Design of a teaching performance evaluation model based on causal methodology

Abstract

Objetive: The investigation consisted in investigating and validating an evaluation model of teacher's performance through the modeling of structural SEM equations applying LISREL. Method: An initial model based on the results of investigations and data from bibliographic referents was proposed. In previous studies they only consider the influence of some variables over the result variable. The proposed initial model consists of fourteen constructs. Six exogenous and six endogenous with their respective observable or measureable variables and in accordance with the causal methodology a diagram of cause is plotted — effect among the constructs, each relationship between the variables is codified for its statistical treatment. Information about the observable variables was collected through previously validated questionnaires from three sources: students, teachers and the immediate authority. The correlation between the observable variables and the constructs, the correlation among the constructs has been codified to build a system of simultaneous equations, for which it is necessary to calculate their factors by solving them. The validation consisted in comparing the matrix of expected covariances and the matrix of reproduced covariances. A measurement of the quality of the adjustments is made to obtain the best results in this case, the model. Result: In the investigation it was demonstrated that in the proposed model only two constructs, condition of the teacher and relationship authority teacher, were not related to the others.   

Conclusion: They must not be considered to evaluate the performance of the teacher.