A multinomial examination of tasks considered as problems in primary education textbooks

Olivares, Daniela; Giaconi, Valentina; Perdomo-Diaz, Josefa; Lupiáñez, José Luis

Keywords: textbook analysis, mathematics curriculum, tasks labelled as problems, bivariate analysis, multinomial regression model

Abstract

Contemporary societies demand problem-solving skills, highlighting the need for tasks that foster these abilities. Although curricular guidelines emphasise such tasks, their practical implementation in resources such as school textbooks varies. Since textbooks are essential for curriculum implementation, an important question arises: how do they frame problem-solving in primary education? We analysed Chilean textbooks used in fourth-grade primary education. Based on the literature, tasks are categorised as straightforward, grey-area or puzzle-like, depending on the solving method and cognitive demand. Additionally, factors such as the number of possible correct answers, data sufficiency, context and curricular area were analysed. By employing a multinomial model to integrate analysis of task type and characteristics, this study reveals significant correlations that deepen our understanding of the tasks textbooks consider problems. The findings show that only a small portion of tasks labelled as problems qualify as puzzle-like problems. Furthermore, bivariate analysis and the multinomial model indicated a strong correlation between certain task characteristics and their classification as problems. These insights suggest a need for strategies to increase genuine problems in textbooks and provide holistic problem-solving perspectives.

Más información

Título de la Revista: INTERNATIONAL JOURNAL OF MATHEMATICAL EDUCATION IN SCIENCE AND TECHNOLOGY
Editorial: Taylor & Francis
Fecha de publicación: 2026
Idioma: inglés
URL: https://doi.org/10.1080/0020739X.2026.2632258
Notas: Scopus, Clarivate Analytics: Emerging Sources Citation Index