Recently I wrote a post checking up on a claim by D2L that seems to imply that their learning platform leads to measurable improvements in academic performance. The genesis of this thread is a panel discussion at the IMS Global conference where I argued that LMS usage in aggregate has not improved academic performance but is important, or even necessary, infrastructure with a critical role. Unfortunately, I found that D2L's claim from Lone Star was misleading:
That’s right – D2L is taking a program where there is no evidence that LMS usage was a primary intervention and using the results to market and strongly suggest that using their LMS can “help schools go beyond simply managing learning to actually improving it”. There is no evidence presented of D2L’s LMS being “foundational” – it happened to be the LMS during the pilot that centered on ECPS usage.
Subsequently I found a press release at D2L with a claim that appeared to be more rigorous and credible (written in an awful protected web page that prevents select - copy - paste).
D2L Launches the Next Generation of BrightSpace and Strives to Accelerate the Nation’s Path to 60% Attainment
D2L, the EdTech company that created Brightspace, today announces the next generation of its learning platform, designed to develop smarter learners and increase graduation rates. By featuring a new faculty user interface (UI) and bringing adaptive learning to the masses, Brightspace is more flexible, smarter, and easier to use. [snip]
D2L is changing the EdTech landscape by enabling students to learn more with Brightspace LeaP adaptive learning technology that brings personalized learning to the masses, and will help both increase graduation rates and produce smarter learners. The National Scientific Research Council of Canada (NSERC) produced a recent unpublished study that states: “After collating and processing the results, the results were very favourable for LeaP; the study demonstrates, with statistical significance, a 24% absolute gain and a 34% relative gain in final test scores over a traditional LMS while shortening the time on task by 30% all while maintaining a high subjective score on perceived usefulness.”
I asked the company to provide more information on this "unpublished study", and I got no response.
Hello, Internet search and phone calls - time to do some investigation to see if there is real data to back up claims.
Details on the Study
The Natural Sciences and Engineering Research Council of Canada (NSERC) is somewhat similar to the National Science Foundation in the US - they are funding agency. When I called them they made it perfectly clear that they don't produce any studies as claimed, they only fund them. I would have to find the appropriate study and contact the lead researcher. Luckily they shared the link to their awards database, and I did some searching on relevant terms. I eventually found some candidate studies and contacted the lead researchers. It turns out that the study in question was led by none other than Dragan Gasevic, founding program co-chair of the International Conference on Learning Analytics & Knowledge (LAK) in 2011 and 2012, and he is now at the University of Edinburgh.
The grant was one of NSERC's Engage grants which look for researchers to team with companies, and Knowillage was the partner - they have an adaptive learning platform. D2L acquired Knowillage in the middle of the study, and they currently offer the technology as LeaP. LeaP is integrated into the main D2L learning platform (LMS).
The reason the study was not published was simply that Dragan was too busy, including his move to Edinburgh, to complete and publish, but he was happy to share information by Skype.
The study was done on an Introduction to Chemistry course at an unnamed Canadian university. Following ~130 students, the study looked at test scores and time to complete, with two objectives reported - from the class midterm and class final. This was a controlled experiment looking at three groupings:
- A control group with no LMS, using just search tools and loosely organized content;
- A group using Moodle as an LMS with no adaptive learning; and
- A group using Moodle as an LMS with Knowillage / LeaP integrated following LTI standards.
Of note, this study did not even use D2L's core learning platform, now branded as BrightSpace. It used Moodle as the LMS, but the study was not about the LMS - it was about the pedagogical usage of the adaptive engine used on top of Moodle. It is important to call out that to date, LeaP has been an add-on application that works with multiple LMSs. I have noticed that D2L now redirects their web pages that called out such integrations (e.g. this one showing integration with Canvas and this one with Blackboard) to new marketing just talking about BrightSpace. I do not know if this means D2L no longer allows LeaP integration with other LMSs or not. Update 6/25: Confirmed that LeaP is still being actively marketed to customers of other LMS vendors.
The study found evidence that Knowillage / LeaP allows students to have better test scores than students using just Moodle or no learning platform. This finding was significant even when controlling for students' prior knowledge and for students' dispositions (using a questionnaire commonly used in Psychology for motivational strategies and skills). The majority of the variability (a moderate effect size) was still explained by the test condition - use of adaptive learning software.
Dragan regrets the research team's terminology of "absolute gain" and "relative gain", but the research did clearly show increased test score gains by use of the adaptive software.
The results were quite different between the mid-term (no significant difference between Moodle+LeaP group and Moodle only group or control group) and the final (significant improvements for Moodle+LeaP well over other groups). Furthermore, the Moodle only group and control group with no LMS reversed gains between midterms and finals. To Dragan, these are study limitations and should be investigated in future research. He still would like to publish these results soon.
Overall, this is an interesting study, and I hope we get a published version soon - it could tell us a bit about adaptive learning, at least in the context of Intro to Chemistry usage.
Back to D2L Claim
Like the Lone Star example, I find a real problem with misleading marketing. D2L could have been more precise and said something like the following:
We acquired a tool, LeaP, that when integrated with another LMS was shown to improve academic performance in a controlled experiment funded by NSERC. We are now offering this tool with deep integration into our learning platform, BrightSpace, as we hope to see similar gains with our clients in the future.
Instead, D2L chose to use imprecise marketing language that implies, or allows the reader to conclude that their next-generation LMS has been proven to work better than a traditional LMS. They never come out and say "it was our LMS", but they also don't say enough for the reader to understand the context.
What is clear is that D2L's LMS (the core of the BrightSpace learning platform) had nothing to do with the study, the actual gains were recorded by LeaP integrated with Moodle, and that the study was encouraging for adaptive learning and LeaP but limited in scope. We also have no evidence that the BrightSpace integration gives any different results than Moodle or Canvas or Blackboard Learn integrations with LeaP. For all we know given the scope of the study, it is entirely possible that there was something unique about the Moodle / LeaP integration that enabled the positive results. We don't know that, but we can't rule it out, either.
Kudos to D2L for acquiring Knowillage and for working to make it more available to customers, but once again the company needs to be more accurate in their marketing claims.