Volunteering With Sex Offenders: The Attitudes of Volunteers Toward Sex Offenders, Their Treatment, and Rehabilitation

The general public has been shown to hold negative attitudes toward sexual offenders, sex offender treatment, and the rehabilitation of sexual offenders. It appears pertinent to the success of sex offender management strategies that utilise volunteers that selected volunteers do not share these attitudes. Here, volunteers for Circles of Support and Accountability (CoSA), a community-based initiative supporting the reintegration of sex offenders, completed three validated psychometric measures assessing attitudes toward sex offenders in general and toward their treatment and rehabilitation. Responses were compared with a U.K. general public sample. The results showed that volunteers held more positive attitudes toward sex offenders, sex offender treatment, and sex offender rehabilitation than the U.K. general public sample. The significance of these findings is discussed alongside directions for future research.


Introduction
There is consensus in the literature that members of the general public hold inaccurate perceptions and openly negative attitudes toward sex offenders (Levenson, Brannon, Fortney, & Baker, 2007;Willis, Levenson, & Ward, 2010), sex offender treatment 1 University of Nottingham, UK (e.g., Olver & Barlow, 2010), and sex offender rehabilitation (Payne, Tewksbury, & Mustaine, 2010). They also have a tendency to grossly overestimate recidivism rates (Brown, Deakin, & Spencer, 2008;Levenson et al., 2007;Olver & Barlow, 2010), and to view sentences as not being sufficiently severe (Olver & Barlow, 2010). However, although remaining skeptical of the efficacy of treatment and of treating sex offenders within the community (Höing, Petrina, Hare Duke, Völlm, & Vogelvang, 2016), the general public have also been shown to subscribe to the treatment and risk management of sex offenders as an alternative to imprisonment alone (e.g., Olver & Barlow, 2010). Overall, general public attitudes are likely to present barriers for sex offender rehabilitation and reintegration (Willis et al., 2010). With the public generally considering it unacceptable for a sex offender to live within their community (Brown, 1999;Brown et al., 2008), sex offenders may find it difficult to form positive relationships resulting in social isolation which has been shown to increase the risk of recidivism (R. J. Wilson, McWhinnie, & Wilson, 2008).
The basis of general public stigma toward sex offenders is unclear. For instance, in research by Payne et al. (2010), few demographic, community-level or victimisation factors were identified as being predictive of attitudes toward rehabilitating sex offenders. Other research by Pickett, Mancini, and Mears (2013) offered partial support for three theoretical models of public opinion on the social control of sex crime: the victim-oriented concerns model, the sex offender stereotypes model, and the risk-management concerns model. In brief, the findings showed that views around victim-harm, concerns of victimisation, and misperceptions of the risk of sex offenders contributed to support for punitive policies but were less predictive of pessimism of sex offender treatment. The authors contended that punitive attitudes toward sex offenders are not driven by a singular influence and highlighted the stereotype of sex offenders as being unreformable as potentially being the main motivator for the underlying hostility toward sex offenders (Pickett et al., 2013).

Differences in Attitudes
The literature on attitudes toward sex offenders has demonstrated that these attitudes vary among different social groups. For instance, front line forensic staff (psychologists and probation officers) have been found to demonstrate significantly more favorable attitudes toward sex offenders than members of the general public (Higgins & Ireland, 2009;Johnson, Hughes, & Ireland, 2007), students (Ferguson & Ireland, 2006;Gakhal & Brown, 2011;Kjelsberg & Loos, 2008), and forensic professionals who are less involved in treatment (Hogue, 1993;Hogue & Peebles, 1997;Lea, Auburn, & Kibblewhite, 1999;Johnson et al., 2007;Kjelsberg & Loos, 2008). More positive attitudes toward sex offenders are thought to be associated with more contact (Hogue, 1993;Nelson, Herlihy, & Oescher, 2002), more training (Craig, 2005;Hogue, 1993Hogue, , 1995, more confidence in working with sex offenders (Griffin & West, 2006;Hogue, 1994), and possessing more accurate information (Church, Wakeman, Miller, Clements, & Sun, 2008;Shackley, Weiner, Day, & Willis, 2014), suggesting that attitudes toward sexual offenders are not static. Notably, Lea et al. (1999) showed that the benefits of more contact with sex offenders extend to paraprofessionals, with volunteers in prison rehabilitation representing more positive attitudes toward sex offenders than prison and police officers. More training has also been shown to influence treatment-specific attitudes of professionals (Craig, 2005). In contrast to these findings, R. J.  found that more experienced volunteers in a community-based initiative, Circles of Support and Accountability (CoSA), were more skeptical about treatment success.

