.N. Myths and facts about sexual offenders: Implications for treatment and public policy. Sex. Offender Treat. 2007, 2, 1–17. [Google Scholar]
  3. Levenson, J.S.; Brannon, Y.N.; Fortney, T.; Baker, J. Public perceptions about sex offenders and community protection policies. Anal. Soc. Issues Public Policy 2007, 7, 137–161. [Google Scholar] [CrossRef]
  4. Grossi, L.M. Sexual offenders, violent offenders, and community reentry: Challenges and treatment considerations. Aggress. Violent Behav. 2017, 34, 59–67. [Google Scholar] [CrossRef]
  5. Hanson, R.K.; Harris, A.J.R. Where should we intervene?: Dynamic predictors of sexual offense recidivism. In Current Perspectives on Sex Crimes; Holmes, R.M., Holmes, S.T., Eds.; Sage: Thousand Oaks, CA, USA, 2002; pp. 300–319. [Google Scholar] [CrossRef]
  6. Willis, G.M.; Grace, R.C. The quality of community reintegration planning for child molesters. Sex. Abus. 2008, 20, 218–240. [Google Scholar] [CrossRef]
  7. Harper, C.A.; Hogue, T.E. Measuring public perceptions of sex offenders: Reimagining the community attitudes toward sex offenders (CATSO) scale. Psychol. Crime Law 2015, 21, 452–470. [Google Scholar] [CrossRef]
  8. Church, W.T.; Wakeman, E.E.; Miller, S.L.; Clements, C.B.; Sun, F. The community attitudes toward sex offenders scale: The development of a psychometric assessment instrument. Res. Soc. Work Pract. 2008, 18, 251–259. [Google Scholar] [CrossRef]
  9. Chui, W.H.; Cheng, K.K.; Ong, R. Exploration of the community attitude toward sex offender scale in a Chinese cultural context. Asian J. Criminol. 2013, 9, 37–48. [Google Scholar] [CrossRef]
  10. Tewksbury, R.; Mustaine, E.E. Parole board members’ views of sex offender registration and community notification. Am. J. Crim. Justice 2012, 37, 413–431. [Google Scholar] [CrossRef]
  11. Conley, T.; Hill, K.; Church, W.T.; Stoeckel, E.; Allen, H. Assessing probation and community corrections workers’ attitudes toward sex offenders using the community attitudes toward sex offenders (CATSO) scale in a rural state. Sex. Addict. Compulsivity 2011, 18, 75–85. [Google Scholar] [CrossRef]
  12. Shackley, M.; Weiner, C.; Day, A.; Willis, G.M. Assessment of public attitudes towards sex offenders in an Australian population. Psychol. Crime Law 2014, 20, 553–572. [Google Scholar] [CrossRef]
  13. Shelton, L.; Stone, J.; Winder, B. Evaluating the factor structure and reliability of the community attitudes toward sex offenders (CATSO) scale. J. Crim. Psychol. 2013, 3, 115–126. [Google Scholar] [CrossRef]
  14. Wevodau, A.L.; Cramer, R.J.; Gemberling, T.M.; Clark, J.W. A psychometric assessment of the community attitudes toward sex offenders (CATSO) scale: Implications for public policy, trial, and research. Psychol. Public Policy Law 2016, 22, 211–220. [Google Scholar] [CrossRef]
  15. Eagly, A.H.; Chaiken, S. The Psychology of Attitudes; Harcourt: Orlando, FL, USA, 1993. [Google Scholar]
  16. Greenwald, A.G.; Banaji, M.R. Implicit social cognition: Attitudes, self-esteem, and stereotypes. Psychol. Rev. 1995, 102, 4–27. [Google Scholar] [CrossRef] [PubMed]
  17. Hogue, T.E.; Harper, C.A. Development of a 21-item short form of the Attitudes to Sexual Offenders (ATS) Scale. Law Hum. Behav. 2019, 43, 117–130. [Google Scholar] [CrossRef] [PubMed]
  18. Harper, C.A.; Hogue, T.E. Press coverage as a heuristic guide for social decision-making about sexual offenders. Psychol. Crime Law 2017, 23, 118–134. [Google Scholar] [CrossRef]
  19. Snape, N.; Fido, D. Sex offenders vs. people with sexual offences: Putting the person before the offence. J. Concurr. Disord. 2021, 4, 93–108. [Google Scholar] [CrossRef]
  20. Bernoski, M. Attitude Changes Toward People Who Committed Sex Offenses: Impact of Person-First Language, Empathy and Humanization. Ph.D. Dissertation, Indiana University of Pennsylvania, Indiana, PA, USA, 2023. PQDT 30569006. Available online: https://www.proquest.com/pqdtglobal/docview/2841619647 (accessed on 30 September 2025).
