Student/trainee-professional implicit theories of paedophilia

Abstract This research investigates student/trainee-professional implicit theories of paedophilia and their explanations for the current crisis of paedophilia. The research methodology is based on a previous research design (Sternberg, Journal of Personality and Social Psychology, 49, 607–627, 1985), and consists of two interrelated studies. Study 1 (n=60) develops scales for student/trainee-professional implicit theories of paedophilia and Study 2 (n=188) uses these scales in conjunction with other scales to define the student/trainee-professional implicit theories of paedophilia and to measure the potential explanations for the current crisis of paedophilia. The findings indicate that student/trainee-professional implicit theories of paedophilic personalities (pathology, abusive and mentally disordered old offender) and behaviours (devious, child focused, cognitive distortions, secretive, child lovers, sexual abusers, socially inept and grooming) seem to reflect both previous research and expert opinion. These tentative findings suggest that student/trainee-professionals' implicit theories of paedophilia, and as such the current crisis of paedophilia in modern society, seem to have developed in part from the transfer of expert knowledge (explicit theories), the impact of the media, and social constructionalism; and not from the student/trainee-professionals' individual personality and coping traits.

Previous research, particularly in psychology, with lay people has focused on their understandings of paedophilia (McCartan, 2004), and has not addressed the possible explanations for the current crisis of paedophilia in modern society. However, other social sciences (including anthropology, criminology, sociology and media studies) have investigated possible explanations for the rise of social risks and moral panics in modern society (Beck, 1992;Cohen, 2003;Furedi, 2002;Goode & Ben-Yehuda, 1994;Hier, 2003;Scourfield, & Walsh, 2003;Thompson, 1998), which they believe can be used to explain the current crisis of paedophilia (Cohen, 2003;Furedi, 2002;Scourfield & Walsh, 2003). Social science research indicates that lay people's perceptions and understandings of social issues may be influenced by professionals, though the 'double-hermeneutic' model of reflexive modernity (Giddens, 1991). The 'double-hermeneutic' model of reflexive modernity indicates that as society modernizes professional knowledge filters down to the public, therefore increasing public understanding and resulting in the public becoming their own 'experts'. As such reinforcing the socially constructed nature of modern society; especially in regard to social concerns, like paedophilia (Scott, Jackson, & Backett-Milburn, 1998). The 'double-hermeneutic' model of reflexive modernity also reinforces the link between explicit and implicit theories in society (Furnham, 1984), which may happen through the media, as the media plays a central role in modern society (Giddens, 1991;McQuail, 2007), especially in regard to the formation and maintenance of public perceptions as well as societal attitudes (Greer, 2002). Therefore, it seems likely that both professional opinion and media influence, via the 'double-hermeneutic' model, may build on and contribute to existing notions of risk in modern society (Beck, 1992;Furedi, 2002;Scourfield & Walsh, 2003), reinforcing existing fears and anxieties (Goode & Ben-Yehuda, 1994), as such creating the current moral panic surrounding paedophilia (Cohen, 2002;McAlinden, 2006a;McDonald Wilson Bradford, 2000;West, 2000). Moral panics seem to strengthen group norms and stereotypes (Cohen, 2003;Hier, 2003), suggesting they may be acting as a coping mechanism in modern society, similar to Terror Management Theory (TMT) (Pyszczynski, Greenberg, & Solomon, 1999), for the ambiguity and risk surrounding social issues. The current research will therefore investigate the social science explanations of the current crisis of paedophilia (i.e. the media, risk, moral panics and coping strategies) through people's implicit theories of paedophilia because implicit theories are tied to ideas surrounding social constructionalism as they are learnt through social (DeFrates-Densch, Smith, Schrader, & Rique, 2004;Ward & Keenan, 1999) and psychological processes (Anderson, 1995). Hence, by understanding the development of implicit theories we can better understand socially sensitive issues, such as paedophilia, because they help to shape our perceptions of other people and groups in society (Bruner & Tagiuri, 1954;Leyens, 1983), and can result in and reinforce stereotyping and prejudice (Levy, 1999). Implicit theory research has had a long history in psychology (Furnham, 1988;Wegner & Vallacher, 1977) focusing on areas such as intelligence (Sternberg, Conway, Ketron, & Bernstein 1981;Sternberg, 1985;Spinath, Spinath, Riemann, & Angleitner, 2003), personality (Wegner & Vallacher, 1977), abnormal psychology (Furnham, 1984;Furnham & Bower, 1992;Furnham & Kuyhen, 1991;Furnham & Rees, 1988;Wegner & Vallacher, 1977), as well as in areas of addiction (Furnham & Lowick, 1984) and crime (Furnham & Henderson, 1983;Mihailides, Devilly, & Ward, 2004;Spinath et al., 2003). Recently there has been an expansion of implicit theory research into the area of sex crime, but this is a growing area particularly in regard to paedophilia. Previous implicit theory research on sex crime has examined lay perceptions of treatment (Fleur & Furnham, 1998), the self-perception of rapists (Polaschek & Gannon, 2004;Polaschek & Ward, 2002) and paedophiles (Marziano, Ward, Beech, & Pattison, 2006); as well as some theoretical work on the implicit theories of sex offenders (Mihailides et al., 2004;Ward, 2000) and child molesters (Ward & Keenan, 1999). Despite this recent expansion more research needs to be done in regard to socially sensitive issues and topics of social concern (Sparks, Girling, & Loader, 2001), like paedophilia, particularly in regard to the social constructionist nature of both implicit theories and social risks (Sparks et al., 2001).
Implicit theory research suggests that implicit theories are those that exist within people's minds; they can exist unconsciously and are related to all aspects of life (Sternberg, 1985). Implicit theories are developed and learnt though individual's interactions with the world (DeFrates-Densch et al., 2004;Ward & Keenan, 1999) and as such they are often based in common-sense perceptions and understandings of the world (DeFrates-Densch et al., 2004). However, because implicit theories are based on personal experiences and constructions, they can be incorrect (stereotypes) (Levi, Stroessner, & Dweck, 1998;Plaks, Stroessner, Dweck, & Sherman, 2001) or maladaptive (cognitive distortions) (DeFrates-Densch et al., 2004;Ward & Keenan, 1999). Therefore, implicit theories can be affected by individual differences, through both situational exposure and personality traits (Anderson, 1995). This highlights the possible impact of both ambiguous professional knowledge and the media's representation of paedophilia, on the development and maintenance of lay people's implicit theories of paedophilia. This highlights the role of the 'double-hermeneutic' model of reflexive modernization in the creation and maintenance of the crisis of paedophilia in modern society. Particularly, as implicit theories are central in regard to social perception (Chiu, Hong, & Dweck, 1997), and stereotype maintenance (Levi et al., 1998;Plaks et al., 2001); therefore affecting how individuals develop and maintain their understandings of themselves, other people as well as the surrounding world (McConnell, 2001;Spinath et al., 2003). Hence, people's implicit theories of paedophilia, whether they be lay people or professionals, are important for they play a central role in the social construction of, and societal reaction to, paedophilia.
As implicit theories exist within people's minds, they are therefore discovered through research (Chiu et al., 1997;Sternberg et al., 1981;Sternberg, 1985). This current research will focus on a student/trainee-professional (i.e. students at university studying to work within the Criminal Justice System or in any capacity with offenders) sample rather than a lay sample because of potential ethical (British Psychological Society, 2000) and sampling (Robson, 2002) issues. Although, it is important to recognize that this participant population may have a greater insight to, and understanding of, paedophilia than the general public because of the specialist knowledge gained through studying for their respective degrees, said participants are still affected by the same social processes (i.e. the media) and cultural attitudes (i.e.stereotypes and bias) towards paedophilia as the general public. The sampled population is also interesting to study in its own right as they are not fully trained professionals (i.e. people who are trained experts and work consistently in the field of paedophilic research and practiceÁtherapists, members of probation, police officers, etc.) meaning that they are still developing their knowledge base in regard to paedophilia. Consequentially, the sample are able to provide a viable mechanism to study the relationship between implicit and explicit theories, the 'double-hermeneutic' model of modern reflexivity and the role of the media in idea formation in regard to paedophilia. The student/trainee-professionals' implicit theories of paedophilia will be uncovered by using a self-report methodology (Furnham, 1988;Furnham & Henderson, 1983;Marziano et al., 2006;Polaschek & Gannon, 2004;Sternberg, 1985) whereby the participants will be asked to describe their understandings of paedophilia (Sternberg, 1985). This research methodology is based on an approach originally used by Sternberg (1985), to uncover implicit theories ofcreativity, wisdom and intelligence. The Sternberg methodology is such that each of the individual research studies builds on one another, to produce a cohesive and reliable construction of people's implicit theories; which is done through the participants providing their own definitions and understandings, with minimal input from the researcher. Study 1 asks two open-ended questions allowing the participants to give their implicit theories of paedophilia, the most common of which will be built upon in Study 2 (those implicit theories that were mentioned by three or more individual participants) (Sternberg, 1985). Study 2 also examines the construction of the participants' implicit theories of paedophilia by addressing a number of issues proposed in previous research, namely the participants' engagement with and attitudes towards the media, as well as the influence of certain personality/coping traits. Consequentially, this research hopes to uncover what the participants' implicit theories of paedophilia are and some of the possible explanations for the formation of these, so to help us better understand the current social outcry surrounding paedophilia in modern society. As this research is exploratory it does not have any formal hypotheses, but rather has a series of research questions. As such, what are student/trainee-professionals' implicit theories of paedophilic personalities and behaviours? What are the student/trainee-professionals' attitudes towards the media, especially what is the relationship between student/trainee-professionals' implicit theories of paedophilia and their attitude towards the media? Is there a relationship between student/ trainee-professionals' personality traits and their implicit theories of paedophilia?

