Abstract
Complex problem solving (CPS), one of the most prominent 21st century skills, is an important educational goal. Previous studies have demonstrated that varying levels of proficiency in students’ problem solving processes exist through computer-based assessments. This present study aimed to identify students’ problem solving processes by focusing on a single item based on the Norwegian PISA 2012 logfile data (N = 353). To specifically identify distinct profiles of students’ problem solving processes, this study derived fine-grained indicators that jointly considered several exploration strategies and time related to problem solving. Using latent profile analysis, this study identified four qualitatively distinct profiles of students’ problem solving processes: inactive, struggling, proficient, and adaptive. Further analysis demonstrated that students’ demographic characteristics (i.e., gender, SES) and motivational characteristics (i.e., openness) differentiated profile membership. In addition, students’ profile membership differentiated their performance not only on a specific item but also on overall problem solving to some extent. Providing information about different profiles can support teachers to adapt instructions to specific students’ needs and develop automated feedback teaching tools that can provide instant feedback. Limitations and future research are discussed