Blackboard’s Data Science Group Reports Findings On Student Performance Patterns


Public results from analytics research initiatives are starting to inform the conversation in our learning community. A recent blog post by Blackboard shows some of the insights they have managed to draw from a Spring-long dataset of customer’s courses across North America this year.

The sample set had responses for 601,544 learners from 18,810 Blackboard Learn and Moodleroom courses. They required at least 10 enrolled students with at least 60 minutes of individual activity each. Data was gathered from 927 institutions but the anonymization of responses did not guarantee any number of respondents from institutions.

Pie chart of percent use of tools in the LMS
Distribution of time spent on course by enrolled students from the sample. Blackboard.

First analysis performed by Blackboard’s data science group focused on time usage. Content takes, expectedly, the bulk of the time, followed by Assessment and Grades. Less is given to announcements and discussions. Students devote minimal time to social activities such as messaging, blogs, and groups. Group member and author of the post John Whitmer clarifies that this distribution does not take into account time spent outside the LMS, working on assignments for submission a prominent example.

Next they tried to dig deeper into the relationshop between time spent and resulting grade. A previous, in-progress post had asserted a positive, if perhaps circumbscribed, correlation. (It also included some technical comments on the research whole). Now, the further exploration segmented the samples in grade and LMS use groups.

Takeaway: the positive correlation with low significance holds. Students spending more time in content do tend to better grades but to a point. After a certain amount, more time spent in course content does not affect final grade significantly; in some cases, it actually hindered performance. This makes me think of a classmates I had, who failed to prepare, and at the time of examination bidded the hour in hopes for last-minute redemption.

In an, arguably, more revealing exercise, they sorted the time spent by students in specific activities, grouped by grades. It revealed a correlation between students looking at their grades (MyGrades section) and getting a high grade. Here’s Whitman:

« The most successful students are those who access MyGrades most frequently; students doing poorly do not access their grades».

In Whitmer and Blackboard’s data science group view, the results are, as a general conclusion, in line with expectations. Hopefully, the patterns that start to unfold are interesting enough to whet the appetite of more moodlers about analytics and data science for learning.

Read the full post for a complete breakdown of Blackboard’s data science research.


Do you gather new or different insights from the research? Let’s have your take in the comment below.


moodlerooms-logoThis Moodle Practice related post is made possible by: MoodleRooms the open source learning experience by Blackboard. Rediscover Moodle. Click here to learn more.