I think he has fitted the categorical independent variables as continues variables. For example, if he had coded gender as 1 for male and 0 for female but if he forget to indicate this variable as a categorical variable while fitting the model (whatever it is in STATA, SPSS or other statistical packages), he may get similar results.