Volunteering With Sex Offenders
In its efforts to reduce reoffending, the criminal justice system in the United Kingdom has increasingly relied upon support from the third sector in the resettlement of offenders (Ministry of Justice, 2010. The benefits of third-sector involvement with offending populations include enhanced responsiveness, social cohesion, and facilitating the transition between prison and the community (see Meek, Gojkovic, & Mills, 2010). In this way, third-sector organisations can be considered "mediators" between the criminal justice system and the community with the volunteers of these organisations mediating between ex-offenders and community members. Consequently, it appears pertinent to the success of community reintegration strategies that rely upon volunteers that the chosen volunteers are more receptive to sex offenders and their treatment and rehabilitation than an average member of the general public (R. J. Wilson, Mcwhinnie, Picheca, Prinzo, & Cortoni, 2007).
The diversity of the volunteer personnel working with offenders compared with paid staff has been considered a benefit of the third-sector involvement in the criminal justice system (Meek et al., 2010). However, research has raised doubts on whether such volunteer workforces are truly representative of their associated communities (Clinks, 2007), and it has been contended that negative attitudes toward offenders may limit the pool of volunteers from which to select from (Clinks, 2006). Research on the profile of volunteers working with sex offenders and the impact this may have on treatment outcomes and public engagement is limited. However, in a case review, McCartan et al. (2014) reported that, of the 172 Circles South East volunteers in the United Kingdom in 2012, 74% were female, 30% reported their profession to be "student," and there was a wide range of ages of volunteers.
Attitudes toward sex offenders (e.g., Church et al., 2008) are thought to have predictive value on punitive judgments (Hogue & Peebles, 1997;Kjelsberg & Loos, 2008) and on attitudes toward sex offender treatment (Wnuk, Chapman, & Jeglic, 2006) and rehabilitation (Rogers, Hirst, & Davies, 2011). For community-based initiatives, a concern may be that inappropriate attitudes may bias volunteer judgments of risk and needs which would likely affect the integrity of the initiative and treatment outcomes. For instance, overly optimistic views that the offenders can change their behaviour may lead to the prioritisation of their needs as opposed to the safety of the community. By contrast, outwardly negative attitudes may manifest in a dismissive and rejecting interpersonal style and a failure to recognise progress (R. J. . It is therefore clear that an imbalance in attitudes can undermine the primary purpose of these kinds of initiatives. An example of a community-based approach to the management and reintegration of former sexual offenders is CoSA. Underpinned by restorative justice principles (Hannem, 2013;Nellis, 2009), the CoSA model attempts to converge both risk-and strengths-based approaches (Petrunik, 2007;R. J. Wilson et al., 2008). Hence, the underlying aim is to support the reintegration of a former sexual offender, known as a "core member," back into the community whilst also monitoring risk and holding them accountable for his or her behaviour. Under the CoSA model, core members are provided with a surrogate social network of three to six volunteers who operate under the supervision of a project coordinator  and are supported by an "outer circle" of professionals (Circles UK, 2009). Evidence for the effectiveness of the model is limited, though some emerging research suggests that it may have benefits in reducing reoffending rates (see Clarke, Brown, & Völlm, 2015, for a review). Some authors have considered the suitability of volunteers for Circles. For example, C. Wilson, Bates, and Völlm (2010), based on a review of the literature, identified the following qualities of volunteers that may contribute to an effective CoSA: appropriate motivation, healthy boundaries, adequate training and support, consistency, and a well-balanced representation of the community. The attitudes of volunteers toward sex offenders, their treatability, and rehabilitation are also important areas to consider and might help inform the recruitment of future volunteers. Indeed, a component of the Dutch CoSA selection criteria is that volunteers should demonstrate a supportive attitude toward restorative justice and social inclusion (Höing, Bogaerts, & Vogelvang, 2016).

