Psychology undergraduate attitudes towards male sex offenders: The impact of the victim’s gender

Abstract

There is a general aspiration for the treatment outcomes of both the sex offender and his/her victim to be fruitful. The relevance of attitudes in this context has been widely studied and there is a substantial body of empirical research pointing towards the importance of attitudes within the treatment and rehabilitation of the sex offender. Still, research that investigates the relevance of attitudes for the successful treatment outcome of the sexual assault victims is scarce. Nonetheless, this study attempts at investigating this gap in literature by measuring the attitudes of students (soon-to-be professionals) towards sex offenders in relation to an important characteristic of the victim, gender. One hundred and twenty-four participants fulfilled two measures of attitudes (PSO and ATS-21) in order to test the hypothesis and to answer the research question. Overall, the sex offenders’ perceived intentions were considered more negative when the victim was a female. No other statistically significant results were found, however on multiple subscales, the attitudes appeared somewhat more negative when the victim of the sexual offences was a female. Findings, limitations and implications are discussed.

Table of Contents

Abstract

Acknowledgements

List of abbreviations

Introduction and review of the literature

Method

Design

Participants

Materials

Case vignettes

‘Perceptions towards Sex Offenders’ (PSO) scale

‘Attitudes Towards Sex offenders’ (ATS-21) scale

‘Balanced Inventory of Desirable Responding’ (BIDR) scale

Procedure

Ethical considerations

Results

PSO Findings

ATS-21 Findings

BIDR Findings

Discussion

Limitations and future directions

Conclusion

References

Appendices

Appendix 1. Social media publication

Appendix 2. Case vignettes

Appendix 3. The Perceptions of Sex Offenders Scale (PSO) and Permission

Appendix 4 – The Attitudes Towards Sex Offenders Scale (ATS-21) and Permission

Appendix 5 – Balanced Inventory of Desirable Responding (BIDR)

Appendix 6 – Participant Information Sheet

Appendix 7 – Consent form

Appendix 8 – Online Debrief Form

Appendix 9 – Data Authenticity Form

Appendix 10 – Ethical Approval Code

Appendix 11 – SPSS Output – Assumptions of Normality

Appendix 12 – PSO Findings

Appendix 13 – ATS-21 Findings

List of abbreviations

SOA – Sexual Offences Act

CSEW – Crime Survey England and Wales

ONS – Office for National Statistics

BPS – British Psychological Society

UK – United Kingdom

TA – therapeutic alliance

GLM – Good Lives Model

RNR – Risk-Need-Responsivity model

PSO – Perceptions towards Sex Offenders scale

CATSO – Community Attitudes Towards Sex Offenders scale

S&M – Sentencing and Management (PSO subscale)

SE – Stereotype Endorsement (PSO subscale)

RP – Risk Perception (PSO subscale)

ATS-21 – Attitudes Towards Sex offenders scale (revised – 21-items)

T – Trust (ATS subscale)

I – Intent (ATS subscale)

SDATS – Social Distance (ATS subscale)

BIDR – Balanced Inventory of Desirable Responding scale

IM – Impression management (BIDR subscale)

S-D – Self-deception (BIDR subscale)

ANOVA – Analysis of variance

Introduction and review of the literature

Data from the most recent 2013/2014 Crime Survey for England and Wales (CSEW) shows that 19.9% of women and 3.6% of men, aged between 16-59 were victims once or more of sexual assault (ONS, 2015). This suggests that a significant fraction of the adult population has experienced sexual assault victimisation and often forensic psychologists, therapists and other professionals that work with sex offenders can be part of this fraction (Willmot, 2013). A sex offender is legally a generic term for all individuals convicted of crimes involving sex. This can range from sexual touching to penetration and the imprisonment time may differ. Within this study, a sex offender is a person that has been found guilty and convicted of a sexual offence under the Sexual Offences Act (SOA; 2003), chapter 42, part 3: sexual assault. For the purpose of this study it is not essential to evaluate any other type of sexual offences. Within the Sexual Offences Act, sexual assault is described as an offence where an individual (A) intentionally touches another individual (B), the touching is sexual, B does not consent to the touching and A does not reasonably believes that B consents (SOA, 2003).

