Grade inflation in introductory physics: the influence of out-of-class assignments

Abstract

We report the results of statistical analysis performed on course grades for calculus-based introductory physics for data collected over a four-year period. We consider two important categories of scores: proctored (in-class proctored exams only) and proctored plus out-of-class (in-class proctored exams plus out-of-class assignments). The analysis revealed significant grade inflation in the proctored plus out-of-class scores. Quantile plots were used to compare the observed data and data modeled using the normal distribution. These plots revealed negligible correlation between the observed and modeled data for the proctored plus out-of-class scores, while a strong correlation is observed for the proctored scores. Using the proctored grade distribution as a reference, we performed goodness-of-fit tests using the Bayesian probability fit and the original reference proportions. Using the expected counts from the two different methods, we found p-values of 0.023 and 0.008. Both p-values support the hypothesis that there is significant difference in proctored plus out-of-class grade distribution compared to the reference. Further analysis showed that approximately 25% of all grades are shifted towards higher grades. Our studies clearly show that grade inflation induced by out-of-class assignments is a crucial issue in assessment that has to be addressed. By comparing the degree of inflation in our grade distribution with the national average, we found it to be about 50% less severe.