INVESTIGANDO LOS CONOCIMIENTOS DE LOS ALUMNOS MEDIANTE TÉCNICAS DE CLUSTER DIFUSO

Abstract

In teaching practice, evaluations usually aim atobtaining individual results, group results are limited toquantities or percentages. Nowadays, the cluster’s analysistechniques allow teachers to create group profiles fromindividual evaluations. In this paper we present a clusteringtechnique applied to structured programming teaching .Problem-solving is one of the most important skills to beacquired by students. In order to evaluate it, the followingvariables have been considered: “syntax”, “instructions'comprehension”, “program's behavior”, “algorithmiccorrectness” and “efficiency”. By applying the Fuzzy CMeansalgorithm, the formation of groups with similarcharacteristics and each individual's degree of belongingwas investigated. The analysis of the results determined that“efficiency” is the variable that should be reinforced andthis contributed to the teaching-learning system's feedback.Index Terms ⎯ education, evaluation, programmingteaching, fuzzy cluster analysis .