La Predicción Del Rendimiento Académico En Alumnos De Bachillerato. Estudio De Un Caso (Centro) En España

Abstract

This paper studies the assessment during the first year of baccalaureate in seventy-six students. In a public educational institution of Saragossa (Spain), no differences in academic achievement between men (forty) and women (thirty-six) were found, but between two main modalities (Sciences, thirty-five students, versus Humanities and social Sciences, with forty-one students), finding differences outstrip students Sciences throughout the year, equaling to end with their counterparts. This leads us to study the hypothesis of whether the ratings could be “inflated” by teachers in Humanities students. For this they were conducted by a retrospective design “ex post facto” an multiple lineal regression analysis, using as a dependent variable grades in the third final evaluation, and as independent, performance in the first and second assessment and all subjects, not including any other variables. The results show different prediction equations that explaining 87% and 88% of the variance in the final performance, in the two types of modalities respectively. The values predicted by the model resulting in a significant percentage of “inflation” in the ratings of 23% in the group of Sciences and 37% in the Humanities and Social Sciences. The mean scores of the subjects discussed in three different moments are analyzed and significant difference was observed in the three evaluations, emerging some evaluation patterns in the teachers. It intuits that the “rating inflation” could begin much earlier than in the preuniversity entrance course.