Developing an Artificial Intelligence Model to Determine Misconceptions About Cell Division of Secondary School Students

ABSTRACT The aim of this study is to develop an AI‐based model to identify students' misconceptions about mitosis and meiosis and to evaluate their responses to open‐ended questions. The sample consisted of 400 secondary school students in a city in eastern Turkey. Two open‐ended questions with answers rated from 0 to 3 by the teacher were prepared and entered into a 3‐column Microsoft Excel spreadsheet. The responses were then examined for OCR errors and organized on a 6‐column table. Student responses were compared with all possible responses to determine similarity rates, and OCR errors were corrected. The responses were then used to build a model using the SequenceMatcher structure. Data analysis, including descriptive statistics, violin plots, box plots, and statistical metrics, was performed via Mathworks Matlab software. The results showed that the developed model was highly effective in identifying students' misconceptions and was both valid and high‐performing in assessing student understanding.