We examine a mediated model in whic

We examine a mediated model in which feedback from family members is posited to influence commitment to family both directly and indirectly through satisfaction with family. These effects are moderated by either gender or family centrality to demonstrate the use of categorical and continuous moderator variables, respectively. Feedback was measured with five items that described reactions to performance of the family role (e.g., My family thinks what I do at home is outstanding). Satisfaction with family was measured with three items that described positive feelings toward the family (e.g., In general, I am satisfied with my family life). Commitment was measured with eight items that described psychological attachment to the family (e.g.,I feel a great sense of commitment to my family). Family centrality was assessed with six items that described the importance of family to life as a whole (e.g., The most important things that happen in life involve family). All items used 7-point response scales in which higher scores represented greater endorsement of the item (1 strongly disagree, 7 strongly agree). Prior to analysis, all continuous measures were mean centered, whereas gender was coded 0 for men and 1 for women (Aiken & West, 1991).