This preview shows page 1. Sign up to view the full content.
Unformatted text preview: crease by
51.81,44.28, and 5.31, respectively for Models I, II, and IU. In addition,
the inclusion of the interaction terms resulted in a change in chi-square
values by 103.62, 88.56, and 10.64, espectively for Models I, 11, and III,
with a corresponding changeof 18, 12, and 6 degrees of freedom (also see
Table 2). The changes in chi-square values are significant at the .001 level
for Models I and II suggesting strong evidence for including interaction
terms in the choice model, and at the . lo1 level for Model ID, suggesting
moderate evidence for including the interaction termsin the choice model.
The results of the multinomial logistic regression analysis are presented in
Table 2. The logistic regression results evaluate the effects interacting
variables when bothare high or low. However, they cannot distinguish the
effects of the interacting variables when one of the variables has a high
value and the other a low value. These effects were therefore evaluated
using chi-square analysis after split...
View Full Document
- Summer '14