Teacher’s performance evaluation at a Mexican public university

Abstract

This paper reports the results of a process of design and psychometric validation of a teaching assessment instrument in a Mexican public university. A team, composed by directors, faculty and students of the educational institution designed the instrument and submitted it to a process of content validity. The final questionnaire, composed by 26 items distributed into four dimensions, was applied to a student sample of 53,192 records, which were the input for the psychometric analyses. Internal consistency analyses using the Crhonbach’a Alfa and split half reliability methods, yielded satisfactory coefficients, both at the level of the total test and of each of its dimensions. Item analysis was carried out by the frequency distribution methods, item-test and item-dimension correlations, and higher and lower group comparisons. Again, all items showed a proper performance on these tests. Finally, construct validation could be verified for three dimensions with an exploratory factor analysis and was confirmed by structural equation system (EQS). It was not possible to do it for the fourth dimension, due to its limited number of items.