"Criminal careers" denotes ways in which offenders develop specialisms and versatility, but studies linking delinquency to social skills deficits have not attempted to explore cognitive, internalised processes by which such "careers" might be chosen. This study investigated criminal minds via script theory: "internal" scripts are used to guide behaviour, "situational" scripts are knowledge of everyday events, and "personal" scripts are a sequence of actions towards a desired goal. This research investigated whether criminal career offenders develop situational scripts for offending and whether such situational scripts express an internalised identity, which manifests as a personal script. Thematic analysis of data derived from "criminal career offenders" supports the notion of criminal situational scripts, with emergent themes considered evidence of personal scripts. "Criminal careers" denotes ways in which offenders develop specialisms and versatility, but studies linking delinquency to social skills deficits have not attempted to explore cognitive, internalised processes by which such "careers" might be chosen. This study investigated criminal minds via script theory: "internal" scripts are used to guide behaviour, "situational" scripts are knowledge of everyday events, and "personal" scripts are a sequence of actions towards a desired goal. This research investigated whether criminal career offenders develop situational scripts for offending and whether such situational scripts express an internalised identity, which manifests as a personal script. Thematic analysis of data derived from "criminal career offenders" supports the notion of criminal situational scripts, with emergent themes considered evidence of personal scripts. Key Words: Criminal Careers, Criminal Versatility, Cognitive Scripts, and Vignette Analysis
(ProQuest: ... denotes formula omitted.)
Some offenders commit a diverse range of offences during their criminal life span; others specialise, or exhibit diversification within a defined range of offence type. The "criminal career" paradigm is concerned with the development of deviant behaviour over time, usually from an early age, and the versatility of, and specialisation in, the crimes that are committed (Piquero, Farrington, & Blumstein, 2007). It has important implications in gaining a greater understanding of criminal behaviour and how that behaviour develops. "Criminal career" is defined as "the longitudinal sequence of offences committed by an individual offender" (Farrington, 1995, p. 511) and offences committed under the label of "criminal career" are usually amongst the most commonly occurring crimes - burglary, car theft, vandalism, and acts of violence. An examination of criminal careers needs to be both a longitudinal undertaking as well as cross-sectional.
West and Farrington (1977) conducted one of the first longitudinal studies on criminal careers, examining approximately four hundred boys, born between 1951- 1954 in a working class area of London, over a fourteen-year period. In interviews conducted between 1971-1973, cross-referenced with official delinquency records, 101 (of the 389 remaining in the study) had official convictions for one or more offences. By 1974, the participants were 20-21 years old, and the number with convictions had increased to 120, with 360 separate convictions. The types of offences were labelled "crimes of dishonesty" (88% of juvenile, 76% of adult convictions) or "crimes of aggression" (6.8% juvenile, 10.6% adult). These labels represent a wide variety of crimes and most of the offenders were convicted of more than one type of offence from either or both of the broad category classifications. Such studies as these represent seminal works in this area. There have been numerous others concentrating on the offences and types of offences. Klein (1984) reviewed 33 separate studies on career criminals, covering 60 cohorts of juveniles. He concluded that there existed a "cafeteria style" of offending, postulating how an offender could browse an array of offences, choosing to commit burglary, theft, vandalism, or violence. For Klein, the ordering of offences was random and not necessarily committed in a specific order, such as escalating seriousness. Four of the reviewed studies provided supporting evidence for patterns of specialisation, but eight showed ambiguous results offering inconclusive evidence for either specialisation or versatility. The remaining 21 failed to identify any evidence of a pattern of offending and supported the view of versatility.
An alternative examination was carried out by Farrington, Snyder and Finnegan (1988), who conducted a very comprehensive study in the USA, using a (then) new technique of measuring offence specialisation known as the "Forward Specialisation Coefficient" (FSC). The FSC is a measure of the probability of the transition from offence i to offence j, given by the following expression:
Where: n(i,i) is the observed frequency
n(j) is the column total, n(i) is the row total
n is the total number of observations.
The FSC can take a value between zero and one. It is zero when there is complete versatility in offending and one when there is complete specialisation.
The study was based on the complete juvenile court careers of nearly 70,000 offenders. A key feature was the "new measure" of the strength of specialisation, combined with a fine-grained classification of 21 offences. In addition, transition matrices of offending careers were also studied, showing a small but significant degree of specialisation on offending, which was superimposed on a great deal of versatility. Furthermore, the degree of specialisation tended to increase with successive referrals. Farrington et al. concluded that 20% of the offenders were specialists and that, whilst offending was generally versatile, delinquency theories should attempt to explain both specialisation and versatility. However, Fisher and Ross (2006) point out that a methodological weakness in such research into criminal versatility is how episodes of offending are classified and categorised. They suggest that the FSC may be too narrow a classification system, leading to inaccurate measures of specialisation/versatility due to under-representation of offence type. Francis, Soothill and Fligelstone (2004) also suggest that official statistics, and studies based upon them, may fall foul of a tendency to classify crime in terms of frequency of offence rather than type. They go even further and show evidence that male offenders exhibit more diversity than female offenders and that this varies with the age profile for the offence type.
Another more recent viewpoint is that derived from analysis of the Cambridge Study on Delinquent Development (CSDD). Kazemian and Farrington (2006) suggest that such studies contribute immeasurably to knowledge about trends in criminal behaviour patterns from a life course perspective. One such contribution is the discovery that self-report data is very much at odds with official statistics, particularly for such data as age of onset of crime and variation in offence type for individuals. Piquero et al. (2007) also suggest that data derived in this way, directly from the individuals involved in crime, offers a valid lifecourse perspective that allows researchers to examine a range of issues including versatility or specialisation.
Such studies then produce evidence of criminal versatility, amongst other things, but offer little explanation for it. The evidence of versatility can be derived from examination of data gathered in a longitudinal manner, from both police records and from empirical data directly gathered from research participants. However, the reasons underlying the statistics are not as clearly explicated. Other theoretical positions attempt to account for a diverse range of criminal behaviours. Toby's (1962) "criminal motivation" theory centres on reference groups and norms of behaviour that represent moral development, and describes how morality is viewed from a vantage point in society and any sub-culture of reference. For a member of this sub-culture, committing an offence may not be the codification of his own ideas or that of his immediate family and friends, but the norms of the reference group. Hence committing an offence is showing loyalty and conformity, posing no moral dilemma. To act in the opposite way would show conformity to the law and non-conducive behaviour to the sub-culture or reference group. Bell (2008) suggests that the subculture's members internalise the social norms of their group, which leads to a higher level of effectiveness in those behaviours showing conformity. Whilst the major work on such types of group identification have focussed on occupational settings, this can be applied to criminal sub-cultural groups too, particularly those engaged in violent crime (van Hiel, Hautman, Cornelis, & de Clercq, 2007). Therefore, it is the lawabiding behaviour that represents the moral dilemma for group members. Through this social processing the individual adjusts to an antisocial role, the delinquent comes to think of himself as a "criminal" and associates with unsuccessful others. He develops a reputation within the group and commits crimes from which he would have originally shrunk. Hence, criminal behaviour can be explained in the same way as any other behaviour provided that the socio-cultural aspects are understood. For this reason, some people can give in to the temptation of antisocial behaviour through the processes of group culture and pressures.
