Date of Award


Document Type


Degree Name

Master of Science (MS)


Computer Science

First Advisor

Brian Dorn

Second Advisor

Neal Grandgenett

Third Advisor

Harvey Siy


Learning analytics is the measurement, collection, analysis, and reporting of data about learners and their contexts. An outcome and primary goal of learning analytics should be to inform instructors, who are primary stakeholders, so that they can make effective decisions in their courses. To support instructor inquiry, I apply theory on reflective practice to learning analytic development. Articulating an instructor's pedagogical expectations is one way to begin facilitating a reflective practice. Expectations based on instructor goals serve as a natural next step and the springboard from which data can be collected. I hypothesize that a learning analytic that encodes and reifies instructors' individual expectations will better support reflective practice for instructors and allow students to more reliably meet set expectations.

I took a user-centered approach to learning analytic research and development. First I triangulated empirical analysis of analytic use with focus groups to understand how instructors interacted with analytics. Instructors had a wide range of behaviors, needs and expectations. For most instructors, analytics were used very briefly (less than 1 minute). Instructors also requested a way to aggregate data from different analytics to better support their information needs. Based on these findings, I developed learning analytics within TrACE to allow for instructors to specify expectations and see student progress related to those expectations. Students could also view their progress towards completing expectations.

Finally, I conducted a field study to compare both instructor analytic use and student compliance to expectations without and with the presence of these analytics. The results of the field study did not support the hypothesis. Instructors for the most part did not change their behaviors with the introduction of these analytics. Students also did not meet expectations more reliably, but one course saw a significant improvement in performance. Without visible expectations, students met significantly fewer posting expectations than other expectations. With explicit expectations, posting performance was no longer significantly less.


A Thesis Presented to the College of Information Science and Technology and the Faculty of the Graduate College University of Nebraska In Partial Fulfillment of the Requirements for the Degree Master of Science in Computer Science University of Nebraska at Omaha. Copyright 2016 Suzanne L. Dazo.

Files over 3MB may be slow to open. For best results, right-click and select "save as..."