Participants
This study had 51 participants; all of who were Criminology postgraduates from the University of Leicester. There was a nearly even split between the gender of the participants, with 23 being male and 28 being female; and an age range of between 21 and 58 years.

Materials
This research was based on previous implicit theory research (Sternberg, 1985) and used an accepted methodology (Furnham, 1988;Furnham & Henderson, 1983;Marziano et al., 2006;Polaschek & Gannon, 2004;Sternberg, 1985). The questionnaire first asked for demographic details (age and gender), and proceeded to ask two open-ended questions; 'What is a paedophile?' and 'What attitudes and behaviours do paedophiles typically display?'.

Procedure
The researcher contacted the University of Leicester's Criminology department and asked their permission to approach their students, to which the department consented and an optimal date as well as a method of sampling (an opportunity sample; Robson, 2002) was agreed upon. The research took part at the end of a lecture given to the whole year group. Prior to the commencement of the study the researcher addressed all the participants, providing them with complete disclosure in regard to the aims and objectives of the study, their rights as participants (i.e. anonymity, confidentially, the right to withdraw, etc.) as well as informing the participants of how their data would be used (British Psychological Society, 2000). The researcher then distributed the questionnaires and the participants were told to commence filling in the questionnaires after reading the preliminary information. Although there was not a strict time frame for the completion of the questionnaire, the majority of participants completed it within 10 minutes. Post completion the researcher offered the participants a short question-and-answer session specifically on the questionnaire and their research in general. No one withdrew from the study at any point.

Data analysis
All the responses were manually collated and any responses (i.e. the same or very similar phases) that appeared three or more times across all participants, for both questions, were deemed to be important and categorized accordingly (Sternberg, 1985).

Results
All the participants' responses to the two questions asked in this part of the study were collated, with any items that appeared three or more times being selected for Study 2 (Tables 1 and 2), as in accordance with previous research (Sternberg, 1985).

Discussion
The findings suggest that a selection of the student/trainee-professionals understand typical paedophilic personality traits to include the following: that they enjoy sexually abusing children, that they are not 'normal' (that they are sick, mad, mentally disordered and not normal); that they are sexually deviant (perverts, rapists); that they mainly focus their abuse on children (feelings for children, they gain pleasure from their interactions with children and they have an unhealthy attitude towards children). Although, some of the student/trainee-professionals suggested that paedophiles do not necessarily have to offend against children (they can use child sexual abuse imagery and they may not offend against children), this was a small cross-section of the participants. These findings correspond with previous research on public (McCartan, 2004) and professional (Howitt, 1998;Silverman & Wilson, 2002) understandings of paedophilic personality traits. The student/trainee-professionals also implied that paedophiles were more likely to be males (Finklehor & Russell, 1984), with only a minority (less than three) believing that females could be paedophilic offenders as well (Howitt, 1995;Saradjian, 1996); however as previous research indicates this is not necessarily the case (Howitt, 1995;La Fontaine, 1990).
Student/trainee-professionals indicated that typical paedophilic behaviours might include that paedophiles were hard to identify (secretive, appearing normal) and are therefore devious (false, liars, manipulative, underhand). A selection of the participants believed that paedophiles were not 'normal' (that they are sick, mad, they have psychological problems, anti-society's norms), suffered from emotional issues (no remorse, impulsive, immature, emotional attitude) and cognitive distortions (believe that children are interested in them sexually, believe that they love the children, blame victim, find their abusive behaviour acceptable). The participants also suggested that paedophiles are dangerous (sexually assault children, predatory, rape). Certain student/trainee-professionals also believed that paedophiles surround themselves with, and interact mainly with, children (spend time near kids, friendly with children, interest in children, sexual interest in children, preference for types of children, friendly with children, gentle with children). Although, a few of the participants indicated that paedophiles have poor social skills are socially inept (socially inept, loners) (Blanchette & Coleman, 2002;Wilson & Cox, 1983), they also considered paedophiles to be capable of complex grooming behaviour (groom family, groom child) (Erikson, Walberg, & Seeley, 1988;McAlinden, 2006b). Study 2: student/trainee-professional implicit theories of paedophilia and potential explanations for the current crisis

Participants
There were 186 participants, all of who were undergraduate students from the University of Leicester, comprising a variety of different schools (Criminology, n 043; Law, n094; Psychology, n051). The students' age ranged from 18 to 40, with a majority in the 18Á21 age range (76%), as well as being female (77%).

