Pattern Discovery in Tertiary Level Subject Grades as Basis for Teacher Bias

ABSTRACT

This quantitative-experimental research aimed to data mine and search for patterns of implicit bias among teachers in terms of assessing, evaluating and grading their students with respect 5 implicit bias factors: students’ gender, age group, family financial background, highs school grade point average (GPA), and cultural background. The dataset came one of the campuses of the largest state universities in the Ilocandia Region of the Republic of the Philippines composed of two (2) years of Information and Communications Technology (ICT) teacher class records among its 450 bachelor of Science in Information Technology (BSIT) students. Student grades were graphed using the Normal Probability Distribution or Bell Curve to easily search for patterns for every class record submitted by the subject teachers. An overall average bell curve graph using cluster analysis via Gaussian Mixture Model was also prepared to determine the commonality among teacher grading patterns in terms of the 5 implicit bias factors. Results revealed that the grade distribution follows the bell curve and that the 18 subject teachers have no implicit bias in assessing and grading their students in terms of gender, age group, family financial background, highs school grade point average (GPA), and cultural background.

FULL TEXT