Abstract

This dissertation proposes the discussion of four empirical studies which show the validity of a new perspective within the social identity approach – the expanded model of organizational identification – and its applications in organizations. The social identity literature provides an important framework for understanding the reciprocal relationships between organizations and their employees, but traditional identity approach has largely neglected a new notion of organizational identification that includes other forms of attachment to the organization. Thus, the purpose of the present contribution is twofold: first, examining the validity of this new expanded perspective, which integrates organizational identification with other notions of identification, namely ambivalent identification, neutral identification and disidentification; and second, investigating its applications in organizational contexts, specifically by investigating how this approach is related to individuals’ outcomes. The first empirical study aims to provide further evidence for this model and to show discriminability and reliability of an Italian-language translation of Kreiner and Ashforth’s scales of a short version of the expanded model of organizational identification (EMOI) in an Italian speaking sample. Results provide good scale reliabilities, and confirmatory factor analyses demonstrate the good factorial validity of the short measure. We also tested the discriminant validity between organizational identification and affective commitment and we examined several antecedents of the four forms of identification. The second empirical study contributes to the understanding of the role of organizational identification for work-related stress by exploring the interactions between organizational identification and the other "problematic" dimensions of the expanded model in predicting employee strain. We hypothesized that ambivalent identification, neutral identification, and disidentification would moderate the negative relationship between organizational identification and exhaustion and ego depletion, such that the link between identification and strain would be stronger when the other dimensions are low. Results largely supported the hypotheses and show reliable interactions for disidentification and neutral identification and marginally significant moderation effects for ambivalent identification. Finally, we tested the interaction effects with a different outcome. Specifically, in the third study, we predicted a negative relationship between organizational identification and counterproductive work behaviors and a moderation of this link by ambivalent identification. We explored both overall counterproductive work behaviors (CWB) and also CWBs toward the organization (CWB-O), and CWBs toward other individuals (CWB-I). A survey-based study of 198 German employees revealed a moderating effect of ambivalent identification on the negative relationship between organizational identification and CWB, and CWB-O. Employees highly identified with their organization reported lower levels of CWB and CWB-O but - and as predicted - only when ambivalent identification was low. We then replicated the study examining CWB, and a scenario study of 228 American employees supported the previous findings: when organizational identification was high, participants in the low ambivalent condition reported lower levels of CWB-O than participants in the high ambivalent condition. The moderating effect of ambivalent and organizational identification was not significant on CWB-I in both studies. These findings provide new evidence for the positive influence of organizational identification under conditions of low ambivalence on counterproductive behaviors toward the organization. Limitations and practical implications of all studies are discussed.