There seems to be a general agreement throughout literature that sex offenders are much more resented by society in comparison to any other offender, even within the prison system (Akerstrom, 1986; Kjelsberg & Loos, 2008). Understandably, sexual assault is a term that often generates strong and negative emotions mostly due to its invasive context but more importantly, as it is considered a violation of one’s fundamental human rights (Corăbian, 2016). Empirical evidence has extensively developed ways in which the resentment the public hold towards sex offenders can be measured such as attitude and perception scales (i.e. Church, Wakeman, Miller, Clements & Sun, 2008; Hogue, 1993; Harper & Hogue, 2014b). Attitudes are most often measured through self-reports that aim to reflect the individual’s belief, feelings or behaviour towards an attitude object (Vogel & Wanke, 2016). The existent attitudes and perceptions scales are mostly developed as a Likert scale (Likert, 1932) or a similar method and thus measured on a spectrum with two opposite ends (i.e. very negative versus very positive) (Vogel & Wanke, 2016). Hogue (2009) suggests that in order for the attitudes towards sex offenders to be socially acceptable they have to fall somewhere in the middle of the spectrum. Accordingly, with the use of scales that ask the appropriate questions (items), it is possible to determine the thoughts, feelings and views that the individual has on a certain topic (Stuart-Hamilton, 2007). Some noteworthy limitations and issues might arise when attempting at measuring attitudes with the use of multi-item scales. The most common concern seems to be the participant’s awareness (‘reactivity’) of the attitude measurement and this may be a threat to the overall validity of the scale (Vogel & Wanke, 2016). In line with the British Psychological Society Code of Ethics and Conduct (BPS, 2010), participants have to be fully informed about the study they are taking part in, therefore this makes it very hard to conceal what the study is actually measuring. Not only that, but it is easy to determine if the study is asking for the opinion, thoughts, feelings and behaviours on a particular topic even with the attempt to conceal the actual purpose of the experiment. In addition, this study attempted at identifying if the participants are distorting their answers to meet the socially acceptable criteria with an added scale at the end in order to determine how much the participants may have distorted their answers. This is perhaps one solution in making sure the collected data is valid and it represents the actual attitudes of the sample. All things considered, measuring attitudes is not as effortless as it may seem and they can easily be influenced by extraneous variables.

The literature on attitudes is vast and mixed in their theoretical approaches (Albarracin, Johnson, Zanna & Kumkale, 2005; Oskamp & Schultz, 2005; Vogel & Wanke, 2016). Therefore, they can have more than one definition in the literature of social psychology, but the basic statement that most are based on is that attitudes are “a summary evaluation of an object of thought” (Vogel & Wanke, 2016, p. 2). This evaluation can be in favour or disfavour towards an attitude object (Eagly & Chaiken, 1993). As mentioned, throughout research, attitudes fall under several theoretical models. The ‘tripartite model’ (Allport, 1935) suggests that attitudes are comprised of affective, behavioural and cognitive responses. This has been challenged within the literature, Oskamp and Schultz (2005) affirm that the three responses of the model are not always related or present together, however this does not determine that the attitude is not present. Another theoretical model is the ‘file-drawer model’ and suggests that attitudes are mental files stored in memory, which one consults when performing an evaluation of an attitude object (Wilson, Lisle & Kraft, 1990). This has been challenged by another theory, the ‘attitudes-as-constructions perspective’ that concludes that individuals do not retrieve a previously filed memory but they create a new attitude based on surroundings (Wilson & Hodges, 1992). Despite the significant discrepancy in definitions and the theoretical basis of attitudes, the common ground implies that attitudes are composed of beliefs, affect and overt behaviour and they influence and interact with each other (Albarracin et al., 2005; Oskamp & Schultz, 2005). Nevertheless, attitudes have the potential to affect behaviour.

The relationship between attitudes and behaviour is complex and not always certain. A considerable amount of research has focused on this relationship extensively and it broadly recognised the relevance of attitudes in predicting behaviour (Ajzen, 2001). Although the link between attitudes and behaviour is somewhat logical, early research has failed to provide empirical evidence that successfully supports this link (Vogel & Wanke, 2016). In a literature review, Wicker (1969) found no significant relationship between attitudes and overt behaviour. Moreover, he suggested that perhaps the concept of attitudes is non-existent and they cannot justify for overt behaviour. Following research began examining the different factors that can affect and moderate the link between attitudes and behaviour (Ajzen & Fishbein, 2005; Fazio & Roskos-Ewoldsen, 2005). They have found various explanations for previous inconsistent findings of the link between attitudes such as the multidimensionality of attitudes, response bias and other moderating factors (Ajzen & Fishbein, 2005). Ultimately, they produced a set of appropriate measures guidance that can increase the finding of an attitude-behaviour link. The ‘correspondence principle’ (Ajzen & Fishbein, 1977) suggests that behavioural criteria must be included into the measure of attitude and it should follow four elements: action, target, context and time. The predictive power of attitudes can be therefore maximised if the ‘correspondence principle’ is adopted in empirical research (Ajzen & Fishbein 1977; 2005; Eagly & Chaiken, 1998). Additionally, general attitudes may provide a better prediction and explanation of a broad variety (situations and contexts) of discriminatory behaviours when compared to particularly specific attitudes (Ajzen & Fishbein, 2005).