  21. Boland, C. Attitude Changes Toward “People Who Committed Sex Offenses”: Impact of Person-First Language in Context of Individual Processing Styles. Ph.D. Dissertation, Indiana University of Pennsylvania, Indiana, PA, USA, 2022. PQDT 29257306. Available online: https://www.proquest.com/pqdtglobal/docview/2703354124 (accessed on 30 September 2025).
  22. Harper, C.A.; Bartels, R.M.; Hogue, T.E. Reducing stigma and punitive attitudes toward pedophiles through narrative humanization. Sex. Abus. J. Res. Treat. 2018, 30, 533–555. [Google Scholar] [CrossRef]
  23. Friestad, C.; Mjåland, K.; Pape, H. Prison officer students’ perceptions of persons convicted of sexual crimes. Eur. J. Criminol. 2023, 20, 1269–1284. [Google Scholar] [CrossRef]
  24. Wurtele, S.K. University students’ perceptions of child sexual offenders: Impact of classroom instruction. J. Child Sex. Abus. 2018, 27, 276–291. [Google Scholar] [CrossRef]
  25. Belisle, L.A.; Marshall, E.A.; Butler, M.M. What about language?: A mixed methods examination of the impact of person-centered language on students’ perceptions of individuals who commit sex offenses. Sex Abus. 2025, 37, 503–528. [Google Scholar] [CrossRef]
  26. Harper, C.A.; Hicks, R. The effect of attitudes towards individuals with sexual convictions on professional and student risk judgments. Sex. Abus. 2022, 34, 948–972. [Google Scholar] [CrossRef]
  27. Rothwell, M.; Fido, D.; Heym, N. Perceptions around adult and child sex offenders and their rehabilitation as a function of education in forensic psychology independent of traditionalism and perpetrator sex. Forensic Sci. Int. Mind Law 2021, 2, 100039. [Google Scholar] [CrossRef]
  28. Harper, C.A.; Bartels, R.M. The influence of implicit theories and offender characteristics on judgements of sexual offenders: A moderated mediation analysis. J. Sex. Aggress. 2017, 23, 139–150. [Google Scholar] [CrossRef]
  29. Osagie, F.O., Jr.; Pica, E. Battle of the sexes: Similarities and differences in lay people’s perceptions about male and female sex offenders. Mod. Psychol. Stud. 2022, 28, 8. Available online: https://scholar.utc.edu/mps/vol28/iss1/8 (accessed on 30 September 2025).
  30. Leon, C.M.; Rollero, C. The role of ambivalent sexism, punitiveness, and ability to recognize violence in the perception of sex offenders: A gender-perspective analysis. Sexes 2021, 2, 495–508. [Google Scholar] [CrossRef]
  31. Biteus, J.; Tuiskunen, M. Attitudes Towards Sexual Offenders: A Thesis Comparing Students’ Attitudes Towards Sexual Offenders and Specific Sexual Offences. Bachelor’s Thesis, Örebro University, Örebro, Sweden, 2018. diva2:1179892. Available online: https://www.diva-portal.org/smash/get/diva2:1179892/FULLTEXT01 (accessed on 30 September 2025).
