«Jo-Anne Wemmers Katie Cyr Université de Montréal While victims are often considered the forgotten party in the criminal justice system, restorative ...»
Four of the five predictor variables are significantly correlated with victim’ procedural justice judgements. Victims' evaluation of how they were treated by the project workers during the preparatory meetings was the only factor not significantly related to their fairness judgements. However, victims' evaluations of their first contact with project workers were related to the procedural justice judgements, suggesting that first impressions are imperative. Also, whether they obtained the information they desired (Chi-square = 14,857, df = 1, p 0.01) and whether they were adequately informed about the project (Chi-square = 12.438, df = 1, p 0.01) were both significantly related to their fairness judgements. Finally, how they were treated by the mediator was significantly related to their fairness judgements (Chi-square = 30.807, df = 1, p 0.01).
The relationship between the three possible determinants of trust (whether they had faith in the mediator, whether they felt safe, and whether they felt understood by the © Applied Psychology in Criminal Justice, 2006, 2(2) 118 FAIRNESS project worker) and procedural justice was then examined. The results are presented in Tables 11 through 13.
The results show that two of the three variables were significantly related to victims' procedural justice judgements.
Victims who said that they did not have faith in the mediator were less likely to judge the procedure as fair than those who did have
faith in the mediator (Chi-square = 36.580, df=1, p 0.01). While victims who felt understood were more likely to judge the procedure as fair than those who did not feel understood (Chisquare = 10,643, df=1, p 0.01), two of the four cells of the Chisquare test have an expected frequency of less than five observations, which brings into question the reliability of this test.
Only three victims said that they did not feel safe prior to the mediation sessions, yet all three felt that the procedure was fair.
Whether victims felt safe prior to mediation was not related to their feelings of fairness.
Next, the factor of neutrality was addressed. Here two possible determinants were examined: whether the mediator was neutral and whether the mediator favoured one party over the other. The results are presented in Tables 14 and 15. Both variables were significantly related to victims' procedural justice judgements.
© Applied Psychology in Criminal Justice, 2006, 2(2) 120 FAIRNESS Regression Analysis Multiple regression analysis shows the degree to which the variance in victims’ procedural justice judgements can be forecast by the predictor variables. In order to find out the relative importance of voice, respect, neutrality, and trust in determining victims’ procedural justice judgements, step-wise regression analysis was carried out. Step-wise regression was chosen because of the relatively strong relationships between the predictor variables. See Table 16 for the correlations between variables.
Step-wise regression uses the semi-partial correlations with the criterion and corrects for that portion of the variance which can be predicted by the other predictor variables.
Before conducting the regression, however, multi-item scales were created for voice, respect, trust, and neutrality. Voice was created using two items: whether the victim felt that he/she had been heard and whether or not he/she felt hindered in making demands. The resulting scale was very reliable (alpha=0.8625).
Two variables not included are whether the victim could make demands, which was found not to be related to victims' fairness judgements, and whether the victim had been able to express him/herself. The latter variable is significantly related to procedural justice, however, including it in the scale reduced the scale’s reliability (alpha=6850).
The variable respect was created using four items: victims’ evaluations of how they were treated at their first contact with project workers, whether they obtained all the information they desired, whether they had been adequately informed, and how respectful the mediator had treated them. The resulting scale was very reliable (alpha = 0.7901). A fifth element, namely victims evaluations of how they had been treated during the preparatory meeting, was not included in the scale as it was not significantly related to procedural justice and its inclusion would have decreased the scale’s reliability (alpha = 0.6945).
Trust was based on the combination of two items: whether victims felt understood by the mediator, and whether they had faith in the mediator. The resulting scale was very reliable (alpha =
0.7679). A scale for neutrality was created by combining the two items of neutrality of the mediator and absence of bias. These two items resulted in a very reliable scale (alpha = 0.9708). Table 17 provides an overview of the means and standard deviation for each of the variables included in the regression.
In order to get an idea of the relative importance of voice, respect, trust, and neutrality in relation to victims’ procedural justice judgements, step-wise regression was carried out using these four variables as independent variables and procedural justice as the dependant variable. In order to control for background variables, the respondent’s age and gender were also included in the analysis. In a first step, voice is included in the regression equation (R-squared =.816; p 0.01) and is able to explain 81% of the variance in victims’ procedural justice judgements. In a second step, the variable trust is included in the regression equation and combined with voice, these two variables are able to explain 88% of the variance in victims’ procedural justice judgements (Rsquared =.886; p 0.01). The remaining variables, respect and neutrality, as well as the background variables were not included in the regression equation.
Amount of Contact In order to examine Lind and Van den Bos’ (2002) assertion that fair procedures may be more time consuming, the amount of contact was also examined in relation to victims’ qualitative appraisals of their interactions. Specifically, victims were asked if they had many contacts with project workers leading up to a possible mediation. The results are presented in Table 18.
© Applied Psychology in Criminal Justice, 2006, 2(2) 122 FAIRNESS The number of contacts victims had with workers was not related to their fairness judgements. Victims who had many contacts with the project workers were more likely to believe that they had been adequately informed about what to expect (chi-sq = 5.481, df = 1, p=0.019). However, they were not more likely to feel that they obtained the information they desired (chi-sq =.592, df=1, p =.442) or that they were heard (chi-sq =.422, df = 1, p =.516). It appears that the quality of the interaction, and not the quantity, is most important for victims' procedural justice judgements.
