A Broad View of the Problem-Based Learning Field Based on Machine Learning: A Large-Scale Study Based on Topic Modeling

Abstract

The aim of the study is to examine Problem Based Learning (PBL) studies in terms of descriptive and semantic content analysis by using topic modeling. For this purpose, descriptive and topic modeling analyzes were used together in the research. In order to include the highest number of articles on Scopus, the term "problem based learning" was searched in the title, abstract and keywords and only journal articles (research and review) were selected. Thus, 7289 articles in 1987-2021 were included in the study. Firstly, the subject area, author and country distributions are listed. In addition, it showed that the most studied topics were education curriculum (39.15%), teaching strategies (14.90%), critical thinking skill (12.29%) and patient simulation (8.88%). When examined in seven five-year periods between 1987 and 2021, it was determined that the most voluminous topic was education curriculum, and the most accelerated topic was clinical education. Considering the number of publications in five-year periods, it was determined that the topics of critical thinking skills and teaching strategies accelerated more in the percentages calculated according to the topics. It is expected that the results obtained will be important reference points for the studies to be carried out in the field of PBL