Conflict Escalation in Intimate Relationships

Duration of Research: 2003-2006

Funding Agency: Israel Scientific Foundation (ISF)

Zeev Winstok (P.I.)
Guy Enosh (P.I.)
Zvi Eisikovits (P.I.)
Eila Perkis (Research Coordinator)

Project Description:
Current research in intimate violence focuses primarily on the frequency and severity of violent acts and their outcomes. These studies operationalize aggression in intimate relationships by using a single and isolated action as the frame of reference (as measured by Strauss et al., 1996 with the Conflict Tactic Scale). By focusing on outcome measures, the process by which aggression emerges remains largely unexplored.

The purpose of this study is to explore a theoretical model that integrates our previous theoretical conceptualizations (Eisikovits, Winstok, & Gelles, in press; Winstok, Eisikovits, & Gelles, in press) with the Planned Behavior Model (Ajzen, 2001). Such theoretical development holds promises to contribute in the development of reliable and valid measures and theory of conflict escalation in intimate relationships. The theoretical approach suggests that conflict escalation is neither random nor situational, but rather constructed in the context of dyadic relationships in which each side takes into account the other's actions. Thus we propose that the variables that are related in predicting an escalatory behavior in intimate relationship are: 1) intention to react in an escalatory manner; 2) perceived ability to perform in accordance with these intentions (called conflict management capabilities); and 3) partner's escalatory intentions as perceived by the actor (based on previous experience). Escalatory intention are predicted by 1) attitudes towards escalation; 2) perception of expectations of significant others regarding escalation (subjective norms); 3) conflict management capabilities, and 4) partner's escalatory intentions as perceived by the actor.

The study has a longitudinal panel design. It consists of three waves of data collection from the same set of cases with an interval of six months between waves. Two samples were drawn for the purpose of this study: One is a sample of heterosexual, cohabiting partners drawn from the general population (200 informants-100 men and 100 women), and the other is a sample of heterosexual men/women and their partners who participate in domestic violence programs (200 informants-100 men and 100 women). Data analysis includes both cross-sectional and longitudinal analyses such as confirmatory factor analysis, latent growth curve models, and change scores models.