The Present Study
The primary aim of this study was to investigate whether people who volunteer with sexual offenders hold more "positive" attitudes toward sex offenders and their treatment and rehabilitation than members of the general public. Demographic differences were explored to identify their relationship with volunteer attitudes. Although we chose to explore attitudes in volunteers for a specific program, CoSA, findings will be relevant to other initiatives that rely on volunteers working with sex offenders and may inform the recruitment of these individuals.
We hypothesised as follows: • • Hypothesis 1: Volunteers would demonstrate more "positive" attitudes toward sex offenders, the effectiveness of treating sex offenders, and the reintegration and rehabilitation of sex offenders than the general public. • • Hypothesis 2: More experienced volunteers may hold less positive views toward the effectiveness of treatment but more positive attitudes toward sex offenders in general than less experienced members.

Participants
An opportunity sample of CoSA volunteers completed an Internet-based survey. Respondents hailed from nine different Circle projects across England. Of the 76 questionnaires completed, 75 were usable. This exceeded the appropriate sample size of 68 participants based on power analyses of public  and paraprofessional samples (Jones, 2013). Of the respondents, 71 (94.7%) had been in an active circle; three quarters of the sample participated in an active circle at the time of the study (73%). The U.K. general public data (n = 210) were supplied by the main researcher from another study. The full methodology of this study is described elsewhere , but in brief the study used existing web panels to explore general public awareness and attitudes regarding sex offenders and sex offender rehabilitation in nine European countries, launched in July/August 2014 . The authors of this current report were provided with the U.K.-specific data from  as a comparison group for our CoSA volunteers. Table 1 provides an overview of the sample characteristics of both groups.

Materials
Demographic variables. Information on age, gender, highest obtained qualification, occupation, knowing a victim of a sexual offence, personal and professional experience with general and sexual offenders, and duration of experience in CoSA (where applicable) was obtained using a simple online questionnaire.

Community Attitudes Toward Sex Offenders-Revised Version (CATSO-R).
Attitudes were measured using the revised version of the CATSO (Church et al., 2008), an 18-item scale measuring lay perceptions and stereotypes of sex offenders. The CATSO-R encompasses four subscales: Social Isolation (e.g., "most sex offenders keep to themselves"), Capacity to Change (e.g., "convicted sex offenders should never be released from prison"), Severity/Dangerousness (e.g., "only a few sex offenders are dangerous"), and Deviancy (e.g., "a lot of sex offenders use their victims to create pornography"). Although some authors have raised concerns regarding the lack of support for the underlying structure of the CATSO (Conley, Hill, Church, Stoeckel, & Allen, 2011;Corabian & Hogan, 2015;Shackley et al., 2014;Shelton, Stone, & Winder, 2013), it has been reported to show good reliability with a reported coefficient of .72 for the CATSO-R (Corabian & Hogan, 2015). Furthermore, the CATSO has been increasingly used in the international literature allowing for comparisons to be made with other studies. In line with Höing et al.'s (2016) research, a neutral midpoint option of "undecided" was included in the scale to circumvent forced attitudes.  have shown that this has also improved reliability, with Cronbach's alpha reported as .81 in their study. In contrast to the original use of a 6-point Likert-type scale, items in this study were therefore rated on a 7-point Likert-type scale from 1 (strongly agree) to 7 (strongly disagree).
Attitudes Toward Sex Offender Treatment (ATTSO). The ATTSO (Wnuk et al., 2006) is a 15-item scale assessing lay attitudes toward the treatment of sex offenders. Treatmentspecific attitudes are assessed by three factors: Incapacitation (e.g., "sex offenders don't deserve another chance"), Treatment Ineffectiveness (e.g., "treatment programs for sex offenders are effective"), and Mandatory Treatment (e.g., "it is important that all sex offenders being released receive treatment"). Items are rated on a 5-point Likert-type scale from 1 (agree strongly) to 5 (disagree strongly). Harper and Hogue (2015) have argued that the scale has not been adequately validated; however, it represents a unique measurement of treatment-specific attitudes and has been used to complement other general attitude measures such as the CATSO (e.g., Church, Sun, & Li, 2011;Conley et al., 2011). Wnuk et al. (2006) reported a Cronbach's alpha of .86, and thus, the scale appears to have strong reliability.
Public Attitudes Toward Sex Offender Rehabilitation (PATSOR). The PATSOR (Rogers et al., 2011) is a 12-item measure exploring lay attitudes toward the rehabilitation and reintegration of sex offenders into society. Again, Harper and Hogue (2015) have noted issues with the validation of the scale, and we have removed the Knowing Offenders' Area of Residence subscale due to its low reliability (.60) compared with the good reliability (.86) of the Rehabilitation subscale (e.g., "sex offenders don't deserve any social support when released"). This also aligns our method with that of Höing et al.'s (2016) which only used the latter subscale. Items are rated on a 5-point Likert-type scale from 1 (strongly agree) to 5 (strongly disagree).

