The role of organizational attachment as a moderator of procedural justice perceptions: The self-interest model vs. the group-value model

Date of Completion

January 1997


Psychology, Industrial|Sociology, Industrial and Labor Relations




Procedural justice research in organizations, which examines the fairness of processes used to arrive at decisions, has focused primarily on two models of subjective procedural justice: the self-interest model and the group-value model. The self-interest model proposes that people's perceptions of fairness are based on how much control they have in the decision-making process whereas the group-value model predicts that people favor procedures that strengthen their security about group membership and their status within the group. Previous research has determined that the group-value model is generally a better predictor of procedural justice perceptions; however, control is still a valuable predictor.^ The current study tests the Tyler and Dawes' (1993) prediction that each model may dominate perceptions in particular settings or for particular subsets of individuals (e.g., the involvement or attachment that people have within a group).^ It is hypothesized that level of attachment, measured by organizational commitment and group cohesiveness, moderates the relative efficiency of the two models in explaining perceptions of procedural justice. Specifically, at low levels of attachment, the self-interest model dominates employees' perceptions of procedural justice and at high levels of attachment, the group-value model dominates employees' perceptions of procedural justice.^ Results indicate that both models accounted for significant variance in procedural justice perceptions. Also, consistent with previous research, the group-value model accounted for significantly more variance than the self-interest model. Results did not, however, support the hypothesized moderating role of attachment.^ In follow-up analyses, a stepwise regression conducted to measure the individual contributions of each predictor (regardless of the model to which it belonged) indicated that Trust (a component of the group-value model) was entered first and accounted for 60% of the variance while Process Control (a component of the self-interest model) was entered next and accounted for an additional 8% of the variance. The total variance accounted for by the full model was 75%. These results suggest that both models of subjective procedural justice contribute to predictions of procedural justice judgments. The lack of a moderating effect with attachment does not preclude the existence of other potential moderators, which is a suggested course for future research. ^