  32. Martens, J.P.; Stewart, J.L. Perceptions of sex offenders’ partners: Associated with perceptions of offenders and influenced by attachment. J. Relatsh. Res. 2020, 11, e19. [Google Scholar] [CrossRef]
  33. Wheeler, C.; Colloff, M.; Brzozowski, A. Attitudes, perceptions and workplace factors predict well-being in forensic healthcare workers caring for sex offendders. J. Psychiatr. Ment. Health Nurs. 2025, 32, 1211–1221. [Google Scholar] [CrossRef]
  34. Bakhtiari, M.; Sohrabzadeh, A.; Bazani, M.; Hosseini, A. An investigation of the validity of two self report questionnaires about attitudes and perceptions towards sex offenders. Ebnesina 2021, 23, 54–60. [Google Scholar] [CrossRef]
  35. Togas, C.; Fotini, M.; Aikaterini, S.; Maria, M. The Greek version of the perceptions of sex offenders scale. Criminol. Soc. Integr. 2022, 30, 167–187. [Google Scholar] [CrossRef]
  36. Deng, L.; Chan, W. Testing the difference between reliability coefficients alpha and omega. Educ. Psychol. Meas. 2017, 77, 185–203. [Google Scholar] [CrossRef] [PubMed]
  37. Keena, L.; Krieger-Sample, L. Empathy-focused learning: Teaching criminal justice students to care. Am. J. Crim. Justice 2017, 43, 389–410. [Google Scholar] [CrossRef]
  38. Moak, S.C.; Walker, J.T.; Earwood, M.; Towery, G. Using reentry simulations to promote changes in attitude toward offenders: Experiential learning to promote successful reentry. Am. J. Crim. Justice 2019, 45, 126–144. [Google Scholar] [CrossRef]
  39. Flora, K. The history of clinical psychology in Greece: A brief review—Legal deficiencies, practical dimensions and challenges for the future. Clin. Psychol. Eur. 2024, 6, e12515. [Google Scholar] [CrossRef]
  40. Pipini, M. Public Attitudes Towards Crime and Punishment in Greece and the Factors Underlying Their Construction. Ph.D. Dissertation, Plymouth University, Plymouth, UK, 2013. 10025742. Available online: https://pearl.plymouth.ac.uk/foahb-theses-other/412/ (accessed on 30 September 2025).
  41. Giulini, P.; Emiletti, L.; Garbarino, F.; Scotti, A.; Bradford, J.M.; Looman, J.; Saleh, F.M.; Brodsky, D.J. The treatment of sex offenders: An international perspective. In Sex Offenders: Identification, Risk Assessment, Treatment, and Legal Issues, 2nd ed.; Saleh, F.M., Bradford, J.M., Brodsky, D.J., Eds.; Oxford University Press: Oxford, UK, 2020; p. 391-C18.P185. [Google Scholar] [CrossRef]
  42. Veggi, S.; Cipresso, P.; Zara, G. Public punitiveness towards individuals with sexual convictions in Italy: A vignette study. Eur. J. Crim. Policy Res. 2024, 30, 673–688. [Google Scholar] [CrossRef]
  43. Lyons-Thomas, J. Interscale correlations. In Encyclopedia of Quality of Life and Well-Being Research; Maggino, F., Ed.; Springer Nature: Cham, Switzerland, 2020; pp. 3352–3353. [Google Scholar] [CrossRef]
  44. Akoglu, H. User’s guide to correlation coefficients. Turk. J. Emerg. Med. 2018, 18, 91–93. [Google Scholar] [CrossRef]
  45. Maruna, S.; King, A. Once a criminal, always a criminal?: ‘Redeemability’ and the psychology of punitive public attitudes. Eur. J. Crim. Policy Res. 2009, 15, 7–24. [Google Scholar] [CrossRef]
  46. Harrington, D. Confirmatory Factor Analysis; Oxford University Press: Oxford, UK, 2009. [Google Scholar]
  47. Neal, T.; Martire, K.A.; Johan, J.L.; Mathers, E.M.; Otto, R.K. The law meets psychological expertise: Eight best practices to improve forensic psychological assessment. Annu. Rev. Law Soc. Sci. 2022, 18, 169–192. [Google Scholar] [CrossRef]
  48. Hanson, R.K.; Morton-Bourgon, K.E. The accuracy of recidivism risk assessments for sexual offenders: A meta-analysis of 118 prediction studies. Psychol. Assess. 2009, 21, 1–21. [Google Scholar] [CrossRef]
