55 Predicting the Effectiveness of the Home-Start Parentingg Support Program Asscher,, J.J., Hermanns, J.M.A., & Dekovic, M. (submitted for publication) TheThe current study examines predictive effects of participant characteristics, programprogram characteristics and their interaction, on changes in the parenting behaviorbehavior of mothers who participated in the Home-Start parenting support program.program. The results confirm previous findings that effects of participant and programprogram characteristics differ by outcome variable. Overall, participant predictorspredictors seemed to have little effect: of the eight predictors, only five (gender of thethe child, income, perceived social support, single parenthood and education of thethe mother) had an effect on some outcome measures, thus suggesting that Home- StartStart is equally effective for all participants. The only program characteristic that affectedaffected changes in two outcomes (rejection and negative control) was the programprogram intensity. Also, the same program characteristic (intensity) interacted withwith several participant characteristics in predicting changes in maternal rejectionrejection and negative control. These results show the utility of examining interactionsinteractions between participant and program characteristics when studying the effectseffects of family support programs. 5.11 Introduction Homee visiting programs for families with young children are widely used. The generall aim of these programs is to support families and prevent family disfunctioning.. Research on the evaluation of home visiting programs has shown somee promising effects for parents and children (Durlak & Wells, 1997; Olds et al.,, 1997; Olds, Henderson, Chamberlin, & Tatelbaum, 1986; Olds & Kitzman, 1993).. Effects are, however, limited in magnitude and the utility of home visiting hass not yet been clearly demonstrated (Sweet & Appelbaum, 2004). One explanationn for the modest effects might be that most effectiveness studies try to findfind overall effects, whereas not all families are likely to benefit from one programm or one approach (Gomby, Culross, & Behrman, 1999). As Farrell, 57 7