COVID-19 UPDATE: Important info for Ecampus students

Skip to main content


Early Detection of Placement for Success in an Online Quantitative Class

Dr. Ping-Hung Hsieh, Dr. Xiaohui Chang, and Dr. Andrew Olstad

Proposed research abstract

In a face-to-face classroom setting, an instructor can gauge students’ reception of a topic via facial expression and body language, and provide alternative examples on the spot to clarify the key ideas and concepts. It is often a challenge in an online environment for an instructor to foresee if students are confused. Students may not even know the right question to ask during virtual office house and the frustration may continue until after an assessment is done. Therefore, it is the main objective of this proposal to investigate and develop key early signals for online classes by taking advantage of vast amount of performance data available on online class management systems such as Canvas and Cengage. By examining data such as number of attempts on a question and time to complete an assignment along with student characteristics and usual performance data (e.g., GPA, class grade), the project applies one of the Cognitive diagnosis models, namely the deterministic inputs, noisy “and” gate model, to explore their associations and to develop key constructs that allow and instructor to identify students in need of assistance and proactively engage with them early in the term. This project is a pilot study for a larger research project that will examine student performance in a sequence of classes using their performance as the prerequisite class as a predictor. When successfully implemented, the project will enhance a learning environment that raises student success.

Return to 2016-17 research fellows and projects.