Across literature, there have been many attempts at providing strong empirical support for the link between attitudes and behaviour together with the development of several theories. The theory of ‘reasoned action’ suggests that an individual’s intentions are the best predictors of one’s behaviour (Fishbein & Ajzen, 1975). The behavioural intention is a conscious decision to engage in certain behaviours and this is mediated by one’s attitude towards the behaviour and subjective norms (Fishbein & Ajzen, 1975). The theory of planned behaviour suggests that individual’s behaviour is in line with their intentions and perceived control over set behaviour (Ajzen, 1991). Collectively, the two theories imply that the base for behaviour are the initial intentions that in turn are influenced by an array of factors including attitudes towards that behaviour.

As noted above, attitudes can also be influenced by social interactions and this can result in participants distorting their answers to meet the desirable criteria for that interaction. However, social interactions are not the only influencer of attitudes. Attitudes can also be influenced by the media and this has been long-established, with empirical evidence suggesting that media plays a significant role in the nature of the attitudes towards sex offenders (Brown, Deakin & Spencer, 2008; Malinen, Willis & Johnston, 2014). It was found that public attitudes were being influenced by the media in most cases, especially with no direct contact with a sex offender (Brown et al., 2008; Kjelsberg & Loos, 2008).

Furthermore, in a UK based study, Brown et al. (2008) found that 90% of the participants reported to have obtained their knowledge about sex offenders from the media and they also felt that the media is somewhat exaggerating the offences in order to incite fear in society. The exaggeration in the media is most likely the generator of negative attitudes and stereotypes about this type of offender. A Canadian study asked professionals and sex offenders about their attitudes on the media’s impact on the reintegration of sex offenders. They found that there is an existent belief that the way media portrays sex offenders can negatively impact on reintegration and affect several risk factors (Corăbian & Hogan, 2012). Harper & Hogue (2017) suggest that the emotionality presented in the media may have a great impact on the way the public shapes their attitudes towards sex offenders. Moreover, the negative portrayal of sex offenders can have such a great negative impact that reintegration and desistance seem hopeless (Fox, 2015). Therefore, it is safe to conclude that the media has negatively shaped and influenced the attitudes towards sex offenders.

Although attitudes may not always be a very good predictor of future behaviour as they are easily influenced, it has been stated throughout the literature that attitudes are the basis for stigma (Goffman, 2009). Attitudes allow stigma to set in because of their evaluative (positive or negative) and subjective characteristics that happen at both a conscious and an unconscious level (Maio, Olson, Bernard & Luke, 2006). Stigma has been defined as an attribute that discredits the individual within the society, somewhat diminishing the individual “from a whole and usual person to a tainted, discounted one.” (Goffman, 1963, p.3). The stigmatisation process occurs when an individual has or is believed to have “some attribute or characteristic that conveys a social identity that is devalued in a particular social context” (Crocker, Major & Steele, 1998, p.505). Essentially, the two definitions accord that if an individual has met the criteria for a stigmatisation mark he or she is somewhat devalued in society. In the case of a sex offender, the stigmatisation mark is often linked to their offending behaviour and they are often considered to be a highly stigmatised population within the society (Evans & Cubellis, 2015; Griffin & West, 2006).

It is important to note that this stigma is relationship and context specific and it is not merely a characteristic of the person except in a social context (Major & O’Brien, 2005). Link and Phelan (2001) added that stigma is the result of factors that can correlate such as labelling, discrimination and stereotyping. The factors can also appear together and when that happens a justification is also present as to why the individual is devaluing, rejecting and excluding the stigmatised person (Link & Phelan, 2001; Ricciardelli & Moir, 2013). The identity of the stigmatised person is often reflective of society’s expectations and assumptions (the ‘virtual social identity’) and it reshapes how the stigmatised individual is seen (Goffman, 1963). The negative evaluations and stereotypes are generally well known and practised in a society (Crocker et al., 1998; Steele, 1997), thus the stigmatised individual is faced with stereotypes, labelling and discriminations across most social interactions and is often excluded and avoided (Major & Eccleston, 2004). Moreover, the literature suggests that the negative information that one holds of another has a greater impact on the general evaluation than positive information in a comparable situation (Ajzen, 2001). Nevertheless, the dehumanisation of the sex offender can affect their rehabilitation (Viki, Fullerton, Raggett, Tait & Wiltshire, 2012).