  49. Wormith, J.S.; Hogg, S.; Guzzo, L. The predictive validity of a general risk/needs assessment inventory on sexual offender recidivism and an exploration of the professional override. Crim. Justice Behav. 2012, 39, 1511–1538. [Google Scholar] [CrossRef]
  50. Chevalier, C.S.; Boccaccini, M.T.; Murrie, D.C.; Varela, J.G. Static-99R reporting practices in sexually violent predator cases: Does norm selection reflect adversarial allegiance? Law Hum. Behav. 2015, 39, 209–218. [Google Scholar] [CrossRef]
  51. Helmus, L.M.; Ahmed, S.; Lee, S.C.; Olver, M.E. Cross-cultural validity of sexual recidivism risk assessments using Static-99R, STABLE-2007, and the VRS-SO. Curr. Psychiatry Rep. 2024, 26, 27–36. [Google Scholar] [CrossRef]
  52. Venner, S.; Sivasubramaniam, D.; Luebbers, S.; Shepherd, S.M. Cross-cultural reliability and rater bias in forensic risk assessment: A review of the literature. Psychol. Crime Law 2020, 27, 105–121. [Google Scholar] [CrossRef]
  53. Evans, D.N.; Cubellis, M.A. Coping with stigma: How registered sex offenders manage their public identities. Am. J. Crim. Justice 2015, 40, 593–619. [Google Scholar] [CrossRef]
  54. Furst, R.T.; Evans, D.N. An exploration of stigma in the lives of sex offenders and heroin abusers. Deviant Behav. 2014, 36, 130–145. [Google Scholar] [CrossRef]
  55. Van den Berg, J.W.; Kossakowski, J.J.; Smid, W.; Babchishin, K.M.; Borsboom, D.; Janssen, E.; van Beek, D.; Gijs, L. Dynamic risk factors in adult men who committed sexual offenses: Replication and comparison of networks found in two independent samples. Psychol. Violence 2022, 12, 424. [Google Scholar] [CrossRef]
  56. McAlinden, A.-M. Restorative justice as a response to sexual offending: Addressing the failings of current punitive approaches. Sex. Offender Treat. 2008, 3, 1–12. [Google Scholar]
  57. Oudshoorn, J.; Stutzman Amstutz, L.; Jackett, M. The Little Book of Restorative Justice for Sex Abuse; Good Books: New York, NY, USA, 2015. [Google Scholar]
  58. Rye, B.J.; Hovey, A.; Waye, L. Evaluation of a restorative justice-based, community-based program for people who have offended sexually: Participant impact. Contemp. Justice Rev. 2018, 21, 276–298. [Google Scholar] [CrossRef]
  59. Wilson, C.; Bates, A.; Völlm, B. Circles of support and accountability: An innovative approach to manage high-risk sex offenders in the community. Open Criminol. J. 2010, 3, 48–57. [Google Scholar] [CrossRef]
  60. Koss, M.P.; Bachar, K.J.; Hopkins, C.Q. Restorative justice for sexual violence. Ann. N. Y. Acad. Sci. 2003, 989, 384–396. [Google Scholar] [CrossRef]
  61. Hanson, R.K.; Harris, A.J.R.; Helmus, L.; Thornton, D. High-risk sex offenders may not be high risk forever. J. Interpers. Violence 2014, 29, 2792–2813. [Google Scholar] [CrossRef] [PubMed]
  62. Brown, M. An exploration of the challenges families experience when a family member is convicted of a sex offence. Prison Serv. J. 2017, 233, 34–41. [Google Scholar]
  63. Evans, D.; Trahan, A.; Laird, K. Shame and blame: Secondary stigma among families of convicted sex offenders. Criminol. Crim. Justice 2023, 23, 78–97. [Google Scholar] [CrossRef]
  64. Quinn, J.F.; Forsyth, C.J.; Mullen-Quinn, C. Societal reaction to sex offenders: A review of the origins and results of the myths surrounding their crimes and treatment amenability. Deviant Behav. 2004, 25, 215–232. [Google Scholar] [CrossRef]
  65. Tewksbury, R.; Mustaine, E.E. Law-enforcement officials' views of sex offender registration and community notification. Int. J. Police Sci. Manag. 2013, 15, 95–113. [Google Scholar] [CrossRef]
  66. Willis, G.M.; Levenson, J.S.; Ward, T. Desistance and attitudes towards sex offenders: Facilitation or hindrance? J. Fam. Viol. 2010, 25, 545–556. [Google Scholar] [CrossRef]
Table 1. Factor and item content of the Community Attitudes Toward Sex Offenders and Perception of Sex Offenders scales.