Another relevant view is "Differential Association" (DA) theory (Sutherland & Cressey, 1978) in which the process of developing into a criminal is related to the circumstances in which criminogenic influences exist. Within DA, the emphasis is on the balance of exposure to criminal norms rather than simply association with criminals. Hoffman (2003) suggests that this macro-level theoretical position can be expanded to include individual level behaviour, taking account of contextual and social learning positions, but that such a link must be made with care. In mapping such a macro-level analysis onto individual behaviour, key variables include weakening of social bonds and rising impoverishment, which lead to difficulty in controlling the behaviour of community members and make higher levels of opportunities for delinquent behaviour. In other words, we can observe, at a societal level, elements such as high levels of unemployment concentrated in specific area, coupled with lack of supervision by figures of authority (parental or community) and make the link to the likelihood of higher levels of delinquency in individuals. However, Hoffman rightly warns that this link may be clear, but that the theoretical perspective needs to be supported by empirical evidence.
Whilst each of the theories outlined above offer explanations about the settings in which criminal behaviour can manifest, they still provide little in the way of an individualistic explanation as to why some develop a criminal repertoire of behaviour. A greater potential for an individual, psychological explanation lies in the "socioanalytic theory" (Hogan & Roberts, 2000). This approach accords a central role to self-concept or self-identity, self-presentation (promotion of the self-image), the reference group and interpersonal skills. The theoretical position is that the individual will develop inadequate social skills and an anti-authoritarian attitude in the appropriate context. When this is added to low educational attainment and limited opportunities, individuals will adopt a deviant role, and in social settings, group polarisation takes place. The individual will then adjust the self-image to accommodate the group norms, involving the self-presentational image of possibly being tough, alienated, reckless and exhibitionistic. The criminal career becomes a rational choice and the image of "criminal" becomes the social identity. This leads us to an allied theoretical position known as rational choice theory (Cornish & Clarke, 1998). Rational choice is the term applied to the decision taken to commit crime after all cost and rewards are considered. In addition, rational choice theory would suggest that the decision (choice) to commit crime holds a specific purpose for the criminal and that this choice varies with the crime type (Guerrette, Stenius, & McGloin, 2005). In this perspective, individuals utilise a cost-reward decision-making process in choosing to commit a crime or not. This is completely bound up with the self-concept and moral code of each person. Hence rational choice theory attempts to provide an understanding of specialisation or versatility in terms of consecutive crime events.
The ground that all these theoretical positions share lies in the form of a "modelling" to group norms. The essence of this common ground exists through the interaction of the individual's behaviour within perceived boundaries of the reference group norms. Various studies have drawn from the different theoretical approaches in order to link criminal behaviour to the modelling of reference groups. This has been done by identifying various components such as persistent egocentric or narcissistic behaviour, a lack of social skills, and interpersonal problem solving deficiency, all strongly correlated with delinquency (Donnellan, Trzesniewski, Robins, Moffitt, & Caspi, 2005) and, through role model training, showing how delinquents' behaviour can be redeveloped.
Chandler (1973) refers to the notion that pro-social behaviour is linked to the development of age appropriate role taking or perspective taking of others and social deviancy is consistent with egocentric thought. Chandler examined what happened when delinquents' pro-social skills and perspective taking were heightened in a comparison study consisting of 45 delinquents and 45 non-delinquents. Each sample was evenly assigned to one of three groups: (a) remedial training in role taking, (b) placebo, and (c) non-treatment. The treatment group received drama lessons and were involved in making video films. The placebo and non-intervention groups were engaged in non-treatment programs. Following pre and post-intervention comparisons, the results showed significant reductions in delinquent behaviour for the treatment group due to a raised ability to appreciate others' perspectives. Chandler concluded that delinquents were less able to adopt the role or perspectives of others and that acquired role-taking skills reduce high levels of social egocentrism. These findings are consistent with socioanalytic theory, which suggests that a characteristic of delinquency is an individual's inability to view another's perspective (Hogan & Roberts, 2000). This is supported by work carried out by Mullins-Nelson, Salekin and Leistico (2006), who found weak or negative relationships between the cognitive components of empathy (such as perspective taking) and psychopathy (an indicator of potential antisocial tendencies) .
A similar approach to reducing delinquency in high school students, by concentrating on cognitive and social skills, was described by Sarason and Sarason (1981). They drew from the idea that maladaptations manifest because individuals have failed to formulate behavioural patterns appropriate for everyday situations. The purpose of their study was not to investigate how this breakdown takes place, but how to readdress it when it happens. The study consisted of 127 students attending an urban multi-ethnic high school with a low attendance rate and poor disciplinary record. Participants were divided into three groups, two experimental groups and a control group. Participants who received the special training were able to think of more adaptive ways of approaching problematic situations and perform more effectively in a self-presentational situation. This effect was seen to persist in a oneyear follow-up study, with lower rates of tardiness, fewer absences and behaviour referrals. The conclusion was that delinquents seemed less able to think of alternatives before acting and that they were less "future" or "present" orientated. So appropriate role model training leads to internalisation of a different set of behavioural patterns. This supports theoretical and empirical positions on deviant behaviour that argue delinquents lack appropriate social skills to deal with problem situations. Welsh and Farrington (2006) contend that such interventions as social skills training and cognitive behavioural approaches are indispensable in crime prevention.
Another example of delinquency identified through inadequate social skills comes from Veneziano and Veneziano's 1988 study of 411 male delinquents aged 12-15 who were assessed for social skill competence via the Adolescent Problems Inventory, and other personality, behavioural, social, intellectual and educational skills tests. From the data, the adolescents were divided into three groups (a) competent in knowledge of social skills, (b) incompetent, and (c) moderately competent. Analysis of variance among the three groups suggested that they differed along a number of dimensions. The group with the lowest scores in terms of knowledge of social skills appeared to have a wider variety of behavioural difficulties. Veneziano and Veneziano concluded that if delinquents who lack social skills also lack knowledge of what is appropriate behaviour, it seems unlikely that a good social response could be exhibited. This would appear to support DA theory, which indicates that the learning of norms takes place through exposure to a higher ratio of group relevant behaviours.