In order to provide ‘best practice’, the work of professionals working with sex offenders has to be “grounded in theoretical understanding” (Willmot, 2013, p.172). Willmot (2013) also suggests that the work of forensic psychologists should be respectful, boundaried and compassionate whilst acknowledging that this sometimes might be difficult. It is also safe the assume that the work of professionals should resemble a stigma-free approach. Willmot (2013) also points out to the importance of a good therapeutic alliance (TA) between the therapist and the sex offender. An investigation into the factors that are most likely to influence the client outcome, the ‘common factors’ were found to have the highest correlation with a positive therapy outcome than any specialised interventions (Lambert & Barley, 2001). Part of the ‘common factors’ and perhaps the most studied are the ‘person-centred facilitative conditions’ and the TA (Lambert & Barley, 2001). TA is plainly defined as the relationship between a client and their therapist or the outcome between the therapist’s approach and the client’s perception of the approach and the therapist (Marshall et al., 2003). TA was found to be at the centre of a positive treatment outcome in sex offender rehabilitation (Ross, Polaschek & Ward, 2008). Although little research has investigated the impact that negative attitudes can have on the successful rehabilitation of the sex offender, it is safe to assume that they have the potential to damage the TA. For example, empirical evidence suggests that the prison atmosphere (therapeutic or not) together with the attitudes of the professionals can negatively affect how stable the treatment is delivered and inevitably the treatment outcome (Schalast, Redies, Collins, Stacey & Howells, 2008; Ward, Day, Howells & Birgden, 2004). When working with sex offenders, the TA should also incorporate the Six Central Themes of Desistance (Farrall & McNeill, 2010). The negative attitudes that the therapist has towards the client could affect the applicability of the six themes of desistance in practice. For example, the second theme is referring to the development and maintenance of hope and motivation (Farrall & Calverley, 2006), however negative attitudes can hinder the development of hope and motivation and so on. Arguably, if the six themes of desistance are not applied in practice, this could lead to reoffending. The six themes of desistance are somewhat the basis to a strength-based approach such as the Good Lives Model (GLM; Ward & Steward, 2003) which was found to be successful in promoting desistance and lowering the risk of reoffending in sex offenders (Ward & Laws, 2010). In addition, the hope and motivation of the sex offender may also be hindered by the attitudes of the therapist. Before the GLM, practitioners were using a ‘What works’ approach with the application of a ‘Risk-Need-Responsivity’ (RNR; Andrews, Bonta & Hoge, 1990) that mostly suggested in targeting the offenders’ criminogenic needs. The main difference between the two approaches is in the focus of the treatment and arguably, another difference is the empirical views on the topic (see Andrews, Bonta & Wormith, 2011 and Ward, Yates & Willis, 2012). The RNR focuses on the weaknesses of the offender (i.e. criminogenic needs), whereas the GLM accentuates the strengths (i.e. primary goods). Thus, the criminogenic needs approach was criticised by empirical evidence suggesting that the identification of risk factors and subsequent treatment solely to reduce the level of risk may not necessarily achieve desistance (Ward, Mann & Gannon, 2007). Nevertheless, regardless of the approach taken in treatment, the focus should primarily be on the six themes of desistance due to their strong empirical basis and their emphasis on TA. If stigma endangers the TA connection, several other issues can arise, such as the loss of motivation to engage with the treatment programme (McMurran, 2003). The motivation and engagement of the offender, or rather the lack thereof is the greatest predictor of recidivism (Levenson & Macgowan, 2004).