Procedure
Following ethical approval from Circles-UK and the University of Nottingham Facuty of Medicine and Health Sciences Research Ethics Committee, participants were invited to complete the survey hosted on the Bristol Online Survey platform. The email request was distributed by a national support officer of Circles UK to local project coordinators who were asked to forward it to volunteers. Included in the email were a study link, a completion date, and details regarding the purpose and nature of the research. Reminder emails were sent to Circles coordinators to again pass on to volunteers. An information sheet followed by a consent form was displayed for participants online before completion of the questionnaire. Consent was given by clicking on to the next page to begin the survey. Demographic information was taken first, followed by a Confidence in Individual and Organisational Attributes questionnaire (not reported here), the CATSO-R, the ATTSO, the PATSOR, and finally, a written debrief at the end of the questionnaire which signposted appropriate support services, if required. No time limit was imposed, though it was suggested that the survey would take about 25 minutes to complete. No identifiable information was recorded to ensure confidentiality and to minimise desirable responding. Volunteers were briefed to "Please think about general sex offenders and not just your own core member(s)" on all attitude measures.

Data Analysis
The Bristol Online Survey data were analysed with SPPS, Version 22. For the purposes of analysis, items were (re)coded so that higher scores on attitude scales represented more negative attitudes toward sex offenders. Item 14 on the ATTSO scale was omitted in the general public data set and thus was computed here from its counterpart on the CATSO-R (Item 18), transforming the data from a 7-point Likert-type scale to a 5-point Likert-type scale for data analysis. Education was categorised into low (general certificate of secondary education [GCSE], equivalent, and below), medium (A-level and equivalent), and high (degree and higher) levels. Age was also split into three groups (<25 years, 26-50 years, >50 years).
Descriptive results on sample characteristics were calculated for the volunteer and general public data sets. The Kolmogorov-Smirnov test revealed that the data failed to meet parametric assumptions across outcome variables. Differences between the two groups were thus tested using Pearson chi-square tests for categorical variables and Mann-Whitney U tests for continuous variables controlling for confounding variables by calculating partial correlation coefficients. Kruskal-Wallis tests were used for within-group differences for volunteers, and Mann-Whitney U tests were used for post hoc comparisons. Intercorrelations between variables were computed using Spearman's rho. Table 1 provides an overview of the demographic characteristics of CoSA volunteers and of the U.K. public sample. Mean comparisons revealed that the volunteer group was significantly more highly educated, t(166.25) = 7.58, p < .001, and had a higher mean age, t(283) = 2.45, p = .02, than the general public sample. Age distribution was also different with the volunteer group having a higher percentage of older (>50 years) and younger (<25 years) participants than the general public sample, F(2) = 9.38, p < .001. Volunteers were also significantly more likely to have known a victim of a sexual crime (inclusive of themselves), χ 2 (2) = 90.65, p < .001, to have known a sex offender, χ 2 (1) = 32.45, p < .001, and were more likely to have professional experience with sexual offenders, χ 2 (1) = 34.60, p < .001. These significant differences remained when controlling for age and gender.

