Developing an efficient computational method that estimates the ability of students in a Web-based learning environment

This paper presents a computational method that can efficiently estimate the ability of students from the log files of a Web-based learning environment capturing their problem solving processes. The computational method developed in this study approximates the posterior distribution of the student’s ability obtained from the conventional Bayes Modal Estimation (BME) approach to a simple Gaussian function in order to reduce the amount of computations required in the subsequent ability update processes. To verify the correctness and usefulness of this method, the abilities of 407 college students who solved 61 physics problems in a Web-based learning environment were estimated from the log files of the learning environment. The reduced chi-squared statistic and Pearson’s chi-square test for the goodness of fit indicate that the estimated abilities were able to successfully explain the observed problem solving performance of students within error. The educational implications of estimating the ability of students in Web-based learning environments were also discussed.

► A computational method that estimates students’ ability in an e-learning environment. ► Fitting the posterior distribution of Bayes Modal Estimation (BME) to a Gaussian function. ► Reduces the amount of computations required in the conventional BME approach. ► Estimated abilities explained students’ problem solving performance within error.