تکانشگری، اسناد و قلدری در زندان: روابط متقابل گروه قلدر و مجرم قربانی

The current study explores bullying behaviours among adult male prisoners, examining its relationship with aggression attribution and impulsivity. Employed are two separate methods of analysis to determine how this may influence results. Participants were 102 prisoners. All completed a revised version of the Direct and Indirect Prisoner behaviour Checklist (DIPC-R), the Barratt Impulsivity Scale: Version II (BIS-12) and the Expressive Aggression Scale (EXPAGG). Analysis included categorical analysis with prisoners placed into one of four groups (pure bully, pure victim, bully/victim and not-involved), and factorial analysis where perpetration and victimisation were assessed as continuous variables and evidence of interactions explored. It was predicted that perpetration would be associated with higher instrumental attributions and higher impulsivity than non-perpetration. It was predicted that a factorial analysis would demonstrate no interactions between perpetration and victimisation across aggression variables, questioning the utility of a distinct ‘bully–victim’ group. Bullies were found to have higher instrumental attribution scores than non-bullies, with no differences for expressive attribution. Victims were more impulsive than non-victims with evidence that perpetration moderated this relationship. A categorical analysis demonstrated that bully/victims were more impulsive, at least in relation to pure bullies. Results suggested that it was the combined effect of indirect and direct aggression which promoted differences between victims and bullies in relation to attribution and impulsivity. Results are discussed with reference to previous research concerning prison bullying, with directions for future research focused on exploration of perpetrator–victim mutuality using a range of variables and distinct methods of analysis.

As an area of academic study, the examination of bullying behaviour among prisoners has had a brief history in comparison to the study of bullying occurring in other settings, such as schools. The first studies exploring prison bullying were published in 1996 (Ireland and Archer, 1996 and Connell and Farrington, 1996), with a marked increase in research since 1999. Since this date, there have been 24 studies published (e.g. Palmer and Farmer, 2002, Ireland and Archer, 2004, Palmer and Thakordas, 2005 and Ireland and Monaghan, 2006). Prison-based research has moved in recent years from a focus on the nature and extent of bullying to an examination of five core areas of interest; refining methods of measurement; recognising the role of the social and physical environment; distinguishing between the different groups involved; predicting group membership and, finally, linking prison research into the wider aggression literature and theory base (Ireland & Qualter, 2006). It is these latter three areas that the current study is focused on, namely an exploration of the groups involved and their distinguishing characteristics, and how prison bullying research can be aligned more to the aggression literature. Whereas school-based bullying research has developed more closely alongside the wider aggression field, clearly defined as a specific subsection of aggression (Olweus, 1996), prison research has developed differently. The latter has utilised methods of measurement that encapsulate a wide range of behaviours, not all of which have been immediately identified as aggression i.e. indirect aggression such as ostracising and spreading rumours, and other forms of more direct aggression such as theft (Ireland and Archer, 1996 and Archer et al., 2007). The focus on the full remit of behaviours that can be classified as aggression has led to prison researchers opting for broad definitions of bullying, an example of which has been offered by Ireland (2002) as follows: “An individual is being bullied when they are the victim of direct and/or indirect aggression happening on a weekly basis, by the same perpetrator or different perpetrators. Single incidences of aggression can be viewed as bullying, particularly where they are severe and when the individual either believes or fears that they are at risk of future victimisation by the same perpetrator or others. An incident can be considered bullying if the victim believes that they have been aggressed towards, regardless of the actual intention of the bully. It can also be bullying when the imbalance of power between the bully and his/her victim is implied and not immediately evident (p. 26).” Addressing definitional issues concerning bullying are a separate paper on their own, although it is suffice to say that even within the more heavily researched areas of bullying study (e.g. schools) there remains no universally accepted definition (Smith & Brain, 2000). This has led some to argue that a fixed, measurable definition of bullying may not exist (Ireland, 2005a), and instead that we should be focusing on defining bullying by what motivates it as opposed to how it manifests itself. What is particularly interesting, however, is the afore mentioned issue of prison bullying not tending to be researched alongside standard aggression, or aggression-related, measures. A number of studies involving adults and children have found that bullies show the expected higher scores on measures of direct aggression, such as violent crime (Farrington, 1993), proactive and reactive aggression (Roland & Idsøe, 2001), and physical aggression (Craig, 1998). This evidence base suggests, therefore, that there are expected links between aggression and bullying worthy of further examination. Extending the exploration of