Using A Machine Learning Approach to Identify Predictive Factors for Early Detection of Procrastination – An Experimental Design

Abstract

Procrastination is a prevalent issue affecting individuals in various domains, including education and personal life. Recognizing the significance of early detection to prevent its negative consequences, this research project aims to identify indicators for early signs of procrastination by conducting a literature review. Previous research identified predictor variables such as self-efficacy, conscientiousness, growth mindset, academic entitlement, and demographic factors. The most important results from the reviewed literature suggest that these predictor variables play significant roles in predicting procrastination tendencies, and integrating these factors into machine learning algorithms can enhance the accuracy of early procrastination detection. Specifically, the incorporation of self-efficacy, personality data, and motivational factors such as academic entitlement and growth mindset were variables that predicted procrastination in machine learning algorithms more accurately than demographic factors. These predictors also had higher levels of correlation with procrastination compared with demographic ones. An experimental design is proposed to study various machine learning models to analyze existing datasets and to assess the predictive performance of each predictor variable. By using ablation experiments, the design will study their individual impact on early procrastination detection and develop effective intervention strategies with the goal of fostering improved academic outcomes and overall well-being.