Data-Driven Instructional Practices: Using Behavior Detection and Model Assignment Techniques to Inform Teaching Strategies for School Education

An increasingly common method for assessing students’ performance on tests is automated grading. It might potentially free up time for teachers and provide students with fast feedback so they can see where they need to improve. For better learning outcomes, automated grading must be utilized in conjunction with other instructional strategies. In this study, we provide Python codes for factor analysis-based automated grading to analyze students’ performance on tests. The code first loads the evaluation data as a matrix with each row denoting a student and each column denoting an evaluation item. After that, we use factor analysis to find latent variables that affect student performance. We establish a grading system in which each letter grade is assigned a specific threshold score. Afterwards, we assign each student a grade based on their factor scores, which are determined by adding their results for each latent factor. Each student’s grade and overall score are output by the code. Additionally, our automated grading system demonstrated the ability to spot trends in student performance that could guide teaching strategies. Our computerised grading system can potentially save time and give students rapid feedback, but it must be used in concert with other instructional strategies to guarantee successful learning outcomes. Our study emphasises the significance of employing automated grading to supplement human evaluators rather than to replace them. In order to further tailor learning experiences for students, future study might examine the use of automated grading in conjunction with other behaviour detection and model assignment strategies. Additionally, the development of increasingly complex algorithms might make it possible for automated grading systems to spot trends in student performance that are difficult for human assessors to spot.