Finally, a particularly pertinent study came from Freedman, Rosenthal, Donahoe, Schlundt, and McFall (1978) who applied their Adolescent Problems Inventory (API) to three groups of 16-17 year-old delinquent and non-delinquent participants. Each inventory item briefly describes a situation reflecting an event that could be encountered by the respondents. For example
1) a male peer or stranger who deliberately bumps into you on the street.
2) you feel hopelessly lost in a geometry class.
Participants were given two sets of instructions, firstly they were required to say what they would do in that situation and secondly, say what they thought the best response would be. The view was taken that both delinquents and non-delinquents were likely to perform competently and incompetently on a number of tasks and that the pattern of deficiencies was likely to vary considerably, not only between groups but also within groups. The findings showed that the non-delinquents performed better on skills regardless of the instructions. It was also noted that the delinquents performed better with the instruction "what would be the best thing to do?" and also when the response format changed from free-response to multiple-choice answers. The findings were taken to offer further supporting evidence that delinquents lack social skills, in line with the socioanalytic theory, which argues that delinquents lack interpersonal skills of sensitivity and competence.
One of the drawbacks with role-play studies is that they tend to rely on the notion that delinquents simply lack social skills. What is not clear is whether they are labelled as delinquents on the basis of these deficits or classed as delinquents because of more specific acts, with the deficits being a part of the overall behaviour. More specifically, it is not specified whether someone can be classed, according to a given criterion, as a delinquent without ever breaking the law. Similarly, do delinquents become delinquents because they lack social skills or can they be delinquents even with equally effective social skills as non-delinquents? With these issues in mind, there are limitations to the use of these studies, as they do not adequately explain why some delinquents are versatile in their offending. However, in addition to the research on sociocognitive functions, the role of the self-concept has also been considered with a view to exploring the underlying processes of deviancy.
According to Blackburn (1993) the self-concept incorporates knowledge and beliefs held about one's self, including attitudes that can affect self-esteem and mediates social interaction. A deviant self-concept may be reflected in antisocial behaviour. For instance, Howell (1978) conducted a case study on the effects of selfperception and deviancy, centring on a poisoner ("J"), eliciting a repertory grid of social perceptions. "J", asked to include someone he considered to be successful, chose another poisoner. Howell concluded that deviant behaviour might occur in the context of an alternative definition of reality by the individual conducting the behaviour. This is further supported by an examination of the self-concept in adolescent offenders carried out by Vermeiren, Bogaerts, Ruchkin, Deboutte and Schwab-Stone (2006). They found that subtypes of self-esteem and self-concept present differently in antisocial adolescents and conclude that this may have clinical or educational implications. Such differences may indicate that cognitive distortions are used to justify behaviour. Taking the view that criminal rationalisations are unacceptable socially, the finding of distorted thinking supports DA theory that it is criminal technique, attitudes and rationalisations that are learnt.
In summary, research shows that distorted thinking maintains deviant behaviour through denial and avoidance of responsibility. Additionally, it is possible to see how behaviour can also be guided by the self-concept and how low self-esteem can motivate and influence dishonest behaviour. This is linked to inadequate social skills in offenders, demonstrated by inappropriate performance in problem solving situations. However, through role-play training, offending rates can be reduced. This indicates that dysfunctional behaviours can manifest through exposure to inappropriate modelling. Another way in which modelling or role-play can be explored is through the use of script theory.
Script theory and criminal behaviour
Schank and Abelson (1977) developed script theory in order to explain the understanding process during a situation or event. Initially a description of language processing and higher thinking skills, it was later extended to an explanation of storylevel understanding. Essentially, script theory maintains that all memory is encoded episodically, in that, in order to be retained, everything must be related in some way to personal experiences. However, in addition to highly personalised episodes, there are generalised "scripts", that aid in conceptualising a narrative in which personal information may not be available. Scripts are used to guide behaviour because the script provides the holder with a set of expectations about what will happen during the unfolding of an event, thus offering a way of predicting the outcome and aid the individual to act accordingly. There are two basic types of script: "situational" and "personal." A situational script relates to knowledge of everyday events, such as going to a restaurant, where it is assumed that most people agree on what actions constitute a restaurant visit. Therefore, a customer would act according to a script when making a visit. Situational scripts are acquired through exposure to relevant experiences. "Personal" scripts consist of a sequence of actions that are designed to lead the "personal" script holder to a desired goal. For example, the restaurant customer may also have the goal of attempting to arrange a date with the waitress, so both types of scripts are operating simultaneously.
Bower, Black, and Turner (1979) used vignettes of everyday events to access how scripts are used in memory. A total of 161 student participants were asked to list the actions that would produce a script for each event, either in free-form response (which produced too much elaboration) or listing in order of occurrence action responses to the script. The findings showed that there was considerable agreement amongst participants about the content and order of actions that went into them. The conclusion was that script knowledge is held in memory and in an ordered form.
A further application of script theory to the examination of deviant behaviour was carried out by Eifler (2007). Eifler presented verbal and visual vignettes of mild deviant behaviour (e.g., not stopping at a red traffic light) to 150 participants and asked what they would do. In the majority of cases the deviant behaviour was the first choice, even though participants identified the behaviour as deviant. Eifler concluded that vignette analysis was a valid method of examining the internal scripts associated with deviant behaviour that we all hold. However, this research was carried out using participants from a non-offender population. This is supported by Fontaine (2007), who suggests that hypothetical vignettes are an appropriate way to examine social information processing and behavioural decision making.
In reviewing the research on role-play and mind scenarios, it appears there is a lack of research attempting to explain the causes, from a psychological viewpoint, of criminal versatility in the individual. Therefore, this study seeks to investigate the role of cognitive scripts in criminal careers. Initially, the study looks at applying situational scripts to account for an offender's repertoire of behaviours. This is followed by investigating the notion of personal criminal scripts in relation to the role of modelling and the self-concept as causal attributions. In other words, the study will explore the research question "do criminal career offenders develop situational scripts for offending, and, if so, are they based on an internalised identity, which manifests as a personal script?"
The research, its intent, protocols and procedures for ethical treatment of participants, was submitted for the scrutiny of the University of the West of England's Research Ethics Committee. Two members of the committee scrutinise the proposal and make recommendations to the committee as to acceptance. The researchers are bound by the Code of Conduct of the British Psychological Society (see http://www.bps.org.uk/the-society/ethics-rules-charter-code-of-conduct/code-ofconduct/) and all procedures were deemed to comply with this stringent code and therefore acceptable to the University Research Ethics Committee. There were two items requiring particular scrutiny. The first was the element of deception involved. Participants were to be naïve as to the psychological interests of the study, as this prior knowledge might have affected their responses. All participants were to be fully debriefed at the end of the session, and allowed the opportunity to withdraw from the study at any time, particularly if they were uncomfortable at this mild deception. Secondly, advice was received from the Ethics Committee on recording of participants' responses, as it was felt that material collected would be sensitive in nature. It was therefore decided that responses would be recorded in hand-written form, and not tape recorded. The research protocols were also submitted to the Probation Service serving as gatekeeper for access to participants. The service also suggested that participants were more likely to be more comfortable with hand written records.