Table 1. Factor and item content of the Community Attitudes Toward Sex Offenders and Perception of Sex Offenders scales.
PSOCATSOFactor Title and Item ContentPSOCATSO
Item Number Factor
Factor: PSO—Stereotype Endorsement (SE),
CATSO—Social Isolation (SI)
66Sex offenders prefer to stay home alone rather than be around lots of people.SESI
77Most sex offenders do not have close friends.SESI
88Sex offenders have difficulty making friends even if they try real [sic] hard.SESI
1414Most sex offenders are unmarried men.SESI
1616Most sex offenders keep to themselves.SESI
Factor: PSO—Sentencing and Management (SM),
CATSO—Capacity to Change (CC)
1 *1 *With support and therapy, someone who committed a sexual offence can learn to change their behaviour. 1SMCC
22People who commit sex offences should lose their civil rights (e.g., voting and privacy).SMCC
1111Trying to rehabilitate a sex offender is a waste of time.SMCC
1212Sex offenders should wear tracking devices so their location can be pinpointed at any time.SMCC
1818Convicted sex offenders should never be released from prison.SMCC
3 (new)--The death penalty should be reintroduced for sex offenders.SM--
10 (new)--People who commit sex offences should be subject to harsh restrictions on their liberty for the rest of their lives. SM--
15 (new)--It’s not a matter of if a sex offender commits another crime, it’s a matter of when. SM--
17 (new)--Sex offenders should have all of their details announced to local communities. SM--
19 (new)--Sex offenders will almost always commit further offences. SM--
Factor: PSO—Risk Perception (RP),
CATSO—Severity/Dangerousness (SD)
9 *9 *The prison sentences sex offenders receive are much too long when compared to the sentence lengths for other crimes. RPSD
13 *13 *Only a few sex offenders are dangerous.RPSD
4 * (new) --People are far too on edge about the risks posed by sex offenders. RP--
5 * (new)--More sex offenders should be given sentences in the community. RP--
20 * (new)--Some sex offenders should be allowed to work in schools.RP--
deleted4 *A lot of sex offenders use their victims to create pornography (originally reported as: Male sex offenders should be punished more severely than female sex offenders).--SD
deleted15 *Someone who uses emotional control when committing a sex offence is not as bad as someone who uses physical control when committing a sex offence.--SD
deleted17 *A sex offence committed against someone the perpetrator knows is less serious than a sex offence committed against a stranger.--SD
Factor: CATSO—Deviancy (D)
deleted3People who commit sex offences want to have sex more often than the average person.--D
deleted5Sexual fondling (inappropriate unwarranted touch) is not as bad as rape.--D
deleted10Sex offenders have high rates of sexual activity.--D
Table Notes: CATSO—Community Attitudes Toward Sex Offenders Scale [8]; PSO—Perceptions of Sex Offenders scale [7]. * Item is reverse-coded. Responses are rated on a 6-point Disagree-to-Agree scale. PSO—Stereotype Endorsement (SE) and CATSO—Social Isolation (SI) are identical factors. The PSO—Sentencing and Management (SM) factor contains all of the CATSO—Capacity to Change (CC) items but adds five additional items. The PSO—Risk Perception (RP) factor is the most different, retaining two original CATSO—Severity/Dangerousness (SD), deleting three CATSO items, and adding three novel items. Harper and Hogue [7] renamed all factors presumably to avoid confusion between the earlier CATSO and PSO revision. 1 Originally, American spelling was used by Harper and Hogue [7] as well as by Church et al. [8].
Table 2. PSO reliability coefficients, study location, and sample source.
Table 2. PSO reliability coefficients, study location, and sample source.