Materials
The questionnaire used in this study was multifaceted and developed from a number of different sources, previous research and the literature, with an aim to better clarify the participants implicit theories of paedophilia, some of the possible explanations of the underlying causes of these beliefs and therefore to help us better understand perceptions of paedophilia in modern society. The participants were first asked for their demographic information. This was followed by a 20-item Implicit Theories of Paedophilic Personality Scale (Cronbach's alpha, r00.82) and a 29-item Implicit Theories of Paedophilic Behaviour scale (Cronbach's Alpha, r00.84); both of which were developed from the most common responses to Study 1 ('what is a paedophile?'and 'what are typical paedophilic attitudes and behaviours?', respectively) (Sternberg, 1985) and used Likert scales (1 (Strongly agree) to 7 (Strongly disagree)). The participants then completed an Attitude towards the Media Scale, which was derived from ideas present in the literature (Howitt, 1995(Howitt, , 1998Silverman & Wilson, 2002;Thomas, 2005). This scale was combination of open-ended questions and Likert scales (ranging from 1 (Every day) to 5 (Never)) on media consumption and exposure, in conjunction with a series of Likert scale on attitudes towards the media (ranging from 1 (Strongly agree) to 7 (Strongly disagree)).
The participants then completed a series of Participant Personality/Coping Trait Scales. These scales were selected in parlance the existing literature (Cohen, 2003;Furedi, 2002;Hier, 2003;Howitt, 1995;Leberg, 1997;McAlinden, 2006a;Silverman & Wilson, 2002;Scourfield & Walsh, 2003;Taylor & Quayle, 2003), and aimed to help us better understand the individual differences, and therefore personality traits (Anderson, 1995), thought to influence the creation and maintenance of implicit theories surrounding paedophilia. However, certain of the personality/coping traits being investigated (i.e. the use of moral panics as a coping strategy, notions of risk and paedophilia, emotional responses to paedophilia) do not have existing scales, which means that these personality/coping traits will be linked, as closely as possible, to existing psychometric scales. A comprehensive literature review was then carried out, in conjunction with discussions with the relevant colleagues, by the researcher whereby they attempted to locate personality/coping scales that matched the theoretical positions from the literature, as much as possible. This was a drawn out process starting with the researcher identifying the main personality aspects of each theoretical approach (i.e. for moral panics and ideas of risk it was decided to look at personality traits relating to cautiousness, tolerance, anxiety and coping), which was followed by a number of potential scales being identified (i.e. The Intolerance of Uncertainty Scale, Buhr & Dugas, 2002; The Revised NEO Personality Inventory (NEO-PI-R), Costa & McCrae, 1992;Jackson Personality InventoryÁRevised, Jackson, 1994; The Abridged Big Five-Dimensional Circumplex (AB5C), Hofstee, de Raad, & Goldberg, 1992) and then narrowed down to the most appropriate ones (Intolerance of Uncertainty, Buhr & Dugas, 2002;Anxiety, John & Srivastava, 1999;Cautiousness, John & Srivastava, 1999). This approach continued until the researcher had assembled a complete, and appropriate, series of scales to measure the personality/coping traits highlighted by the literature which play a central role in the development of personal as well as social perceptions, and therefore implicit theories, of paedophilia (Cohen, 2003;Hier, 2003;Howitt, 1995;Furedi, 2002;Leberg, 1997;McAlinden, 2006a;Scourfield & Walsh, 2003;Silverman & Wilson, 2002;Taylor & Quayle, 2003). The current study has paired the following scales and theoretical positions: moral panic and risk (Anxiety, John & Srivastava, 1999;Cautiousness, John & Srivastava, 1999; Intolerance of Uncertainty, Buhr & Dugas, 2002); group stereotyping especially in regard to in-group/out-group prejudice (Social Complexity, Wann & Hamlet, 1994); social and cultural norms (Public Self Consciousness, Buss, 1980); emotional (Stability, Hogan & Hogan, 2002) and fixed attitudes (Cognitive Flexibility, Martin & Rubin, 1995; Judgemental/Openmindedness, Peterson & Seligman, 2004); as well as coping skills (Functional Dimensions of Coping, Ferguson & Cox, 1997). The personality scales where measured on a series of Likert scales, the range of which changed depending on the scale in question and all the scales have strong levels of reliability with Cronbach's alphas ranging from r00.73 to r 00.88.

Procedure
This study used University of Leicester undergraduate students, from three different departments (Criminology, Law, and Psychology) with the respective departments' permission. The same procedure as used in Study 1 was used to sample and test both Criminology and Law students; the only exception being that the questionnaire generally took 20 minutes to complete. However, a different procedure was used in regard to Psychology students, as Psychology participants received course credit for completing the study and, as such, the sampling process was initiated and ran through the Experimental Participation Requirement (EPR) system. The researcher placed an overview and details of the study on the online EPR system, with the students being able to pick up the study from a convenient central location, complete it and return it; they were informed that the study should take no more than 20 minutes to complete. They were also provided with a full de-briefing sheet explaining the nature of the study and indicating the ethical issues involved. No one withdrew from the study at any point.

Data analysis
The results will be analysed by doing a series of factor analysis, one for each of the four main sections of the questionnaire (implicit theories of paedophilic behaviour, implicit theories of paedophilic personalities, attitudes to the media, and personality/ coping traits). Each factor will be shaped by and labelled in accordance with its most salient item loadings (i.e. those with a loading of 0.44 or higher) (Maltby et al., 2004). The author recognizes that the process of deriving and labelling coherent/ meaningful factors from a factor analysis can be quite subjective; in light of this when each factor was established a series of possible labels were developed, analysed and discussed with other academic colleagues, until the most coherent and applicable label was agreed upon. After the factors for each factor grouping (i.e. implicit theories of paedophilic personality) have been determined a series of higher order correlations will be done across the different factor groupings to determine if there are possible theoretical overlaps (Maltby et al., 2004).

