Working Paper Series: Special Edition of 2016 to 2018 Interns

Diagnostics Model 1 ECCU Sub-Sample

Table. 17 F-Test Baseline Logit Null Hypothesis:β1- β15=0 Alternative Hypothesis: Not H o

P-Value

Decision

F-Statistic

0.0001***

The coefficients are not simultaneously equal to zero as such including them in the model improves it fit.

* p<0.1, ** p<0.05, *** p<0.01

Table. 18 Goodness of Fit Test Null Hypothesis: The Model is Correctly Specified Alternative Hypothesis: The Model is incorrectly Specified P-Value

Decision

Pearson χ 2

0.4013

There is insufficient evidence to reject the null, as such it can be concluded that the model is correctly specified. Like the Pearson goodness of fit test there is insufficient evidence to reject the null at all conventional levels, hence the model is a good fit.

Hosmer-Lemeshow

0.4608

* p<0.1, ** p<0.05, *** p<0.01, Hosmer-Lemeshow 8 groups.

47