Attitudes Toward Sex Offenders
There were significant differences between the volunteer and general public samples with volunteers reflecting significantly more positive attitudes toward sex offenders, sex offender treatment, and the rehabilitation of sex offenders (see Table 2). Across CATSO subscales, volunteers regarded sexual offenders as being significantly less sexually deviant and more capable of change. However, there were no significant differences between the two groups with regard to sex offenders being socially isolated or dangerous. On the ATTSO subscales, the volunteer group viewed treatment as being significantly more effective, and they were significantly less supportive of incapacitation than the general public sample. However, there were no significant differences between the groups in attitudes toward mandatory treatment. These results were maintained when controlling for education and age using partial correlation coefficients.

Relationships Between Attitudes and Volunteer Characteristics
Age category had a significant effect on volunteers' views on mandatory treatment, χ 2 (2) = 8.32, p = .02, and on the Severity/Dangerousness scale, χ 2 (2) = 8.44, p = .02. The eldest group of volunteers believed that sex offenders were more dangerous, U = 251.00, p = .006; however, they were less inclined to support mandatory treatment compared with the middle-age category, U = 254.50, p = .007. Education level had a significant effect on CATSO-R Total Score, χ 2 (2) = 6.49, p = .02, and its Deviancy subscale, χ 2 (2) = 7.93, p = .04. Post hoc paired comparisons (with α adjusted to .017) revealed that volunteers with lower qualifications (GCSE and below) held less positive attitudes toward sex offenders than those with medium, U = 8.00, p = .01, and high levels of qualifications, U = 56.50, p = .007. Moreover, lower educated volunteers thought sex offenders were more sexually deviant than those with higher levels of qualifications, U = 68.00, p = .01.
Volunteers who had any experience of sexual offending outside of Circles, through knowing a sex offender, knowing a victim, or working with a sex offender, were more pessimistic of sex offender treatment, as measured by the ATTSO, U = 281.00, p = .025. Moreover, volunteers who had worked with sexual offenders held less supportive views toward mandatory treatment than those who had no professional experience with sex offenders, U = 430.00, p = .04. Volunteers who had known a sex offender outside of CoSA thought they were more socially isolated, U = 517.00, p = .05. Conversely, knowing a victim of a sexual crime or working with general (nonsexual) offenders had no impact on attitudes. Equally, attitudes were not affected by experience levels (number of months or circles).

Bivariate Results
The CATSO-R scale was significantly correlated with both the ATTSO and the PATSOR Rehabilitation subscale which were in turn significantly correlated. Intersubscale correlations (see Table 3) were strongest between attitudes optimistic of treatment effectiveness, a sex offender's capacity to change, favoring rehabilitation and opposing incapacitation. Severity/Dangerousness and Mandatory Treatment were not correlated with any of the subscales indicating that sexual offenders were perceived as dangerous regardless of their capacity to change whereas compulsory treatment was favored independent of views on treatment efficacy.

Discussion
This research examined volunteer attitudes toward sex offenders, their treatment, and their rehabilitation in the context of attitudes in the public domain. Using CoSA volunteers as a cohort of volunteers, our findings will have wider implication for the selection of volunteers working with sex offenders. The first hypothesis of this study was supported: Volunteers held markedly more positive general attitudes toward sex offenders, sex offender treatment, and sex offender rehabilitation than a U.K. general public sample. These attitudes were interrelated, supporting previous findings (Rogers et al., 2011). Although interstudy comparisons on sex offender attitudes are problematic (see Shackley et al., 2014), more generally these findings add to the wealth of evidence that suggests that more contact (Hogue, 1993(Hogue, , 1995Nelson et al., 2002) and more experience with sexual offenders are linked with more favorable attitudes (Higgins & Ireland, 2009;Johnson et al., 2007;Lea et al., 1999). Perhaps most promising for sex offender management strategies that rely on volunteers, is that the volunteers in this sample did not share in the general public pessimism pertaining to sex offenders' rehabilitation and capacity to change Shackley et al., 2014). Representing more positive attitudes generally toward sex offenders is also likely to allow volunteers to more effectively engage sex offenders in an empathic relationship, the importance of which has been expressed by both volunteers and core members in CoSA (R. J. . Nevertheless, it was found that volunteer attitudes did not differ significantly from the general public on all subscales, and they were not always more positive. Rather, volunteers viewed sex offenders as similarly socially isolated and slightly more dangerous. Although causation cannot be presumed, it is possible that these specific attitudes were shaped by the volunteers' experiences working with sex offenders. Core members in CoSA are often selected due to a lack of social support (Höing, Vogelvang & Bogaerts,

