The most striking difference between the two models was that the total household expenditure was not significant in the homoscedastic probit model, whereas it was highly significant in both mean and latent scale models (heteroskedastic probit model), indicating that the inference based on the ordinary probit model is inaccurate in case of violation of the homoscedasticity assumption.
← all excerpts
Predictors of catastrophic out-of-pocket health expenditure in rural Egypt: application of the heteroskedastic probit model.
1
—
—