aggression related variables to prison bullying is of interest, particularly with a view to confirming if the behaviour examined within prisons can be linked more broadly into the aggression literature. There is a risk that research conducted within forensic (prison) samples may develop in isolation from the wider academic research base unless there is a concerted effort to align forensic research with aggression research. To date only three studies have begun to explore the role of aggression in prison bullying (Ireland and Archer, 2004, Palmer and Thakordas, 2005 and Archer et al., 2007). All have reported a relationship between aggression and bullying. Ireland and Archer (2004) found, among young and juvenile prisoners, that all four subscales of the Aggression Questionnaire (AQ: Buss & Perry, 1992) were moderately correlated with measures of different forms of bullying. Similar findings were reported in a study of adult offenders (Palmer & Thakordas, 2005). Both Ireland and Archer (2004) and Palmer and Thakordas (2005) also employed a participant group-classification approach to explore how aggression measures related to bullying behaviour, whereby prisoners were classified into one of four categories; pure bullies (sole perpetrators); pure victims (sole victims); bully/victims (mutual perpetrator–victim category) and those not-involved. This approach has been commonly applied to prison (see Ireland, 2002) and school-based studies (e.g., Baldry and Farrington, 1998, Boulton and Smith, 1994, Craig, 1998 and Unnever, 2005). In Ireland and Archer (2004) prisoners classified as pure bullies or bully/victims reported higher levels of physical and verbal trait aggression than other prisoners, with bully/victims reporting higher levels of hostility and anger than the other categories. In Palmer and Thakordas (2005), categorical differences were limited to the hostility subscale with bully/victims presenting with higher levels of hostility than those not-involved. In a more recent study among adult prisoners, Archer et al. (2007) extended exploration of bullying behaviour and aggression related-variables to an assessment of aggressive attributions using the EXPAGG (Expressions of Aggression), and impulsivity. Both pure bullies and bully–victims showed significantly higher values than the not-involved category with regards to impulsiveness. Only bully–victims showed higher expressive attributions than the not-involved category. A relationship between aggression variables and bullying should not be surprising if theories outlined to explain prison bullying are supported. Ireland (2004) postulated, for example, that within prisons, pure bullies may be largely proactive aggressors (i.e. planned/controlled) whereas bully/victims may display aggression driven more by reactive motives (i.e. by negative emotions such as anger, fear, hostility). Ireland (2004) cited the findings of Ireland and Archer (2004), where bully/victims presented with higher levels of anger and hostility, as supporting evidence for this. It was further theorised that aggression demonstrated by bully/victims may be motivated by a drive to prevent their own future victimisation by communicating to their peer group that they are not ‘easy targets’. This was extended further in 2005 with the development of the Applied Fear Response [AFR] model (Ireland, 2005b). This model described how fear can drive reactive responses to bullying, with such responses (particularly those by bully/victims) including aggression towards others. The AFR model argued further that reactive responses can be both immediate and delayed, with the latter occurring sometime after the provocation. However, although reactive aggression in this instance was thought to be driven primarily by a negative emotion (e.g. fear), it does arguably have a proactive goal, namely that of aggressing towards peers in order to communicate that they are likely to fight back in the future. In this way, reactively-aggressing can be considered an attempt to restore self-image and to prevent the stigmatization that being labeled as a victim who does not fight back is likely to carry (Ireland, 2004). Thus, the aggression of bully/victims may be mixed motive (Bushman & Anderson, 2001) and not easily classified as either reactive or proactive. Furthermore, if it was accepted that those reporting both perpetration and victimisation (i.e. bully/victims) were more likely to demonstrate reactive motivations in relation to their use of aggression than the more proactively-motivated perpetrator (i.e. pure bullies), predictable differences in aggression attribution should also be evident. The question would then become whether or not aggression attributions would relate more to expressive attributions (i.e. a loss of control) or instrumental attributions (i.e. that the aggression was necessary). These two types of attribution have been routinely researched in the aggression literature (e.g. Campbell, Muncer, McManus & Woodhouse, 1999), with instrumental beliefs found to associate more with measures of aggression than expressive beliefs ( Archer and Haigh, 1997 and Smith and Waterman, 2006), and particularly with physical aggression (e.g., Alexander et al., 2004, Archer and Graham-Kevan, 2003 and Ramirez et al., 2001). On the basis of this it could be argued that both perpetrator groups (i.e. pure bullies and bully/victims) should be expected to hold more instrumental attributions than the other groups (i.e. pure victims and those not-involved). It has further been suggested that bully/victims may be more likely to hold expressive attributions than pure bullies. This is based on the findings of Archer et al. (2007) which has offered some