Results and discussion
Implicit theories of paedophilic personality Results. Prior to running this factor analysis the 'Paedophiles are male' and 'Paedophiles are female' items were removed from the analysis, as these were biological traits, not personality traits. Hence, the scale measuring the student/ trainee-professionals implicit theories of paedophilic personality traits consisted of 17 items. Preliminary analysis indicated that the data was suitable for a principal components analysis (PCA), with the KaiserÁMeyerÁOlkin measure being 0.84 (Pallant, 2007) and the Bartlett's test of Sphericity reaching statistical significance (pB0.01) (Pallant, 2007). The PCA revealed the presence of four components with eigenvalues exceeding 1, explaining 32.7%, 10.1%, 7.7% and 4.2% of the variance, respectively. The inspection of the Scree plot indicated a massive drop between the first and second component, with the first three being 1 eigenvalue or above, and a slight levelling off after the fourth using Catell's (1966) Scree test criteria; it was decided to use only three components for further analysis. Although this contradicted the results of the Parallel Analysis, which showed only two components, an Oblimin rotation was carried out and revealed that the two components were not completely separate or easily defined. The rotation solution (Table 3) indicated that three components were better separated and easier to define; with the three component solution explained 50.6% of the variance; with component 1 explaining 32.7%, component 2 explaining 10.1% and component 3 explaining 7.7%. In deciding the strength of the components a cut-off point of 0.44 was adhered to for the various items (Maltby et al., 2004); the reason for this being that it would reveal more powerful and better-focused components. However not all the items loaded on one of the three factors; and a four component version was run to see if the remaining factors would load on the extra component, however this did not add anything to the mix instead resulting in an increase in double-item loadings. It was decided to stay with the three-factor model, with component 1 representing a 'pathological' factor, with component 2 displaying an 'abusive' factor and component 3 displaying a 'mentally disturbed, old sex offender' factor. As part of the factor analysis a Pearson product moment correlation (Table 7) was produced indicating a medium positive correlation between the pathological and devious factors (r 00.42, n0172, pB0.01) and a medium negative correlation between the pathological and the mentally disordered old offender factors (r 0(0.37, n0181, p B0.01).
Discussion. The emergent factors from the student/trainee-professionals' implicit theories of paedophilic personalities (sexual degenerate, abusive and pathological sex offender) fit with previous research (La Fontaine, 1990;Leberg, 1997;Howitt, 1998;McCartan, 2004;Silverman & Wilson, 2002). Although, interestingly the student/ trainee-professionals believe that a paedophilia is pathological sex offender, whereas previous research indicates that there is no consensus on this (van Dam, 2001), with paedophilia being depicted as a mental illness (Bagley, Wood & Young, 1994), a mental deficiency (Blanchard et al., 1999), an abnormality in development (Lee, Jackson, & Ward, 2002), which is reinforced by the Diagnostic and Statistical Manual of Mental Disorders (DSM) changing its definition with every publication condition (American Psychiatric Association (APA), 1952,1968,1980,1987,2000).
These results also show an interaction between different combinations of the three paedophilic personality factors. These findings indicate that the participants do not seem to believe that the sexual degeneracy and abusive factors of the paedophile's personality are related, which appears to contradict previous research (Howitt, 1995). Interestingly, the participants believe that the sexual degeneracy aspect of the paedophile's personality is not related to them being pathological sex offenders, as such reinforcing the heterogeneous nature of paedophilic offenders (Howitt, 1995;Leberg, 1997;van Dam, 2001;Whiskin, 1997).
These findings provide support for the 'double-hermeneutic' model of reflexive modernity (Giddens, 1991) demonstrating that there seems to be some filtering down of professional knowledge into the public domain. Also, even though, these personality factors seem to reflect widespread public attitudes to paedophilia  (McCartan, 2004;Silverman & Wilson, 2002), they do focus on general definitions, thereby reinforcing existing stereotypes of paedophilia, rather than the actual complexity of the paedophilic personality (Howitt, 1998). These findings seem to reinforce previous implicit theory research indicating that implicit theories are based on common-sense perceptions of the world and therefore can be incorrect, biased and/or stereotyped (DeFrates-Densch et al., 2004;Levi et al., 1998;Plaks et al., 2001;Ward & Keenan, 1999).