Ten male participants were recruited via the probation service and fit the criteria of "criminal career offender" (Farrington, 1995) in that they had been convicted on at least two separate occasions, with a minimum of two types of offences. No deliberate decision was made to only include male participants, but access to female participants simply did not happen due to potential female participants not being available at the time of the study. Age ranged from 18-26 years, (mean age 22.5). The participants were naive to the psychological interests of the study, but were fully debriefed at the end of the data collection.
Vignettes similar to the Freedman et al. (1978) study were developed with two sets of instructions for responding. This clarified the participants' scripts and provided a comparison to a perceived appropriate script. In view of the problems experienced by Bower et al. (1979) with regard to the free response format, a more structured option was chosen. Furthermore, because contributory attributions to the existence of the scripts are sought, the vignette responses were followed by a semi-structured interview. The data obtained from the vignettes and interviews were subjected to thematic analysis as described in Gavin (2005, 2008). Due to the sensitive nature of the material, interview responses were hand written, not recorded. The responses were written by the researchers and verified as correct by the participant at the end of each interview. Including the participant in the process of verification is known as member checking (Schwandt, 2007), and is seen as a crucial technique by which data can be checked for its validity. In order to establish the recorded data and its interpretation as a credible account of the participants' view, the participant is given the opportunity to assess the accuracy of the record made. Such a process is regarded as a plausible way to minimise misrepresentation and misinterpretation of the account the participant wishes to give (Onwuegbuzie & Leech, 2007)
Four separate vignettes were used. Each set of four was administered twice, in two rounds. The first round a standard instruction in how to complete them was given. The second round contained a modified version of the first instructions. The first round of vignettes was coded V1a, V2a, V3a, and V4a and indicated the use of the standard instructions. The second round was coded V1b, V2b, V3b, and V4b to indicate the modified instruction format.
Participants were asked to read the instructions at the top of the vignettes and proceed by answering each one in turn, using their own words. All participants were interviewed separately, on a one-to-one basis. The first round of vignettes was given to the participants. Each vignette response was completed before the next one was presented. Once completed, the second set of vignettes was then presented and delivered and completed one at a time. All participants completed all vignettes. After completion of the second round, a few minutes were then taken to compare responses between the pairings of vignettes from the two sets. Following this, a short, standardised open-ended interview was then conducted. This method uses a predetermined set of questions, but there are no set responses (as would be the case in closed questions) and participants are allowed flexibility in how they respond (see Gavin, 2008). There were a total of four questions for each participant, with one question relating to each pairing of vignettes. The questions centred on the responses for each of the vignettes in set "a" and the corresponding vignette in set "b". For example, if a participant indicated that he would commit an offence in a given vignette from set "a" but then in the corresponding vignette from set "b" indicated recognition of the socially desirable response, the question would be structured in the following way:
"Why would you commit an offence in that situation when you know that the best answer is what you have said in the second story?"
Alternatively, if a participant indicated in both corresponding vignettes that they would commit an offence, the question would be structured in this way.
"Why do you think that the best answer is to commit the offence?"
The interviews lasted on average 20-25 minutes. After the interview participants were then invited to examine the contents of the responses they had made and were given the opportunity to agree or disagree on accuracy.
Coding of Vignette Data
Assignment of responses to vignettes were based on the researchers' interpretation of what is a clear indication of a response to either commit an offence or not and what is interpreted as a non-committal answer. This was then counter-checked for inter-rater reliability.
The responses to the vignettes were assigned to one of three predetermined categories. These categories were decided on the basis of a clear indication to:
1) Commit an offence, NEGATIVE
2) No clear indication to either commit an offence or not commit an offence, NEUTRAL
3) A clear indication not to commit an offence, POSITIVE
Transcription and Analysis of Data
Data transcription provided a preliminary opportunity to gain an overview, and then the vignette and interview data were studied over a longer period, using a theory-led thematic analytical approach (see Smith, Harre', & Van Langenhore, 1995) in order to gain an understanding of the emerging themes. A thematic analysis allows the examination of underlying themes in responses by identifying commonalities in the terminology used. This process of analysis is well-established, and has precedents in the area of examining offenders' responses in qualitative interview approaches. For example, Kellett and Gross (2006) examined the accounts of "joy-riders" when describing the act of stealing and driving cars, using a thematic analysis approach. They note that the accounts can be examined and coded in a hierarchical manner, and more and more implicit themes are revealed as the more explicit are explored. Thus, using thematic analysis, the analyst aims to explicate structures that represent attitude and belief systems. Hence, the transcripts were read and re-read, items that showed high frequency of occurrence were noted. The themes were identified on the basis of recurring elements seen in the transcripts and/or answers to vignettes. This was done by writing, in the margins, abbreviated codes to summarise the main focus of each sentence. A summary of each theme was developed to capture the focus of the material within it. In this way, the data was reduced to a manageable set. Various themes were identified, and there was significant agreement about appearance of these themes amongst the material. As new topics arose, all transcriptions were reexamined to find material relevant to that topic theme. The material was then rechecked to identify overlaps and networks of meaning. This process resulted in the production of proto-themes (Hayes, 2000), initial groupings in terms of similarity of information contained in the text. All material was then re-read to extract data that confirmed or disputed each proto-theme. These themes were then titled and defined and are explored and discussed in the relevant sections.
The researchers are both academic psychologists, who have had experience working in the criminal justice arena. This facilitated the contact with the probation service and informed the research in a particular way. As researchers, we were not explicitly part of the criminal justice system, but understood enough about it to be comfortable with elements of the system. Additionally, as we are not authority figures within the system, the participants may have been more comfortable taking part in interviews with us than with, for example, probation officers. The research was carried out as part of a qualification assessment for one researcher, and supervised by the other.
The researchers conducted the interviews personally. In doing so, it is recognised that interviews were not objective and were affected by the researcher role, experiences, perspective and mannerisms and delivery of the questions. Whilst there was considerable similarity in the questions, they were tailored according to the individual's responses to the vignettes.