confirmation for increased expressive attributions among bully/victims, but only in comparison to those not-involved and not in comparison to other perpetrators. As noted, there is evidence that bully/victims demonstrate higher levels of hostility (Ireland & Archer, 2004; Palmer & Thakordas, 2005) and anger (Ireland & Archer, 2004) which would equate more with expressive attributions, although increased anger was not found by Palmer and Thakordas (2005). Evidence of increased impulsivity could provide further evidence for an expectation for higher expressive attributions, since such attributions focus on reduced control. Impulsiveness has been consistently associated with higher levels of direct aggression (Barratt, 1994, O'Connor et al., 2002 and Smith and Waterman, 2006), and involves a lack of inhibition and the inability to delay immediate responses to a situation. There has been some support for increased levels of impulsivity among bully/victims (Archer et al., 2007), but this has also been demonstrated for pure bullies (Archer et al., 2007). On the basis of the (limited) research to date, therefore, it would appear to suggest that predicting evidence of increased instrumental attributions and increased levels of impulsivity for both perpetrator groups (pure bullies and bully/victims) is most likely to be confirmed. The findings in relation to expressive attribution and the distinction between the perpetrator groups on this variable appear equivocal. An interesting point to make, however, relates to the methods of analysis employed in prison studies to date. It is possible that the equivocal findings across perpetrator groups in relation to some aggression-related variables may be an artefact of the specific method of analysis adopted. The majority of research has accepted a four-group classification approach and conducted analyses accordingly (e.g. across the categories of pure bully, pure victim, bully/victim and not-involved). The reliability of this classification approach in terms of its application to analysis has been questioned (Archer et al., 2007). To date only one study has explored a different approach to analysing prison bullying data: Archer et al. (2007) opted for an extended method of analysis whereby a factorial approach was also employed to explore the association between perpetration and victimisation using main effects, with interactions used to examine if mutual perpetration and victimisation (i.e. bully/victims) did relate to aggression variables. The factorial analysis demonstrated that those classed as bullies showed higher levels of impulsiveness, instrumental and expressive attributions, whereas those classed as victims showed higher scores than non-victims on impulsiveness. There were no interactions. Archer et al. (2007) argued, therefore, that all of these differences between the four categories could be attributed to either a main effect of bullying or a main effect of victimisation or the additive effects of both. This was argued to rule out the possibility that being a victim moderates the impact of being a bully on any of the measures used, or that being a bully moderates the impact of being a victim. The current study aims to add to the existing literature base by exploring if aggression-related variables, namely aggression attribution and impulsivity, are associated with bullying behaviour among a sample of adult male prisoners. This area has received limited attention to date, with only one previous study exploring attribution and impulsivity. Importantly, the current study aims to add to the existing research by exploring the value of the four-classification approach routinely used to assess the characteristics of prisoners involved in bullying. Only one study to date (Archer et al., 2007) has attempted to do this and thus there is a need to attempt to confirm or disconfirm the findings of Archer et al. If it can be further confirmed that there are no characteristics of being both a bully and a victim beyond those that can be predicted from the main or additive effects of being a bully or a victim, then this would question reference to ‘bully–victims’ as a separate category. This category has received considerable attention in prison bullying research to date, with it argued that it is one of the most frequently occurring and most observable groups (Ireland, 2002). Confirming, therefore, that the difference between perpetration and victimisation can be predicted by the main or additive effect of being a bully or a victim would have implications for anti-bullying programs that base intervention on a four-category classification approach, and for future studies aiming to explore bullying behaviour. Adult male prisoners were required to complete a self-report checklist of bullying behaviour, a measure of aggression attribution and an assessment of impulsivity. In summary, there were three core predictions tested in the present study, namely that: 1. All forms of bullying will be associated with instrumental attributions, with higher instrumental attributions expected among perpetrators than non-perpetrators (based on Archer and Haigh, 1997, Alexander et al., 2004, Archer and Graham-Kevan, 2003 and Ramirez et al., 2001). 2. Based on research demonstrating that impulsivity has been consistently associated with higher levels of direct aggression (Barratt, 1994, O'Connor et al., 2002 and Smith and Waterman, 2006), it is expected that impulsivity will be higher among perpetrator than non-perpetrator groups. 3. Based on the findings of Archer et al. (2007), a factorial analysis will demonstrate no interactions between perpetration and victimisation across aggression attribution and impulsivity.