DISCUSSION This study examined the circumstances in which victims feel that they are treated fairly. The findings show victims desire more than the ability to make demands. Most victims in the sample said that they had been able to make demands, but what mattered most was that they felt they were heard and were not hindered in making demands. The findings seem to support the point of view, which is commonly found in the victimological literature, that victims seek recognition by expressing their point of view and having their point of view taken into consideration (Kilchling, 1995; Shapland, et al., 1985; Wemmers, 1996). The findings suggest that voice is not just about expressing one’s needs but also, and perhaps more importantly, about being heard.
Regarding the determinants of procedural justice judgements, clearly the present study, which is based on post-test data only, does not provide a test of the different models developed by Lind & Tyler (1988), Tyler (2003), Tyler & Lind (1992), and Wemmers (1996). The results of the regression analysis show that victims' procedural justice judgements are largely determined by the amount of voice victims feel that they have in the process as well as their trust in authorities. The process of mediation was viewed as fair when victims felt that they were sufficiently able to make their point of view heard, when they did not feel hindered in making certain demands, when they felt understood by the mediator, and when they had faith in the mediator. In contrast, when victims felt that they had not been able to make themselves
heard, felt hindered in making demands, felt misunderstood by the mediator, and had little faith in him/her, they were more likely to feel that the process was unfair.
The variables of respect and neutrality were not included in the regression equation. However, we cannot conclude that these variables are not important as they are both strongly correlated with procedural justice. The predictor variables are all highly intercorrelated, and while they are strongly related to procedural justice, they add little to the equation in terms of predictive power.
Taken together, the results from the regression analysis and the chi-squared analyses suggest that victims find mediation fair because it offers them recognition and respect through consultation, not because it allows them to make demands. These findings seem to contradict certain advocates of restorative justice, such as Fattah (2001) and Roach (1999), who argue that restorative justice appeals to victims' sense of fairness because it allows them to make demands. Beyond making demands, the quality of interaction with project workers is important to victims' perceptions of fairness. The present findings suggest that the ongoing legal debate about victim participation in the rights of the accused (see Ashworth, 2000 and Roach, 1999) may be superfluous. The way in which outcomes are received is important to victims (Wemmers, 1996), and in order to be fair, procedures must include victims while being impartial and free from bias.
With respect to mediation programs, the findings suggest that the contact between project workers and victims is key. It is not enough that programs offer victims input; the quality of the interaction with project workers has a significant impact on victims’ procedural justice judgements. From the very first contact, project workers have to develop a good rapport with the victim.
They must allow the victim sufficient opportunity to express him/herself, give the victim the feeling that he/she has been heard, and communicate that they understand their point of view. They must establish a sense of trust between the victim and themselves, while remaining neutral and impartial. These findings underscore the importance of proper training for project workers.
© Applied Psychology in Criminal Justice, 2006, 2(2)124 FAIRNESS
From a managerial perspective, the finding that qualitatively positive interactions are not synonymous with timeconsuming interactions is of great importance. Lind and Van den Bos (2002) expressed concern that time constraints may be a major impediment to procedural justice in organizational settings.
However, the present study shows that victims’ evaluations of the quality of the interpersonal interactions with project workers are not systematically related to the number of contacts. The number of contacts between victims and project workers was not correlated with their procedural justice judgements. Victims who had several contacts with project workers were, however, more likely to feel that they had been adequately informed. The lack of relationship between number of contacts and procedural justice judgements suggests that criminal justice professionals can invest in fairness at relatively little cost in terms of human resources; thus, time constraints should not impeded procedural justice.
In return for their investment in procedural fairness, criminal justice professionals can reap many benefits. Tyler and Huo (2002) argue that perceptions of fairness are vital to securing confidence and cooperation with the criminal justice system.
Research shows that victim collaboration with the justice system is low and victims who have previously had negative experiences with the justice system are less likely to report future criminal acts to the police (Besserer and Trainor, 2000; Shapland, et al., 1985;
Van Dijk, 1999). By treating victims with respect and recognition, programs like victim-offender mediation can enhance victims’ faith in the criminal justice system and their willingness to collaborate with authorities.
Moreover, because they appeal to victims’ sense of justice, procedures that allow victims to be heard can effectively reduce the risk of secondary victimisation. Insensitive reactions by authorities to victims can augment the victim’s suffering (Maguire, 1991). Several authors have criticized the conventional criminal justice system because of its failure to provide victims with a formal role other than that of a witness (Fattah, 2001; Langevin 2002; Roach, 1999; Shapland, et al., 1985). Providing that project workers are able to develop a good rapport with the victim, victim
offender mediation can provide victims’ with the recognition and respect that they seek to restore their sense of justice and reduce their suffering (Wemmers and Cyr, 2005).
There are noteworthy limitations to this study. It was based on a small sample and therefore the findings cannot be generalized to victims in general. The study needs to be replicated, using a larger and more diverse sample. Furthermore, the overrepresentation of victims who participated in mediation may have affected the results. Further research, which addresses the limitations of the present study, is needed before conclusions can be drawn about victims in general.
Adams, J.S. (1965). Inequity in social exchange. In L. Berkowitz (ed.), Advances in Social Psychology, Volume 2. (pp. 267-299). New York: Academic Press.
Ashworth, A. (2000). Victims’ rights, defendants’ rights and criminal procedure. In A.
Crawford, and J. Goodey (eds.), Integrating a victim perspective within criminal justice (pp. 185-206). Aldershot: Darmouth Publishing.