Number of Public Laws (Per Congress, divided by 2); expected direction positive
Dependent Variable: Whether the first chamber successfully overrides the veto (0-No, 1-Yes)
Table 8:
Logistic Regression Analysis of Second House Override Model, Truman
Presidency
Variable Coefficient SE Z Mean
Constant -20.47 14.85 -1.38
Year Within Term 2.79 1.63 1.71* 3.13
First House Override % .16 .13 1.23 81.60
_____________________________________________________________________________
_
N=15
Pseudo R2=.32
Chi2 (2)=5.60*
Mean of Second House Overrides=.73 (11/15)
Percent Predicted Correctly=.80 (12/15), model increases predictive accuracy of 1 decision
*significant at .10
Variable Specification of Presidential Support Model (Reduced Version)
Year Within Term (1 to 4); expected direction positive
First House Override Percentage: expected direction positive
39