Part 1: The Vignettes
The results show that there is general agreement between participants in respect of how they would respond to the scenarios for the first set of vignettes "a" (what would you do?). The total column indicates that between all 10 participants and across all four vignettes, there would be a total of 27 criminal responses from a possible 40. There were also seven responses that indicated a non-committed or ambiguous answer and six responses that indicate no offence would be committed. However, there is slightly less agreement with respect to the responses for the scenarios in the second set of vignettes "b" (what is the best thing to do?). There were a total of four responses indicating that an offence would be committed, a total of 15 responses indicating a non-committal or an ambiguous response and a total of 21 responses indicating no offence would be committed. Figure 1 shows the data from the table graphically.
Part 2. The Interviews
Data extracted from the interviews were thematically analysed.
Theme one: Role switching
This was the only theme that was formed solely from vignette data. The theme highlights the switching from one role to another as participants explained what they would do, compared to what they thought was the right thing to do.
This particular example was taken from responses to Vignette 3, which asked what the participant would and should do, given the opportunity, in a situation where another driver had driven badly, causing the participant to swerve. Note: the numbers in brackets at the end of the responses relate to which participant made the comment.
Vignette responses to instruction (a) what would you do?
"I'd smash it up" (1)
"I'd do something, probably wait a while first" (7)
This compared to responses for instruction (b) what is the best thing to do?
"you don't do anything" (1)
"I would leave it" (7)
In these examples the participants were switching from one perspective to another, according to which role they were being asked to perform. However, there was less certainty about what was the right thing to do for other participants.
For example, in response to a different vignette, using instruction (a) one participant said:
"I'd ask him what he's doing" (2)
When responding to the same vignette but with instruction (b) the same participant said again:
"I would still ask him what he's doing" (2)
Another participant responded to instruction (a) in the following way:
"I'd hit him" (1)
and the same participant to the same vignette but with instruction (b):
"don't know" (1)
For some participants there are different options from which they will make a choice, but for other participants there does not seem to be the same opportunities.
This theme illustrates how the participants are able to switch between scripts they perceive to be appropriate to a given scenario. Several participants indicated two options in response to the different sets of instructions, from a chosen script or an alternative script. This relates to the Freedman et al. (1978) findings, which showed that participants could respond more positively to a specific instruction set, such as in the second round of vignettes. Some, when asked the best thing to do, responded "don't know". This indicates scripts from which to guide their own behaviours, but no alternative scripts. This is supported by the findings of Veneziano and Veneziano (1988) who concluded that delinquents were less able to adopt the role of others, and that of Welsh and Farrington (2006), who show that early training in perspective taking overcomes this. Furthermore, in DA theory, it is assumed that criminal motivation is learnt. However, in considering the findings of the social skills studies, it maybe that in some circumstances the behaviour is not driven by motivation, but activated in the absence of an alternative script. This theme demonstrates the existence of criminal situational scripts, illustrated not only by the negative responses to the vignettes, but by the alternative responses and in the other example by the absence of an alternative.
Theme two: Goal orientated
This theme is concerned with the view that responses to particular situations have a goal aim. In some cases, the goal appears to be clear, such as when a participant indicates that they would take an item because it may contain money (the implication being that money is the goal). However, in other cases, the goal may not always be so obvious. An example of this might take the form of a response where a participant indicates that they do not really know why they would act in a particular way. They just know that they would. In this case, the implication is that the goal is more theoretically based, such as self-presentational drives.
Vignette 2 was concerned with whether a participant would steal in a given opportunist moment, describing how one could easily take a bag from the seat of a parked car whilst walking past it.
"I'd take it cos I'm a thief." (6)
"it's everyone for themselves in this world, innit!" (10)
Vignette 1 was about what to do in a situation where another individual seemingly deliberately barges into the participant.
"I don't really know, I just wouldn't stand for it" (2)
"If you let them take advantage they'll do it again" (3)
The first two examples were relatively straightforward in respect of what might appear to be an obvious goal. It is in the second set of examples where the goal aims were less obvious.
This theme is, in many ways, the backbone of the personal script notion. It deals with the goal aims of the participants, in respect of their vignette responses. In some cases there are clear and obvious examples of participants goal aims. For example, "I'd take it cos I'm a thief" and "it's everyone for themselves in this world, innit". These comments indicate clear goals and were in response to taking an item that could lead to financial gain. However, in other examples the goal aim is not so clear. For example, "I don't really know, I just wouldn't stand for it" and "if you let them take advantage they'll do it again." These comments were in response to dealing with a potentially violent encounter. The participants could not fully explain their decision to act, just that they would react in a negative way. One interpretation is that they may lack the appropriate skills or scripts to deal with the problem appropriately. When considering the findings of Freeman et al. (1978) this is a viable view. However, it is also appropriate, in light of other literature reviewed, that the underlying goal aim is a cognitively determined one and this can coexist alongside the "limited social skills" argument (Mullins-Nelson et al., 2006). The research interpretation is that the participant is motivated to present an image, which he has come to internalise and considers his identity. Supporting evidence comes from Toby (1962) who talks of the delinquent perceiving himself as a criminal, and developing a reputation, committing crimes that he would not have originally considered. Similarly, Hogan and Roberts explain the behaviour by self-presentational needs involving role-play to promote the self-image. They discuss the self-concept as playing a prominent role in the individual's influence on behaviour choice. Furthermore, Vermeiren et al. (2006) argue that a deviant self-concept manifests in antisocial behaviour. Another argument is that participants may have deliberately chosen to answer in a way that they know not to be accurate, indicating forms of demand characteristics (Gavin, 2008). However, another interpretation is that an inaccurate answer also serves the self-presentational need, i.e., the participants have a personal script based on the self-concept of a criminal and the goal aim is to present that image.
Theme three: The belief system
This theme is concerned with how perceived appropriate behaviour is used to make choices about what to do. This is closely related to theme two; however, there is a difference. Theme two was concerned with what an individual "would" do, this theme is concerned with what an individual "would not" do.
"stealing is wrong, I'm no thief" (5)
"well I don't think it's right to steal from other people, businesses and all that, but not ordinary people" (7)
Vignette 4 was about whether a participant would, whilst in police custody provide the police with the required information concerning other suspects.
"If you get caught doing something, you have to take the consequences, so I would, but I wouldn't get someone else nicked" (8)
Within this theme, there is an indication that participants do possess a sense of right and wrong. However, it is construed from different reference points to what is seen as legally correct or socially acceptable views. This theme is inherently concerned with what a participant would not do. It dovetails with theme two in that it represents the participants' expressed morality and or the presentational component of the self. For example, one participant said, "well I don't think its right to steal from other people, businesses and all that, but not ordinary people". In this case the respondent offers a rationale for behaviour that is a compromise between what is socially acceptable and what is appropriate behaviour to 'some' criminals. The question is from where does this abnormal belief system come? Toby talks about how reference groups and norms of behaviour represent moral development for the criminal. Sutherland and Cressey (1978) point to criminal rationalisations being learnt. Hogan and Roberts refer to the reference group, which is the internalised view of the expectations of significant others. This point, then, represents the self-presentational component of the theme. The criminal may hold an abnormal belief system or make cognitive judgements rationalised from a perceived appropriate behaviour that may be due to social pressures or motivation and attitudes, and reflects through self-presentational needs. In this way, the theme supports the notion of a personal script for a 'criminal'.