2015)
, and child sexual abusers are a group particularly targeted for support, a group that is generally perceived to be the most dangerous (Bates, Williams, Wilson, & Wilson, 2014; R. J. . These attitudes may also be adaptive in carrying out both supportive and accountability functions. Of note, there were no correlations between Severity/Dangerousness and any other subscale, indicating that sexual offenders were perceived as dangerous regardless of their treatability. This finding appears to link volunteers' views with the "risk-management concerns" model of public opinion (Pickett et al., 2013) which is fitting given the risk-monitoring function of CoSA volunteers in the United Kingdom. However, some authors have suggested that the Severity/ Dangerousness subscale of the CATSO-R lacks face validity (e.g., Corabian & Hogan, 2015) and may therefore not accurately capture relevant attitudes. These findings may have important implications for organisations utilising volunteers in working with sexual offenders. Third-sector organisations might "work" by volunteers, who do not parallel the attitudes of the broader social context, acting as "mediators" between former sexual offenders and the wider community. There may also be certain qualities or particular groups of volunteers that demonstrate more positive attitudes and/or may be better geared toward facilitating the desistance and reintegration of ex-offenders into the community. This was explored here in the context of demographic characteristics and their relationships with attitudes.
Females were overrepresented in the sample, which is consistent with national data on CoSA volunteers (McCartan et al., 2014) and volunteers within criminal and restorative justice interventions more generally (Crawford, 2003). Advancing this, however, a more elaborate volunteer profile can be conceived which is a highly educated female with around 2 years experience, external professional experience of general offenders, and some form of experience (professional and/or personal) with sexual offenders outside of volunteering. The age distribution and employment status of volunteers were also distinct from the general public sample. The volunteer sample was made up of more older and more younger people who tended to be in employment, education, or retirement. Although it is encouraging that volunteers with these characteristics demonstrated a positive attitude toward sex offenders, this kind of profile may limit the effect of influencing general public attitudes by acting as a catalyst for such change.
At odds with predictions and previous findings, experience levels did not influence attitude ratings (Craig, 2005;Hogue, 1993;Nelson et al., 2002). Specifically, R. J.  reported that more experienced volunteers were more pessimistic about treatment success. There may have been a general shift in volunteer attitudes in parallel with observed trends in the general public, who have more recently shown support for volunteering with  and treating (Olver & Barlow, 2010) sex offenders. Confirming findings from other studies though, higher education was associated with more positive general attitudes toward sex offenders in volunteers (Church et al., 2008;Shackley et al., 2014).
Having any experience of sexual offending was related with pessimism of treatment effectiveness. This is in contrast to previous research that has generally shown the tendency of reported victims of sexual abuse, or those who have familiarity with a victim, to regard sex offenders less negatively than nonvictims (e.g., Ferguson & Ireland, 2006), and that more contact with sex offenders is associated with more positive attitudes (Hogue, 1993). Rather, it may be that more positive attitudes are associated with more positive contact with sex offenders. Similarly, this finding also adds a potential dimension to the "victim-oriented concerns" model outlined by Pickett et al. (2013) that more direct experience including knowing (or being) a victim of a sexual offence can predict more pessimistic views with regard to sex offender treatment. Of particular note was our finding that the majority of volunteers stated that they have known a victim of sexual abuse, including themselves. Although our data do not allow a breakdown into those who were a victim themselves and those who have known someone else in that situation, this finding raises a number of issues. One might speculate that having such experience might be a potential motivator that has not been described in previous published research (see C. Wilson et al., 2010, for a review). Such potential restorative justice motivation would be in line with Circles principles, though it is not possible to conclude from our research that such motivation does indeed exist. Nonetheless, based on a small interview study with CoSA volunteers, Wager and Wilson (in press) further conclude that motivation to volunteer was based on a number of considerations and was not necessarily always related to the status of victim/survivor.
Understanding the role of victims as volunteers clearly has important implications for selection, training, and safeguarding of and support for volunteers. An extreme position may be that volunteers with this background are unsuitable for working with sex offenders. Indeed, Brampton (2010) argued, in the context of Sex Offender Treatment Programmes (SOTP), individuals (professionals) with sexual victimisation should be excluded from working on SOTP on the basis of their potential vulnerability and impact upon their well-being. She also suggested that workers with this background may be more likely to sue their employer for work-related stress. Wager and Wilson (in press) challenge these assumptions and propose that such views of sexual victimisation "serve to proliferate demeaning notions of victimhood." They suggest instead to adopt a salutogenic approach which emphasises the possible attribute of survivors, including the potential roles they could take on in working with sex offenders. There also appears potential for these volunteers to influence wider societal perceptions of victims of sexual offences, and by extension, the general public support for punitive sex crime measures (e.g., Pickett et al., 2013).
It is important to continue to explore these issues with prospective volunteers and to provide the necessary support for volunteers who have been victims of a sexual offence. It may also be beneficial for projects to try to engage more male volunteers and more people less acquainted with offending behaviour to attain a more balanced representation of the community. Additional training may be warranted for the benefit of less educated volunteers who may be more susceptible to stereotypes and media sensationalism (Shackley et al., 2014) and those with less professional experience.

