It’s a post on this blog that was heavily shared and read, and it was also a chapter in one of our Urban Myths books: Universal Design for Learning. At the time of writing, little research was present. We didn’t describe it as a myth, but we did warn of a possible ‘mentos effect’.

For instance if you look at the checkpoint “Offer ways of customizing the display of information”, you get a large list of research on this topic, but this research wasn’t necessarily conducted from a UDL point of view. Again this is isn’t necessarily a bad thing, but this doesn’t guarantee that the combination in UDL might have the desired effect.  Compare it with Diet Coke and Mentos. If you want to refresh your breath, than Mentos will do. If you want to quench your thirst, for some Diet Coke will do. If you put a Mentos in a bottle of Coke Light, well, you get this (Btw, this is not a so uncommon problem in educational sciences, but most of the time people don’t pretend to be universal).

But now I received, through my colleague in Utrecht Jeroen Janssen, a new meta-analysis with interesting and positive information about UDL. In this study King-Sears et al examined studies to the following criteria:

Five inclusion criteria were applied: (a) original research in English; (b) experimental design with treatment and control groups (including true experimental and quasi-experimental studies); (c) learners’ achievement measured; (d) data available for effect size calculations; and (e) UDL was intentionally and proactively applied to interventions’ designs. There were no age or grade restrictions, nor were there restrictions for content of what was taught. Studies were excluded if UDL was identified but had scant, if any, expansion describing how UDL was operationalized. For example, studies that cursorily mentioned UDL, but did not specify how the components of an intervention were designed to align with UDL, were excluded.

This left 20 studies to be included.

What about what elements were checked from UDL?

Fifteen studies indicated their corresponding interventions featured all three UDL principles (engagement, representation, expression), one study featured one principle (representation), and four studies did not report information about the UDL principles. From the nine UDL guidelines (see udlguidelines.cast.org), five studies aligned with five or more guidelines, seven studies featured between one to three guidelines, and eight studies did not report alignment with guidelines.

And now, most importantly, the results?

The analysis revealed a medium positive combined effect for UDL (g = 0.43; SE = 0.11; 95% C.I. [0.19, 0.67]; p = 0.00), indicating that the academic achievement of learners from UDL-based instructional settings was moderately better than that of learners from non-UDL-based instructional settings.

Abstract of the study:

This meta-analysis examined learners’ academic achievement in Universal Design for Learning (UDL) environments compared to business-as-usual conditions. Twenty studies, consisting of 50 individual effects, met the eligibility criteria, focusing uniquely on participants’ learning and treatment/control designs. Academic achievement was analyzed for pre-kindergarten to adult participants. Results yielded a moderate positive combined effect for learners receiving UDL treatments (g = 0.43), indicating moderate efficacy of the UDL-based instruction. Five significant moderators were identified. In addition, the UDL Reporting Criteria were employed to assess whether studies included information regarding UDL-based design components. UDL’s emergence as a research-based practice for diverse learners is discussed.