Theme four: Contradictory perception
This theme is concerned with responses that contradict not only the participant's own belief system, but that of the legal one as well.
"because you don't do that" (1)
"the right thing to do is let it go, you can't fight on the street, there's too many witnesses" (5)
"No I wouldn't steal a woman's handbag, it's not right it could be my mums" (2)
There is a clear three-way contradiction between what the participants believe to be right, what is legally right, and what the participants say they would actually do.
The theme is concerned with contradictions expressed by the participants in respect of their own declared morality and that of the legal position. In some ways this is a continuation from theme three, concerned with a belief system and selfpresentational needs, theme four is only related to self-presentation. For example, responses reiterated, "because you don't do that" or "the right thing to do is let it go, you can't fight on the street there's too many witnesses." In this theme, the participants show that they have scripts for specific situations. However, there is no logic in which social norms can be followed. It appears that a script is in place and is salient during these particular scenarios. However, in doing so the opportunity to appraise the situation differently is lost by the restrictions of the script. This is indicated in the Sarason and Sarason (1981) study, and Welsh and Farrington's (2006) examination of interventions, where it was concluded that delinquents were less able to think of alternatives before they act until trained to do so. The dysfunctional belief system links into the self-concept of the "criminal".
Theme five: Avoidance and distorted justification
The fifth and final theme is concerned with how participants use cognitive distortions and avoidance of issues to continue with their deviant behaviour. Responses indicate victim blaming "you can't blame anyone but yourself if you leave things around like that" or avoidance "it's up to them, I'm not getting involved." In these two examples, the participants are avoiding the issue of taking any responsibility. This theme can be related to Chandler's (1973) study on egocentric thought, which refers to antisocial behaviour being due to a delinquent's inability take the perspective of others. The delinquent cannot or does not want to see the reality of the situation. This denial suggests that, whilst high self-esteem may act as a deterrent to dishonesty, low self-esteem does not. Therefore, given that deviancy is linked to low self-esteem and egocentric thought or the inability to view events from another's perspective (Mullins-Nelson et al., 2006), it is possible to see how those with low self-esteem can be influenced by egocentric thought and behaviour. This results not only in deviant acts, but also in avoidance of taking responsibility due to a lack of self-worth.
The theme relates to participants who find ways of justifying their behaviour by blaming other people.
"you can't blame anyone but yourself if you leave things around like that" (9)
"that's their problem for leaving it there" (5)
In another example, a participant responds in the following way:
"it's up to them, I'm not getting involved" (4)
Through these examples, it is possible to see how some behaviour can be rationalised.
Summary of results
The research aim was to identify whether script theory could be used as a way of explaining criminal versatility. This was explored in respect of offenders holding situational scripts for offences and operating from a personal script. The methodology employed was designed to extrapolate any possible mindsets for potentially criminal situations and to determine whether the scripts were situational or personal. The results reflected this aim by exploring the two different sets of data derived from the same participants. The first examination dealt with the issue of script or mind set responses to the vignettes. The results showed that there was a high proportion of responses to committing an offence (negative) for the instruction "what would you do?" This compared with much lower responses to not committing an offence (positive) and non-committal or ambiguous responses (neutral). The second instruction set "what is the best thing to do?" revealed a high proportion of responses to not committing an offence and also a similar number of non-committal or ambiguous responses. Additionally, there were a low number of offence responses with the second instruction set.
The second data set and its examination dealt with the issue of personal scripts. In this section, the data was thematically analysed and revealed a number of themes that are related to the research aims and reviewed literature.
In respect to the responses obtained from the vignettes containing the "what would you do?" instruction, support the notion that offenders do have scripts for committing offences. This is demonstrated in a number of ways. Initially the coded results show that out of a possible 40, there were a total of 27 offence responses across the full range of scenarios. Situational scripts constitute a set of actions in response to a situation, which is acquired through exposure to that event, and they are held in memory. Therefore, the basic criteria for script theory is met through the methodology employed and the results obtained. Secondly, from the four vignettes, participants averaged nearly three criminal script responses, and related offences of more than one type, indicating versatility. The interpretation therefore is that these findings support the notion of offenders possessing a range of scripts for different offence types, i.e., versatile criminals acquire a cluster of criminal "situational scripts" to guide behaviour in specific or relevant situations. Thirdly, it is possible to see how learning opportunities for single and multiple "criminal script" development can take place. In DA theory, learning is through social interaction and the process of learning criminal patterns involves the same mechanisms entailed in any other learning. Therefore, criminal scripts can be acquired instead of or alongside other scripts that are associated with acceptable behaviours. Fourthly, support for criminal situational scripts comes from the corresponding vignettes, where the instruction set was "what is the best thing to do?" In this set of results there are almost as many responses indicating a non-committal or ambiguous answer, as there are indicating a positive response. Hogan and Roberts (2000) and Mullins-Nelson et al. (2006) argue that offenders lack interpersonal skills of sensitivity and competence with respect to the ability to view another's perspective. This links into responses from the participants who did not offer a positive answer to the instruction "what is the best thing to do" and indicates that they do not have an appropriate script for that scenario, particularly when considering their negative response to the "what would you do" scenario. At this stage, it is appropriate to consider the results in terms of a limited knowledge base and within the realms of the studies reviewed on deviant behaviour, a lack of social skills and role-play. For example, Sarason and Sarason (1981) argue that maladaptations of behaviour manifest because individuals fail to formulate behavioural patterns appropriate to situations that are encountered daily. They showed that, through targeting specific behaviours and by retraining the individual through the use of role-play techniques, different behavioural sets can be learnt. The interpretation here is that the existing behaviours are scripts, so newly acquired behaviours can also be thought of as scripts. To emphasise this point further, Veneziano and Veneziano (1988) reported that those who scored lowest on social skills also lacked knowledge of appropriate behaviours and it therefore seemed unlikely that a good social response could be exhibited. Such a finding is supported by the positive results of various interventions in social skills training as described by Welsh and Farrington (2006). This can also be interpreted as a lacking in appropriate scripts. Collectively, the evidence indicates that there is strong support for criminal versatility being explained by script theory.
The second set of results produced a number of themes providing support for the notion that versatility is linked to personal scripts. The themes drawn from the interviews offer an insight into the cognitive style and rationale of the participants' functioning.
