A Regression Model To Predict The Risk Of Incomplete Grading Of Student Assessments In Higher Education Institutions

Abstract

The accuracy and completeness of student assessment data are paramount in higher education institutions, serving as a cornerstone for informed decision-making, equitable education, and student success. However, the issue of incomplete grading, where grades for assessments are missing or inaccurate, poses a significant challenge. This research presents a regression model designed to predict the risk of incomplete grading of student assessments in higher education institutions. By leveraging historical data, the model identifies factors contributing to incomplete grading, such as grading errors, data entry issues, and technological challenges. Moreover, it examines the consequences of incomplete grading, encompassing student well-being, academic performance, and institutional accountability. The model, built using a comprehensive dataset and machine learning techniques, serves as a valuable tool for educational institutions to proactively address and mitigate the issue of incomplete grading. The research targeted a population of 367 and higher education students from Kenyan universities. Online questionnaires were used to get data from the respondents and SPSS was used to convert data into numerical values. The data collected was analyzed using python data analysis tool to identify patterns and generate the model. The source-code was written in Python. The ANOVA statistic showed that the independent variables are significant to the dependent variable. Subsequently, the independent variables in the study have a significant impact on the dependent variable of Sustainable prediction of incomplete grading. The findings of the research are significant to the education sector as it adds knowledge that will help guide the institutions on how to manage missing marks.