Table 10.
Likelihood ratio test comparing probit versus logit models
| Likelihood criteria | Probit | Logit | L–P |
|---|---|---|---|
| Equation 2—FEAH | |||
| AIC | 3821.365 | 3815.994 | 5.370 |
| BIC (df = 31/31/0) | 3979.668 | 3974.297 | 5.370 |
| Result: difference of 5.370 in BIC provides positive support for logit model | |||
| Equation 3—FEAFH | |||
| AIC | 3429.889 | 3422.091 | 7.798 |
| BIC (df = 31/31/0) | 3588.193 | 3580.395 | 7.798 |
| Result: difference of 7.798 in BIC provides strong support for logit model | |||
| Equation 4—FEAH | |||
| AIC | 3809.935 | 3802.539 | 7.396 |
| BIC (df = 32/32/0) | 3973.185 | 3965.790 | 7.396 |
| Result: difference of 7.396 in BIC provides strong support for logit model | |||
| Equation 5—FEAFH | |||
| AIC | 3423.081 | 3414.133 | 8.948 |
| BIC (df = 32/32/0) | 3586.331 | 3577.383 | 8.948 |
| Result: difference of 8.948 in BIC provides strong support for logit model | |||