The data from the vignettes together with the themes extracted from the interviews suggest that script theory is a plausible form of investigation and explanation of criminal versatility. According to various theoretical viewpoints discussed above, the learning of criminal behaviour takes place through the process of social interaction and association with criminal others. In addition, within DA it is considered that the learning of criminal patterns involves the same mechanisms entailed in any other learning. Given this, combined with Schank and Abelson's (1977) view that scripts are developed through exposure to everyday events, it is entirely possible to relate the notion of acquiring a cluster of situational scripts to account for a diverse range of behaviours. Individuals may develop scripts for situations based on the learning opportunities and experiences presented to them within their environment. This is supported by the studies reviewed on role-play, which show that a set of behaviours associated with a particular situation can be learnt.
Themes two, three, four, and five provide support to the notion that criminal career offenders can develop a personalised script. These themes consist of a goal aim, a criminal belief system, a criminally motivated perception, and a self-serving set of distorted cognitions that protect the individual's low self-esteem. The personal script interpretation is supported most by the socioanalytic theory, which refers to the self-concept as promoting the self-image or presentation through role-play. The notion is also a general consideration of criminal motivation theories, which state that the delinquent comes to think of himself as a criminal. This then becomes the personal script, acquired through social pressures, exposure to group norms or internalising the view of the reference group expectations. It is the interpretation here that situational scripts are the mechanisms in which the personal script is expressed either intentionally or in some cases through a lack of alternatives. They provide the individual with specific behavioural sets for particular situations. Instead of thinking of the situational scripts as inadequate response patterns of behaviour, or a lacking in behavioural response sets, they might be thought of in terms of a fully developed set of behavioural responses appropriate to the individuals' self-concept. Criminal motivations and drives underpin this personal script. Therefore, if situational scripts are altered, as in role-play studies, the personal script would also need to be altered in order to effect real change. Similarly, if an individual's motives and attitudes were addressed and thereby altering the personal script, the cluster of situational scripts would again need to be redeveloped to prevent the individual from lacking the necessary social skills needed to function in an acceptable way. Script theory is an approach that aids explanation of criminal versatility and because the theory operates at a cognitive functional level, it can be superimposed on existing theories of criminal development.
Limitations and suggestion for future research
The study was concerned with the use of script theory as a method of explaining criminal versatility, and the use of vignettes to explicate the internalised scripts. Its primary aim was not to research the acquisition and development of criminal behaviour, but to establish the existence of script use. Therefore, the participants were already well established in habitual criminal behaviours, as it is this group of offenders who exhibit criminal versatility and are therefore subject to script theory. Similarly, extraneous variables such as motives and drives relating to drug dependency were not a research consideration and limit the findings to non-dependent drug users. However, it maybe that these issues form part of the motivational drives discussed with the various theories reviewed in the literature, which was established without reference to high drug dependency. Therefore, future research should consider drug dependency and addictive behaviour with respect to its direct influence on script theory.
A further consideration here is whether the participants were responding to questions in a truthful and meaningful manner. Using qualitative research approaches often means that judgments about usefulness and credibility of design, data collection and analysis are left to the researcher and the reader. One means of explicating this is to gather data in different ways and to verify with the participants that the interpretation is an accurate account. In this research the vignette responses were further explored in interviews, examining the ways in which the participants reacted to their initial responses and how they explained them. This allowed a form of verification of each type of data against the other. The participants were also naïve to the research objectives. Whilst this does not mean that they were unaware of the research needs, or did not understand their place in the process, this mild deception meant that there was consistency in the way participants and their data were treated. Furthermore, each participant was asked if he wished to change any of the responses once they were defined, and none took that opportunity. Hence there were several opportunities made for triangulation of data, as suggested by Lincoln and Guba (1985) in order to verify the findings, convergence between interpretations of the emerging account checked at various points (Mays & Pope, 1995). This study used semi-structured interviews and thematic analysis to gain an insight into the notion of criminal behaviour and script theory. Whilst this was entirely appropriate in view of its exploratory nature, it is accepted that there will be alternative methods to be considered. Such alternatives might include in-depth interviewing, or a more observational method. For example, Holmqvist (2008) describes using semistructured in-depth interviews and standardised questionnaires to examine the lack of affect in juvenile psychopaths. He concludes that the lack of consciousness with respect to emotions such as shame might be, notwithstanding any neurological findings, attributed to a poorly integrated situational script. Also, Fontaine (2008) reports on a study in which antisocial behaviour is monitored online. In this way, observers can determine real time social decision making processes in respect of the adherence to or rejection of aggressive social scripts.
It is also acknowledged that there is little scope for generalising from this small sample of young adults, to either a wider criminal population or the very young exhibiting deviant behaviour.
Criminal versatility is a description of the way some offenders commit a diverse range of offences whilst others specialise. The aim of this study was to examine criminal versatility via the examination of internalised cognitive scripts that repeat offenders might hold. It was postulated that offenders my hold situational scripts for hypothetical criminal scenarios, but operate from a personal script that may deviate for the more socially desirable situational script. Participants meeting the criteria of "career criminal" were recruited for the study, and discussed scenarios in terms of "what is the best thing to do?" and "what would you do?" Results demonstrated that there were contradictory responses to each set of events, with participants revealing a knowledge of socially acceptable behaviour (the best thing to do) but intention to commit offences (what would you do?) in more than two-thirds of cases.
The study has revealed how criminal career offenders can be in possession of a cluster of criminal scripts and a personal script of the "criminal." These scripts can be acquired through the various relevant psychological processes discussed in each of the three theories. Script theory, used in this context, deals with the cognitive mechanisms of those underlying processes and has revealed some interesting items for further research.
Bell, D. (2008, February 11). Identity and social innovation. Structure and Dynamics: e-Journal of Anthropological & Related Sciences, 2(3), Article 2. Retrieved September 1, 2008, from http://repositories.cdlib.org/imbs/socdyn/sdeas/vol2/iss3/art2
Blackburn, R. (1993). The psychology of criminal conduct: Theory, research, and practice. Chichester, England: Wiley.
Bower, G. H., Black, J. B., & Turner, T. J. (1979). Scripts in memory for text. Cognitive Psychology, 11, 177-220.
Chandler, M. J. (1973). Egocentrism and antisocial behaviour. Developmental Psychology, 9(3), 326-332.
Cornish, D., & Clarke, R. (1998). Understanding crime displacement: An application of rational choice theory. In S. Henry & W. Einstadter (Eds.), Criminology theory reader (pp. 45-56). New York: New York University Press.
Donnellan, M., Trzesniewski, K., Robins, R., Moffitt, T., & Caspi, A. (2005). Low self-esteem is related to aggression, antisocial behavior, and delinquency. Psychological Science, 16(4), 328-335.
