Table 4-8. Husbands' Violated Expectations Model Coefficients
Model B P t Significance
1 Constant -1.609 -3.599 < .001
WLB-I .016 .225 2.630 .010
2 Constant -1.080 -2.199 .030
WLB-I .015 .209 2.479 .014
VOL-I -.200 -.202 -2.400 .018
Question 3: Do task differentiation (of family household tasks and child-related
tasks), role dissatisfaction, violated expectations, individual well-being, age, infant age,
income level, and each of the couple conflict type scores (volatile, validating, conflict-
avoiding, and hostile) serve as predictors of marital disaffection?
A stepwise multiple regression was conducted with marital disaffection designated
as the dependent variable, and the set of couple-level variables (household task
differentiation, child-related task differentiation, role dissatisfaction, couples' averaged
violated expectations, individual well-being, age, infant age, income level, and the
volatile, validating, conflict-avoiding, and hostile couple conflict type scores) designated
as the independent variable. There were six steps in the stepwise regression analysis,
producing a final model accounting for variance in marital disaffection with an R2 of
.417. The first step in the analysis identified role dissatisfaction as a significant predictor
variable (3 = .406, t = 5.063, p < .001). The role dissatisfaction variable had an R2 of .165
and was a significant predictor in the initial model (F (1, 130) = 25.637, p < .001). The
couple's average age was added as a predictor in the second model resulting in an
increased R2 of .260 and an R2 change of .095; the model continued to be significant
(F (1, 129) = 22.686, p < .001). The third step added couples' individual well-being as a
predictor in the third model (R2 = .327, F (1, 128) = 20.732, p < .001), resulting in an R2
change of .067 over the second model. The hostile couple conflict type score was added