Patterns of Misconception Change in Stellar Evolution Under Differentiated Instruction

This study maps changes in misconceptions about stellar evolution by emphasising three student-level categories: positive, stable, and negative change. Misconceptions in astronomy often persist in definitional concepts and causal relations. A quasi-experimental pretest–posttest design with experimental and control classes was employed. Data were collected using a fifteen-item, four-tier diagnostic; an item was considered understood when both the answer and the written reason were correct. We analysed each student’s total understanding score to classify individual change and then computed class-level proportions. Results show the experimental class was dominated by positive change 62.9%, followed by stable 25.7% and negative 11.4%; the control class showed positive 48.6%, stable 24.3%, and negative 27.0%. These findings indicate that differentiated instruction is more effective in reducing misconceptions at the individual level.