Statistical analysis of the evolutive effects of language development in the resolution of mathematical problems in primary school education

Abstract

For primary school students, the difficulty in solving mathematical problems is highly related to language capacity. A correct solution can only be achieved after being able to deal with different abstract concepts through several stages: comprehension, processing, symbolic representation and relation of the concepts with the right mathematical operations. A model linking the solution of the mathematical problems (PS) with the mental representation (MR) of the problem statement, while taking into account the level of the students (which has influence in the linguistic abilities), is presented in this study. Different statistical tools such as the Analysis of Covariance (ANCOVA), ROC curves and logistic regression models have been applied. The relation between both variables has been proved, showing that the influence of MR in PS is similar in the different age groups, with linking models varying just in the constant term depending on the grade level. In addition, a cutoff in the mental representation test is provided in order to predict the student’s ability in problem resolution.