Abstract
Logit modelling in training program evaluation. Unstable conditions of intervention contexts imply lack of standard instruments to evaluate programs. Frequently programs are translated into categorical variables using ad-hoc instruments. The logit modelling, with flexible statistical assumptions, allow the linear modelling of the intervention program design variables. We describe the theoretical basis of logit modelling and present an empirical study in the staff training program of the University of Sevilla. Objectives of the program (objectives, methodology and rating) predict participant satisfaction, showing its usefulness as a global quality criterion in the institution.