Limitations and Future Directions
The cross-sectional design, unrecorded response rate, and limited sample size of the study limit the conclusions that can be drawn from the data, particularly concerning the direction of causality of volunteer attitude formation. Prospective studies should look to explore this further to decipher the unique attitudes of volunteers and, through this, seek ways to address general community stigma. Considering these attitudinal differences within the theoretical models of public opinion referenced earlier may help with this endeavor as it appears that volunteer attitudes described here map onto these frameworks to some degree, namely, the "risk-management concerns" model and the "victim-oriented concerns" model (see Pickett et al., 2013). There may also be a need to establish a definition of an "appropriate" volunteer attitude, and from there determining whether volunteers demonstrate such attitudes.
Extant differences in attitudes between volunteer populations and the general population are likely to be enhanced to those reported here due to the overrepresentation of higher education in the public sample compared with the general population . Higher education has been linked with more liberal attitudes toward sex offenders (e.g., Church et al., 2008). Equally, child sexual abusers are overrepresented within Circles. This "type" of sexual offenders are often perceived to be the most dangerous and least likely to change their behaviour (Bates et al., 2014;. For intergroup comparisons, forthcoming research may profit from including a question to gauge what group of sex offender(s) participants were thinking of.
Despite these limitations, this study offers an empirical showcasing of volunteer attitudes in the context of the general public and individual differences. It is a challenge for research to establish what kind of impact volunteer factors (such as attitudes) may have on the awareness and engagement of the general public, the interpersonal experience of the sex offender, and on overall risk-management outcomes. Volunteers' interpersonal attitudes to the sex offender(s) they are working with, for example, rather than their attitudes to general sex offenders, may be more indicative of interactive behaviours and may have a greater affect on overall treatment outcomes (Harkins & Beech, 2007). Volunteer attitudes may also be amenable to change as a function of emotions, interpersonal style, group dynamics, the "type" (Payne et al., 2010), and behaviour of the sex offender, the victim, the offence, and wider organisational factors. In this sense, to fully appreciate the nuances of volunteer and public attitudes, there is a need to further consider individual-level, organisational-level, and community-level variables.

Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.