This follows a conditional logit specification (i.e. single choice equal to maximal latent utility, with coefficients constant across choices) that has already been specified.
Predictors include distance and some predictive factors (like observed student variables) with coefficients in common across options, there should be treatment specific fixed effects (centerd around zero) in drawing attendance.
This follows a conditional logit specification (i.e. single choice equal to maximal latent utility, with coefficients constant across choices) that has already been specified.
Predictors include distance and some predictive factors (like observed student variables) with coefficients in common across options, there should be treatment specific fixed effects (centerd around zero) in drawing attendance.