Classroom Interaction Evaluation Through Sequential Analysis of Observational Data

Abstract

In this paper, an example of research in which interaction within the classroom is evaluated using a technique of sequential analysis starting from log-linear and logit models is presented. The research was carried out among students in their last year of secondary school (17-18 year olds). The general aim was to improve the teaching-learning environment through evaluating interaction. The context of study, the categorization process, the records and the observational sampling carried out and the records obtained and briefly described, are sequentially analyzed using log-linear and especially logit models, and by placing special emphasis on the concordance of the observational measurement . From the chosen model, a set of aspects of great interest is studied in depth to examine the interpretation of the results (mainly the adjusted residuals and the -ers estimated from the model). Finally, there is a discussion about a series of considerations relating to the choice of the logit models and to the usefulness and applicability of the sequential analysis in the evaluation of interaction in educational contexts.