Implicit theories of paedophilic behaviour
Results. Preliminary analysis indicated that the data was suitable for a PCA, with the KaiserÁMeyerÁOlkin measure of 0.82 (Pallant, 2007) and the Bartlett's test of Sphericity reaching statistical significance (p B0.01) (Pallant, 2007 (Table 4). In deciding the strength of the components a cut-off point of 0.44 was adhered to for the various items (Maltby et al., 2004). Although a number of items failed to load on the four components (Table 4), alternative component numbers did no better. It was decided that the four-factor model had the best component groupings; consequentially component 1 seems to be displaying a 'predatory child sexual abuser' factor, component 2 a 'child lover' factor, component 3 a 'socially inept' factor and component 4 a 'secretive' factor. As part of the factor analysis a Pearson product moment correlation (Table 7) was produced, indicating a weak negative correlation between the child lovers and socially inept factors (r0(0.16, n0173, pB0.05) and also a medium negative correlation between the predatory child sexual abuser and socially inept factors (r 0(0.30, n 0173, p B0.01).
The student/trainee-professionals also believe paedophiles to be secretive, and as such devious, about their attraction to and interactions with children (Howitt, 1995;Kleinhaus, 2002). Thus this seemingly reinforces research on the grooming techniques used by paedophiles (Erikson et al., 1988;Howitt, 1995;Leberg, 1997;McAlinden, 2006b). However, almost contrary to this, the student/traineeprofessionals suggest that paedophiles can be socially inept, arguing that they are more comfortable with children than adults, as such possibly explaining the paedophilic relationship (Blanchette & Coleman, 2002;Wilson & Cox, 1983;Whiskin, 1997). Consequentially, the student/trainee-professionals' implicit theories of paedophiles behaviour reflect the complexity of previous research and professional opinion in this area. These results seem to provide greater support for the 'double-hermeneutic' model of reflexive modernity, than the student/traineeprofessionals' implicit theories of paedophilic personalities do (Giddens, 1991). This suggests that it is not always the case that the common-sense perceptions (DeFrates-Densch et al., 2004), which help to create implicit theories, are always misinformed and/or biased (DeFrates-Densch et al., 2004;Ward & Keenan, 1999).These results also show an interaction between the four different paedophilic behaviour factors.
With student/ trainee-professionals countering the idea that paedophiles, whether they were perceived as predatory child sexual abusers or child lovers, sexually abuse children because they are socially inept (Blanchette & Coleman, 2002;Taylor, 1981), and subsequently unable to function in age-appropriate adult relationships (Groth & Birnbaum, 1978). This finding seems to negate explanations given by paedophiles for their offending behaviour, particularly that the paedophilic relationship is a normal loving relationship (Brongersma, 1984;La Fontaine, 1990;Leberg, 1997;Leo, 1983;Howitt, 1995;O'Carroll, 1980;Taylor & Quayle, 2003). Simultaneously the student/ trainee-professionals seem to be suggesting that all types of paedophiles are capable of complex grooming behaviours (Erikson et al., 1988;McAlinden, 2006b;Wyre, 1987).