As previously mentioned, attitudes are widely measure on a spectrum ranging between too negative or too positive. Throughout his work, Hogue (2009) has examined the impact of both negative and positive attitudes on the development of TA in the treatment of sex offenders. He suggests that negative attitudes allow a punitive and/or dismissive approach, which can significantly weaken the TA (Hogue, 2009). Although they are widely used in practice, confrontational approaches can discourage the offenders from taking responsibility for their actions and disempower them from changing their behaviour (Kear-Colwell & Pollock, 1997). Forcing them to accept the label of a sex offender can also hurt the TA, inhibit their engagement in treatment, and incite resistance in the client (Kear-Colwell & Pollock, 1997). Another significant issue that can arise as a result to negative attitudes for the sex offender is the failure to reintegrate into communities. For example, Levenson and Hern (2007) found that sex offenders could often be denied rent property and employment, which unavoidably can debilitate the successful reintegration. Subsequently, the offender can return to risky behaviours and eventually re-offend (Hanson & Bussiere, 1998). On the other side of the spectrum, when the attitudes are too positive towards the client may unlock different problems, which can also be regarded as negative. The most obvious effect that positive attitudes can have on the TA are boundary issues (Smith & Fitzpatrick, 1995). In this case, the healthy TA cannot be maintained which can also negatively affect the treatment outcome (Beech & Hamilton-Giachritsis, 2005). Moreover, overly positive attitudes can be viewed as a perceived normalisation of cognitive distortions and an acceptance of the offending behaviour (Hall & Hirschman, 1991) which arguably is inversely correlated with desistance. Positive attitudes towards the sex offender can also predispose the therapist to the vulnerability of being victimised (Farrenkopf, 1992). Overly positive attitudes somewhat disregard the victim and can influence the objective assessment of clinical change (Hogue, 2009). Across previous research, the positive attitudes towards sex offenders varies in regards to the participant’s gender. A significant amount of empirical evidence found women’s attitudes to be more positive (Ferguson & Ireland, 2006; Green, 2004; Radley, 2001) or males with more positive attitudes (Smith, 2008 cited in Hogue, 2009). Other studies found no difference between males and females (Hogue, 1993; 1995; Johnson, Hughes & Ireland, 2007). However, none of the above mentioned studies explored the idea of ‘relateability’ between the participant’s gender and the gender of the victim portrayed in the sexual offender presented to them.

The attitudes of psychology students are important and relevant as they are more likely than other graduates to be involved in the rehabilitation of the sex offender or the victim (Harper, 2012). The public attitudes towards sex offenders have been widely researched in different settings and under different contexts, however no research has been found that investigated the attitudes in relation to the victim. Data from the 2013/2014 CSEW (ONS, 2015) also shows that the victim can be predisposed to stigma perhaps as much as the sex offender can. One in eleven people thought that the victim was ‘completely’ or ‘mostly’ responsible for a sexual assault or rape by someone if they have been heavily flirting beforehand (9%), when they were under drug influence (8%) and while they were drunk (6%). Although the results are indicative of the members of the general population, professionals are part of the general population just as any individual and they are too predisposed to showing negative attitudes towards the victims to sexual offences. The literature around the impact of negative attitudes on the treatment outcomes of the victims of sexual violence is limited. The existent research in this area has mostly investigated the attitudes towards rape victims (Barber, 1974; Ward, 1988), suggesting that the attitudes towards victims of sexual violence are important as they can affect the quality of victim care. Therefore, it is argued that the victim is important as it can influence the attitudes towards sex offenders. For example, Weekes, Pelletier and Beaudette (1995) found that the sex offenders that victimise women and children are viewed as more immoral and ‘mentally ill’ than other sex offenders due to their victim’s perceived vulnerability factor. This is important as the treatment outcome and the desired desistance is hindered by the very important factors that can be somewhat controlled (influenced by attitudes) and perhaps the characteristics of the victim is just another factor. Therefore, this study is set out to examine what is the potential impact of victim’s gender on the general attitudes towards sex offenders as measured in a university sample. In line with previous research, the hypotheses used to guide this study are:

H1 – The attitudes towards sex offenders will be significantly more negative when the victim is a female.

H2 – The attitudes towards sex offenders will be significantly more positive when the victim is a male.

The predictions will be investigated in order to answer the following research questions:

Does the victim’s gender affect (positively or negatively) the general attitudes towards sex offenders? Does the gender of the participant reflect the attitudes?

Method

Design

The research follows a repeated-measures design where each participant covered both male offender – male victim and male offender – female victim conditions. Student attitudes (dependent variable) are measured in both conditions. The independent variables in this study are the gender (male or female) of the victim and the gender of the participant (male or female).

Participants

The participants are part of an opportunity sample recruited with the use of an advertisement published on social media websites (e.g. Facebook; see Appendix 1) asking all students that were enrolled at university studying a degree in Clinical, Counselling or Forensic Psychology for their willingness to take part. Some participants have also been recruited with the use of a publication on the Psychological Research on the Net website for further recruiting in an attempt to meet the desired number of participants.

One hundred and seventy-five participants have taken part, with N=124 completing all questionnaires. Out of the final sample, 87% are females (N=108) and 13% males (N=16), N=41 claim to study for a Forensic degree, N=47 a Clinical degree and N=36 a Counselling degree at an UK University across all three levels (see Table 1). Participants age ranges between 18 and 51 (M=21.94, SD=4.63).

Part of the departmental research participation scheme in place at the University of Worcester, first and second year students enrolled at this university received 1 credit (1=15 minutes) that can be used for their final year research project. No other participants received incentives. None of the participants were informed about the hypotheses of the study.