Resumen de Using Artificial Intelligence to Assess the Level of Cognitive Complexity Involved in Didactic Tasks. A case study on Historical Thinking.
Connie Cofré Morales, Eduardo Puraivan, Karina Huencho Iturra, Macarena Astudillo Vásquez, Carlos Hervás-Gómez, María Dolores Díaz Noguera
The design of didactic activities is aimed at developing learning objectives at different levels of cognitive complexity. In the field of History teaching, one of the purposes is to develop Historical Thinking, which requires specific tasks. However, several investigations report that many of the activities presented in school textbooks are limited to the use of the lowest cognitive levels, not achieving the intended Historical Thinking of students. Faced with this singularity, it is valid to ask how we can use some tools to validate that the activities fulfill the expectations of cognitive complexity. In this paper, we discuss the use of an AI to evidence whether it can help as a support tool for this task. The results show that the task of assigning the predominant cognitive level is challenging even for highly qualified experts, and that the AI results match at least with expert’s assessment using some of the taxonomies considered, except for one activity. On the other hand, there is a high appreciation by experts of the potential of ChatGPT to both classify and argue its decisions, although there are also some risks to be considered.
