Forum - an online learning platform - classroom data for tasks like tracking attendance and student talk time. It is also used to show talk time per student via green red and yellow indicators, accessible to the instructor at any time via a hot key.
However, we wondered how else we might leverage this data for educational purposes.
The Problem
How might we better use Forum’s classroom data to inform instructors and administrators of student engagement, throughout a program?
I began by discussing existing reports with multiple partner program managers and examining the classroom data collected by Forum. I explored ways to summarize key engagement metrics such as reactions, talk time, and chat activity to provide insights on student participation.
Two Directions
I created two potential dashboard designs, each displaying engagement data in different formats and styles. A survey of 14 instructors helped guide design decisions:
Feedback Summary
Iteration
In response to the survey feedback, I explored program, section, and class level overviews, as well as including technical or internet connection issues.
I collected feedback from 16 instructors and administrators on the test prototype.
Feedback
Overall, instructor feedback was overwhelmingly positive. Helpfulness ratings for each section improved in almost all areas, and the section overview was rated as the most helpful page.
Quotes from the survey.
The Final Designs
The final dashboard provides instructors with:
3 key metrics: student talk time, instructor talk time, and attendance
Additional metrics such as polls, chat, breakouts, and reactions
Student talk time history for the previous 4 classes
The final design used color to indicate engagement 'health' in an easily glanceable way.
Results & Looking Ahead
The new dashboard was able to replace the unscalable practice of creating custom reports for every partner three times a term, saving over 400 hours annualy.
In addition, instructors are already starting to utilize this data to better inform their decisions during class.
Future improvements could include using this data to better understand student engagement trends and predict changes, though more data would be needed for a high level of confidence.