Author(s)Region of StudySample TypeReported Reliability (α)
SMSERPTotal
Harper & Hogue [7]United KingdomGeneral Public0.930.850.810.92
Biteus & Tuiskunen [31]SwedenStudents—2 versions0.810.810.310.77
0.860.780.380.83
Harper & Hogue [18]United KingdomGeneral Public0.930.840.800.92
Harper & Bartels [28]United KingdomGeneral Public0.930.840.870.90
Harper & colleagues [22]United KingdomStudents0.920.830.65
Martens & Stewart [32]United KingdomStudents0.880.780.650.83
Leon & Rollero [30]ItalyGeneral Public and Students0.870.830.700.82
Rothwell et al. [27]United KingdomForensic Students
General Public		0.890.900.750.89
0.880.910.740.89
Snape & Fido [19]United KingdomGeneral Public0.910.780.22
Harper & Hicks [26]United KingdomStudents and Forensic Professionals0.920.840.72
Togas et al. [35]GreeceGeneral Public0.900.83 0.88
Bakhtiari et al. [34]TehranConvenience (mostly students, young professionals)0.850.610.76
Osagie & Pica [29]United StatesGeneral Public0.870.790.65
Friestad et al. [23]NorwayPrison Officer0.850.730.52
Students0.900.810.64
Table Notes: SM = Sentencing and Management; SE = Stereotype Endorsement; RP = Risk Perception; and Total = Whole Scale. Rothwell and colleagues [27] reported two sets of PSO reliability scores for adult victims (first row) and child victims (second row). Togas and colleagues [35] endorsed a two-factor model, removing Risk Perception as well as removing two items (i.e., 4 and 9). Friestad and colleagues [23] reported two sets of PSO reliability scores for before the training (first row) and after training (second row). Embedded in a within-subjects design, Biteus and Tuiskunen [31] also reported two sets of reliability scores for two types of offences: harassment (first row) and rape (second row).
Table 3. Interscale correlations reported between PSO scales.
Table 3. Interscale correlations reported between PSO scales.
Author(s)Correlation Coefficients (r)
SM–SESM–RPSE–RP
Bakhtiari et al. [34]0.580.310.80
Togas et al. [35]0.16nana
Friestad et al. [23]0.30 (T1)
<0.30 (T2)		<0.30
0.38 (T2)		<0.30
<0.30
Harper & Hicks [26]0.280.520.04
Leon & Rollero [30]0.370.14−0.33
Table Notes. SM = Sentencing and Management; SE = Stereotype Endorsement; RP = Risk Perception; T1 = Time 1 pre-test; and T2 = Time 2 post-intervention.
Table 4. Correlations of the PSO with related Measures.
Table 4. Correlations of the PSO with related Measures.
Convergent Validity ScaleAuthor(s)Correlation Coefficients (r)
SMSERPTotal
ATSHarper & Hogue [7] −0.84
ATSHarper & Bartels [28] −0.89
ATSHarper & Hicks [26]−0.84−0.23−0.61
8-item Risk Judgement Index (conceptually most similar to Risk Perception)Harper & Hicks [26]0.670.230.53
GPSHarper & Hogue [7]0.78
GPSLeon & Rollero [30]0.670.260.16
5-item Punitiveness Index (conceptually similar to Sentencing and Management)Togas et al. [35]0.680.17
6-item Stereotype Index (conceptually similar to Stereotype Endorsement) Togas et al. [35]0.610.36
Table Notes. ATS = Attitudes Toward Sex Offenders [17]. GPS = General Punitiveness Scale [45]. SM = Sentencing and Management; SE = Stereotype Endorsement; and RP = Risk Perception. Scales labelled as indices were novel measures created for the specific studies.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Rye, B.J.; Tuer, A.G. Perceptions of Sexual Offenders (PSO) Scale. Encyclopedia 2025, 5, 168. https://doi.org/10.3390/encyclopedia5040168

AMA Style

Rye BJ, Tuer AG. Perceptions of Sexual Offenders (PSO) Scale. Encyclopedia. 2025; 5(4):168. https://doi.org/10.3390/encyclopedia5040168

Chicago/Turabian Style

Rye, B. J., and Amy G. Tuer. 2025. "Perceptions of Sexual Offenders (PSO) Scale" Encyclopedia 5, no. 4: 168. https://doi.org/10.3390/encyclopedia5040168

APA Style

Rye, B. J., & Tuer, A. G. (2025). Perceptions of Sexual Offenders (PSO) Scale. Encyclopedia, 5(4), 168. https://doi.org/10.3390/encyclopedia5040168

Article Metrics

Article metric data becomes available approximately 24 hours after publication online.
Back to TopTop