Attitudes towards the media
Results. Prior to running this factor analysis it was decided to remove three of the items from the analysis; these were 'do you read a newspaper?', 'do you watch news programmes on the TV?' and 'do you listen to news programmes on the radio?', as these items are reflections of media consumption, and not of the participants' attitudes of the impact of the media. The 10 items comprising the student/traineeprofessionals' attitude towards the media were reduced to seven items. Preliminary analysis indicated that the data was suitable for a PCA, with the KaiserÁMeyerÁ Olkin measure being 0.71 (Pallant, 2007) and the Bartlett's test of Sphericity reaching statistically significance (p B0.01) (Pallant, 2007). The PCA revealed the presence of two components with eigenvalues exceeding 1, explaining 33.7% and 17.1% of the variance, respectively. The inspection of the Scree plot indicated a massive drop between the first and second component and a slight levelling off after the second, using Catell's (1966) Scree test criteria; it was decided to use only two components for further analysis. This was further supported by the results of a Parallel Analysis. An Oblimin rotation was performed to help interpret the two components. The rotation solution (Table 5) indicated that the two components were separate and easily defined. The Oblimin rotation indicates that the two component solution explained 50.8% of the variance with component 1 explaining 33.7% and component 2 explaining 17.0%. In deciding the strength of the components a cut-off point of 0.44 was adhered to for the various items (Maltby et al., 2004). Component 1 seemed to be describing 'media impact' with component 2 describing 'media opinion'. As part of the factor analysis a Pearson product moment correlation (Table  7) was produced, indicating a small positive correlation between the media impact and the media personal opinion components (r00.23, n 0173, p B0.05).
Discussion. The two factors that emerged from the student/trainee-professionals' attitudes towards the media (media impact and personal opinion) support previous research showing that although the media has an impact on our understandings of the world (Howitt, 1998;McQuail, 2007), we are not fully aware of the degree of said media impact (Howitt, 1998;McQuail, 2007). Furthermore, the positive correlation between the two factors offers support for the 'uses and gratification' model of the media impact (Howitt, 1998), which suggests that student/trainee-professionals' personal opinions and their interpretation of the media collaborate to allow them to understand the world. This result highlights the complexity in understanding media impact because individuals react to different media forms and styles in different ways (McQuail, 2007). Thus reiterating that implicit theories are learnt though interactions with the world at large (DeFrates-Densch et al., 2004;Ward & Keenan, 1999) and that the media seems to play a central role in the development of implicit theories through the 'double-hermeneutic' model of reflexive modernity (Giddens, 1991).
Participant personality/coping traits Results. Preliminary analysis did not consistently show that the data was suitable for a PCA, with the KaiserÁMeyerÁOlkin measure being 0.59 (Pallant, 2007), but the Bartlett's test of Sphericity reached statistical significance (pB0.01) (Pallant, 2007). The PCA revealed the presence of five components with eigenvalues exceeding 1, explaining 19.5%, 15.5%, 9.9%, 9.3% and 8.1% of the variance, respectively. The inspection of the Scree plot indicated a massive drop between the first and second component and a slight levelling off after the second, using Catell's (1966) Scree test criteria; although there were five factors with an eigenvalue of 1 or more. However, the results of a parallel analysis showed only four components, as such, it was decided to look at both four and five components for further analysis to see which best described the factor groupings. An Oblimin rotation was performed to help interpret the four components. The rotation solution (Table 6) indicated that the four components were generally separate and easily defined, although there were s a number factors that failed to load on any of the components (total cautiousness and total emotional flexibility), therefore it was decided to look at five components. In doing so it was discovered that there was a double loading of 'coping; emotional regulation' on two factors, hence it was decided to use the four component model instead; consequentially, component 1 seemed to be describing 'cognitive flexibility', with component 2 describing 'positive coping', with component 3 describing 'sociability', and with component 4 describing 'negative coping'. As part of the factor analysis a Pearson product moment correlation (Table 7) was produced indicating that there were no relationships between the participant personality/ coping trait factors.
Discussion. There were four factors that emerged from the participants' personality/ coping traits (cognitive flexibility, positive coping, sociability and negative coping). The findings reflected earlier research highlighting that people tend to have positive and negative coping strategies (Buhr & Dugas, 2002;Ferguson & Cox, 1997;Pyszczynski et al., 1999); that there is a greater deal of inflexibility surrounding social concepts that are prevalent in modern society (Furedi, 2002); and that society's reaction to risk has lead people to become less sociable and more cautious (Furedi,   2002). There were, however, no strong correlations between any of the four personality/coping trait factors (Table 7). This lack of interaction seems to suggest that there is no relationship between these different traits and how people form their implicit theories. As such, seemingly rejecting the premise laid out in the introduction that moral panics may be coping mechanism to deal with social risks.
Higher order correlations Results. A series of Pearson product moment correlations were carried out across the different factor domains (implicit theories of paedophilic personalities, implicit theories of paedophilic behaviours, attitudes towards the media and participant personality/coping traits) ( Table 7) to determine if there were any higher-level interactions between the different factor groupings. Preliminary analyses were performed to ensure no violation of the assumptions of normality, linearity and homoscedasticity. The results indicate that there were some interactions between the various components of the four different factor groupings.
