coefficients you see that both the centered Hassles and the interaction terms are signifi-
cant (p 5 .000 and .037, respectively), but the social support variable is not significant.
By convention we leave it in our regression solution, because it is involved in the inter-
action, even though the associated tvalue shows that deleting that variable would not
lead to a significant decrease in R^2.
Our regression equation now becomes
YN 5 .086 chassles 1 .146 csupport – .005 chassupp 1 89.585.15.14 Mediating and Moderating Relationships 559Hassles
1.000
–.167
.577**
.910**
1.000**
–.167
–.297*Hassles
Support
Symptoms
hassupp
chassles
csupport
chassuppSupport
–.167
1.000
–.134
–.510**
–.167
1.000**
.402**Symptoms
.577**
–.134
1.000
.585**
.577**
–.134
–.391**hassupp
.910**
–.510**
.585**
1.000
.910**
–.510**
–.576**chassles
1.000**
–.167
.577**
.910**
1.000
–.167
–.297*csupport
–.167
1.000**
–.134
–.510**
–.167
1.000
.402**chassupp
–.297*
.402**
–.391**
–.576**
–.297*
.402**
1.000CorrelationsPearson Correlation
** Correlation is significant at the .01 level (2-tailed).
* Correlation is significant at the .05 level (2-tailed).Sum of
Squares
9427.898
14839.816
24267.714Model
1 Regression
Residual
Totaldf
3
52
55Mean
Square
3142.633
285.381F
11.012Sig.
.000aANOVAbCoefficientsaModel
1 (Constant)
chassles
csupport
chassuppB
89.585
8.594E-02
.146
–5.06E-03Std. Error
2.292
.019
.305
.002Beta.509
.057
–.262t
30.094
4.473
.479
–2.144Sig.
.000
.000
.634
.037Unstandardized
CoefficientsStandardized
CoefficientsaPredictors: (Constant), chassupp, chassles, csupport
bDependent Variable: SymptomsaDependent Variable: SymptomsTable 15.5 Regression solution for moderated relationship between hassles
and symptomsR
.623aModel
1R Square
.388Adjusted
R Square
.353Std. Error
of the
Estimate
16.8932Model SummaryaPredictors: (Constant), CHASSUPP, CHASSLES, CSUPPORT