For mothers, the indirect effect of paternal alcoholism on warmth and sensitivity through marital satisfaction was significant. When the direct effect of father's alcoholism on 36 month maternal warmth and sensitivity was included in the model, the fit of the model did not improve. Furthermore, this path was not significant. Thus, marital satisfaction mediated the association between paternal alcoholism and warmth and sensitivity. For fathers, the indirect effect of paternal alcoholism on warmth and sensitivity through marital satisfaction was also significant. However, when the direct effect of paternal alcoholism on warmth and sensitivity was included in the model, the fit of the model significantly improved. Thus, marital satisfaction partially mediated the association between father's alcoholism and negative parenting. Interestingly, the association between father's alcoholism and warmth and sensitivity was significant and in the opposite direction to what was predicted: paternal alcoholism predicted increases in paternal warmth and sensitivity at 36 months. However at the correlational level paternal alcoholism was not associated with negative parenting at 36 months. Thus, it is likely that a suppression effect was occurring with fathers.