There was a small positive correlation between the sexual degenerate and the media impact components (r00.15, n0181, p B0.05) and a small positive correlation with the sexual degenerate and the secretive components (r00.25, n 0 172, p B0.01). There was also a medium positive correlation between the pathological sex offender and socially inept components (r00.44, n0172, p B 0.01) and a strong positive correlation between the sexual degenerate and the predatory child sexual abuser components (r00.54, n0172, pB0.01).
There was a weak negative correlation between the participants' cognitive flexibility component and the predatory child sexual abuser components (r 0575 0.20, n 0172, p B0.05); a weak negative correlation between the socially inept and the personal opinion components (r0(0.15, n 0170, pB0.05); a weak negative correlation between the child lovers and the abusive components (r 0(0.23, n 0 172, p B0.01); a weak negative correlation between the socially inept and the sexual degenerate components (r 0(0.28, n0172, p B0.01); and a weak negative correlation with the pathological sex offender and personal opinion components (r 0 (0.27, n 0178, p B0.01). There was also a medium negative correlation between the abusive and predatory child sexual abuser (r0(0.34, n0172, p B0.01) and a medium negative correlation between the pathological sex offender and the predatory child sexual abuser components (r0(0.35, n 0172, p B0.01).
There was a small positive correlation between the sexual degenerate and the media impact components (r00.15, n0181, p B0.05) and a small positive correlation with the sexual degenerate and the secretive components (r0 0.25, n 0 172, p B0.01). There was also a medium positive correlation between the pathological sex offender and socially inept components (r00.44, n0172, p B0.01) and a strong positive correlation between the sexual degenerate and the predatory child sexual abuser components (r0 0.54, n 0 172, pB0.01).
There was a weak negative correlation between the participants' cognitive flexibility component and the predatory child sexual abuser components (r 0 (0.20, n 0172, p B0.05); weak negative correlation between the socially inept and the personal opinion components (r0(0.15, n 0170, pB0.05); a weak negative correlation between the child lovers and the abusive components (r 0(0.23, n 0 172, p B0.01); a weak negative correlation between the socially inept and the sexual degenerate components (r 0(0.28, n0172, p B0.01); and a weak negative correlation with the pathological sex offender and personal opinion components (r 0 (0.27, n 0178, p B0.01). There was also a medium negative correlation between the abusive and predatory child sexual abuser (r0(0.34, n 0172, pB0.01) and a medium negative correlation between the pathological sex offender and the predatory child sexual abuser components (r0(0.35, n 0172, p B0.01).
Discussion. The correlations between the different factors groupings revealed that the majority of the significant and strong correlations were between the student/traineeprofessionals' implicit theories of paedophilic personalities and paedophilic behaviours, many of which seem to reflect existing social attitudes (Bell, 2003;Silverman & Wilson, 2002), previous research and professional opinion (American Psychiatric Association, 1952, 1968, 1987Bagley et al., 1994;Blanchard et al., 1999;Lee et al., 2002). However, the student/trainee-professionals seem to suggest some counterintuitive interactions including that the pathological sex offenders personality type is more closely linked to the paedophile being socially inept (Whiskin, 1997), rather than them being a predatory child sexual abuser. Also, contrary to previous research (Howitt, 1995;Silverman & Wilson, 2002;Thomas, 2005), the participants seem to believe that there is no relationship between paedophilic behaviours, either as child lovers or as predatory child sexual abusers, and paedophiles having abusive personalities. Interestingly, the student/traineeprofessionals insinuate that the paedophiles' sexual degenerate personality is not related to them being socially inept, which seems plausible as paedophiles may use sophisticated grooming techniques during their offending (McAlinden, 2006b). These seemingly counterintuitive findings tentatively suggest that the participants seem to have an understanding of the heterogeneous nature of paedophiles, realizing that paedophilia is a complex and difficult area, with offenders having a variety of personality and behavioural traits which can sometimes be dichotomous.
However, there were fewer interactions between the participants' implicit theories of paedophilia and their attitudes towards media and/or their personality/coping traits. The student/trainee-professionals seemed to believe that the media, more so than their personal opinion, has an impact on their understandings of paedophilia (Silverman & Wilson, 2002), suggesting that the media has helped shape their belief in the sexual degenerate nature of paedophiles' personalities, the paedophile pathological sex offender personality type and their beliefs that paedophiles are socially inept. This tentatively suggests that the media plays a role in the formation of implicit theories, possibly through the 'double-hermeneutic' model of reflexive modernity, contributing to the creation of the current mis-perceptions of paedophilia in modern society (Bell, 2002;Cohen, 2002;McAlinden, 2006a;Silverman & Wilson, 2002;Thomas, 2005).
Interestingly, the student/trainee-professionals in suggesting that paedophilic behaviours seemed to be those of predatory child sexual abusers tended to believe that they used a degree of cognitive flexibility and were therefore open minded in thinking this. This finding is seemly counterintuitive in the first instance, however it is important to realize that this perception may be the most socially acceptable understanding of paedophilia, given current attitudes in modern society (Bell, 2003;Howitt, 1995;Scourfield & Walsh, 2003;Silverman & Wilson, 2002).

