or predictor variable. There were four steps in the stepwise regression analysis, producing

a final model explaining the variance in role dissatisfaction with an R2 of .375. The first

step in the analysis identified child-related task differentiation as a significant predictor

variable (P = .488, t = 6.367, p < .001). The child-related task differentiation variable had

an R2 of .238 and was a significant predictor in the initial model (F (1, 130) = 40.536, p <

.001). The second step added couples' averaged violated expectations as a predictor in

the second model (R2= .294, F(1, 129) = 26.870, p < .001), resulting in an R2 change of

.056. The hostile couple conflict type score was added as a third predictor in the next

model resulting in an increased R2 of .342 and an R2 change of .048; the model continued

to be significant (F (1, 128) = 22.199, p < .001). The fourth and final model added

individual well-being as a fourth predictor of role dissatisfaction resulting in an R2 of .375

and a change in the R2 of .032. This model also was significant (F (1, 127) = 19.012, p <

.001). The final model accounted for 37.5% of the variance in role dissatisfaction. The

results of the stepwise regression analysis are presented in Table 4-3 and the model's

coefficients are found in Table 4-4.

Table 4-3. Role Dissatisfaction Stepwise Regression Model Summary
Step Variable R R2 Radj R2chg F chg dfl df2
1 TD (C) .488 .238 .232 .238 40.536 < .001 1 130
2 EXP-C .542 .294 .283 .056 10.303 < .001 1 129
3 HSTL .585 .342 .327 .048 9.371 < .001 1 128
4 WLB .612 .375 .355 .032 6.558 < .001 1 127


Table 4-4. Role Dissatisfaction Model Coefficients
Model B 3 t Significance
1 Constant 2.475 7.478 < .001
 TD (C) .046 .488 6.367 < .001
 2 Constant 2.717 8.271 < .001
 TD (C) .036 .387 4.822 < .001
 EXP-C -.425 -.258 -3.210 .002