Eifler, S. (2007). Evaluating the validity of self-reported deviant behavior using vignette analyses. Quality & Quantity, 41, 303-318.
Farrington, D. (1995). Human development and criminal careers. In M. Maguire, R. Morgan, & R. Reiner (Eds.), The Oxford handbook of criminology (pp. 511- 548). Oxford, UK: Oxford University Press.
Farrington, D. P., Snyder, H. N., & Finnegan, T. A. (1988). Specialisation in juvenile court careers. Criminology, 26(3), 461-487.
Fisher, G., & Ross, S. (2006). Beggarman or thief: Methodological issues in offender specialisation research. Australian & New Zealand Journal of Criminology, 39(2), 151-170.
Fontaine R. (2007). Toward a conceptual framework of instrumental antisocial decision-making and behavior in youth. Clinical Psychology Review, 27(5), 655-675.
Fontaine, R. (2008). On-line social decision making and antisocial behavior: Some essential but neglected issues. Clinical Psychology Review, 28(1), 17-35.
Francis B., Soothill, K., & Fligelstone, F. (2004). Identifying patterns and pathways of offending behaviour: A new approach to typologies of crime. European Journal of Criminology, 1, 47-87.
Freedman, B. J., Rosenthal, L., Donahoe, C. P., Schlundt, D., & McFall, R. M. (1978). A social-behavioural analysis of skill deficits in delinquent and nondelinquent adolescent boys. Journal of Consulting & Clinical Psychology, 46(6), 1448-1462.
Gavin, H. (2005). The social construction of the child sex offender explored by narrative. The Qualitative Report, 10(3), 395-413. Retrieved May 5, 2009, from http://www.nova.edu/ssss/QR/QR10-3/gavin.pdf
Gavin, H. (2008). Understanding research methods and statistics in psychology. London: Sage.
Guerette, R., Stenius, V., & McGloin, J. (2005). Understanding offense specialization and versatility: A reapplication of the rational choice perspective. Journal of Criminal Justice, 33, 77- 87.
Hayes, N. (2000). Doing psychological research. Buckingham, UK: Open University Press.
Hoffman, J. (2003). A contextual analysis of differential association, social control, and strain theories of delinquency. Social Forces, 81(3), 753-785.
Hogan, R., & Roberts, B. W. (2000). A socioanalytic perspective on person/environment interaction. In W. B. Walsh, K. H. Craik, & R. H. Price (Eds.), New directions in person-environment psychology (pp. 1-24). Hillsdale, NJ: Lawrence Erlbaum.
Holmqvist, R. (2008). Psychopathy and affect consciousness in young criminal offenders. Journal of Interpersonal Violence, 23(2), 209-224.
Howell, K. (1978). The meaning of poisoning to a person diagnosed as a psychopath. Medical Science Law, 18(3), 179-184.
Kazemian, L., & Farrington, D. (2006). Length and residual number of offenses for two generations of repeat offenders. Journal of Research in Crime & Delinquency, 43(1), 89-113.
Kellett, S., & Gross, H. (2006). Addicted to joyriding? An exploration of young offenders' accounts of their car crime. Psychology, Crime & Law, 12(1), 39- 59.
Klein, M. W. (1984). Offence specialisation and versatility among juveniles. British Journal of Criminology, 24(2), 185-194.
Lincoln, Y., & Guba, E. (1985). Naturalistic inquiry. New York: Sage.
Mays, N., & Pope, C. (1995). Education and debate qualitative research: Rigour and qualitative research. British Medical Journal, 311, 109-112.
Mullins-Nelson, J., Salekin, R., & Leistico, A. (2006). Psychopathy, empathy, and perspective-taking ability in a community sample: Implications for the successful psychopathy concept. International Journal of Forensic Mental Health, 5(2), 133-149.
Onwuegbuzie, A., & Leech, N. (2007). Validity and qualitative research: An oxymoron? Quality & Quantity, 41, 233-249.
Piquero, A., Farrington, D., & Blumstein, A. (2007). Key issues in criminal career research: New analyses of the Cambridge study in delinquent development. Cambridge Studies in Criminology, New York: Cambridge University Press.
Sarason, I. G., & Sarason, B. R. (1981). Teaching cognitive and social skills to high school students. Journal of Consulting & Clinical Psychology, 49(6), 908-918.
Schank, R. C., & Abelson, R. P. (1977). Scripts, plans, goals and understanding, an inquiry into human knowledge structures. Hillsdale, NJ: Lawrence Erlbaum.
Schwandt, T. (2007). Dictionary of qualitative inquiry (3rd ed.). London: Sage.
Smith, J. A., Harre', R., & Van Langenhore, L. (Eds.). (1995). Rethinking methods in psychology. London: Sage.
Sutherland, E. H., & Cressey, D. R. (1978). Criminology (10th ed.). Philadelphia: Lippincott.
Toby, J. (1962). Criminal motivation, a sociocultural analysis. British Journal of Criminology, 2(4), 317-336.
van Hiel, A., Hautman, L., Cornelis, I., & de Clerq, B. (2007). Football hooliganism: Comparing self-awareness and social identity theory explanations. Journal of Community & Applied Social Psychology, 17, 169-186.
Veneziano, C., & Veneziano, L. (1988). Knowledge of social skills among institutionalised juvenile delinquents: An assessment. Criminal Justice & Behaviour, 15(2), 152-171.
Vermeiren R., Bogaerts, J., Ruchkin, V., Deboutte, D., & Schwab-Stone, M. (2006). Subtypes of self-esteem and self-concept in adolescent violent and property offenders. Journal of Child Psychology & Psychiatry, 45(2), 405-411.
Welsh, B., & Farrington, D. (2006). Preventing crime: What works for children, offenders, victims and places. Dordrecht, Netherlands: Springer.
West, D. J., & Farrington, D. P. (1977). The delinquent way of life. London: Heinemann Educational Books.
University of Huddersfield, Huddersfield, United Kingdom
University of the West of England, Bristol, United Kingdom
Dr Helen Gavin is Principal Lecturer in Forensic Psychology at the University of Huddersfield, UK. Correspondences regarding this article should be addressed to: T: +44 (0)1484472789; E-mail: firstname.lastname@example.org.
David Hockey is a forensic psychologist in training in Bristol, UK.
Copyright 2010: Helen Gavin, David Hockey, and Nova Southeastern University
Gavin, H., & Hockey, D. (2010). Criminal careers and cognitive scripts: An investigation into criminal versatility. The Qualitative Report, 15(2), 389-410. Retrieved from http://www.nova.edu/ssss/QR/QR15-2/gavin.pdf…