General discussion
These findings, although preliminary, do seem to suggest that the student/traineeprofessionals implicit theories of paedophilia tend to reflect preconceived public perceptions (McCartan, 2004), professional understandings (McCartan, 2008b, paedophiles' self-perceptions (Howitt, 1995;La Fontaine, 1990;Leberg, 1997;O'Carroll, 1980;Silverman & Wilson, 2002) and media representations (Bell, 2002;Greer, 2002;Silverman & Wilson, 2002) of paedophilia predominate in modern society. This seems to reinforce that implicit theories are common-sense perceptions and understandings of the world (DeFrates-Densch et al., 2004) developed through interactions with the world at large (DeFrates-Densch et al., 2004;Ward & Keenan, 1999); and therefore can be incorrect (Levi et al., 1998;Plaks et al., 2001), or maladaptive (DeFrates-Densch et al., 2004;Ward & Keenan, 1999).Consequentially, this research seems to support the 'double-hermeneutic' model of reflexive modernity (Giddens, 1991), indicating that expert knowledge and explicit theories play a role in the development of non-expert (i.e. lay, student, trainee professional) implicit theories of paedophilia. Although the current research suggests that greater dissemination of expert and explicit knowledge is important in understandings of paedophilia, the research concurrently reveals that personal opinion seems to play an important role as well. Unfortunately the research does not reinforce ideas suggested by the literature that student/trainee-professionals' personality/coping traits could play a significant role in the development of their implicit theories of, and current crisis of, paedophilia in modern society. However, the research findings do highlight that the student/trainee-professionals are not inflexible in all of their implicit theories of paedophilia; as such reinforcing the social constructionalist nature of paedophilia in modern society (McCartan, 2008a). In conclusion this preliminary research stresses the need for wider public education and debate surrounding paedophilia, an increased media sensitivity and a more balanced reporting of paedophilia, as well as a clearer and better defined understanding of expert implicit theories of paedophilia.