Feedback is a critical component of supporting student learning, however, delivering effective and timely feedback is a challenge in large courses. Often in large courses, feedback is not personalized, and students receive the same message, regardless of an individual’s engagement or needs. Learning analytics-based feedback could help meet this need and create more relevant feedback. 

Learning analytics-based, personalized feedback was implemented in a large introductory course. The instructor identified groups of students to target based on their engagement and performance. Feedback messages were sent via email after each exam (a total of four times) during the semester.  

Five guiding principles were applied in crafting the feedback messages: 

1. Use data to consider the student’s performance and engagement with course resources. 
2. Target one “behavioral” aspect of the course. 
3. Acknowledge hard work and possible frustration.
4. Incorporate metacognitive/reflective prompts. 
5. Offer one or two study strategies (keep it simple). 

A total of 374 students (39% of the class) received at least one message from the instructor during the semester. Students’ perceptions of the feedback message were examined, and their engagement pre- and post-message was tracked to identify behavioral changes. Several engagement behaviors were traced, including lecture attendance, homework completion/scores, practice problem completion/scores, and course material access.  

Main Findings 

Most students perceived the feedback message as helpful and motivating. Over 80% of students indicated that they took action due to the feedback message, and the following types of actions were reported:

1. Actively sought and used more external resources (e.g., supplemental instruction). 
2. Tried to change study strategies or adopt new strategies.  
3. Reached out for help from the professor or teaching assistants.
4. Engaged more seriously with the lecture and course materials. 
5. Continued to work hard and study even more with the same strategies.  

Among the tracked behaviors, there was some indication that practice problem engagement increased after receiving the messages. However, engagement with the lecture and homework remained steady.

If you are interested in learning more about how learning analytics-based feedback can be implemented to help your students’ learning, please contact us at its-ra@iowa.uiowa.edu.