Will Thalheimer: Does eLearning Work? What the Scientific Research Says!

by Mirjam Neelen & Janet Benson – Learning Experience Designers

What does the scientific research tell us about elearning effectiveness, particularly as it compares with classroom instruction?

These are the questions that Will Thalheimer examines in his fresh-from-the-press report. But wait. Why would we even ask these questions? Don’t we already know that eLearning works? After all, it’s used abundantly; tons of people use it every day. Thalheimer rightly says that eLearning, on the other hand, has a reputation for being boring and ineffective and wildly hyped by vendors and eLearning evangelists. So, in summary: No, it’s not clear if it works and it’s not clear when it works, what causes it to work. Do we make sense?

By the way, it’s just awesome that the world of learning professionals can avail of someone like Thalheimer who does all this solid research and then makes it available for free so that we can all benefit from it! Perhaps you can consider hiring him or buying his book to pay him back (and to get more awesomeness!).

We, as learning experience designers at the Learnovate Centre are big advocates of backing everything that we do up with science. Actually, we feel it’s our responsibility. Mind you, Thalheimer’s research includes thousands of scientific studies as his research includes several meta-analyses to find the answer to the two questions.

Comparing eLearning to Traditional Classroom Instruction

Interestingly, based on the five meta-analyses that Thalheimer conducted, he concludes that “eLearning tends to outperform classroom instruction” (p 4). Blended learning creates the largest benefits (so, a mix of classroom and eLearning instruction).

Are you getting excited by this outcome? Don’t be. When diving deeper, it becomes clear that it’s NOT the eLearning that improves learning effectiveness, it’s the LEARNING METHODS that are typically used in eLearning that impact eLearning’s benefits.

Paul A. Kirschner tweeted, in response to Thalheimer’s article, that Clark posed a similar conclusion all the way back in 1983 (also referenced by Thalheimer, by the way), saying that

The best current evidence is that media are mere vehicles that deliver instruction but do not influence student achievement any more than the truck that delivers our groceries causes changes in our nutrition (Clark, 1983, p 445).

Which Learning Methods are Most Effective?

So, why would eLearning outperform classroom instruction and why would blended learning outperform both? One possibility is that eLearning design invites the usage of more effective learning methods. But why then would blended learning be even better? Thalheimer suggests that one reason could be that the eLearning used in blended learning experiences uses more effective learning methods. Perhaps it’s the combination of those effective methods and the comfort of a face-to-face setting? Another option is that, because the research analysed on blended learning didn’t compare the specific learning methods used, so perhaps, coincidentally, the blended learning methods investigated in the research used even more effective learning methods than the learning methods used in the eLearning studies. We’re not sure and Thalheimer doesn’t speculate any further.

So, what are these effective learning methods? Not surprisingly (to us anyway) they include elements such as “providing learnings with realistic practice, spaced repetitions, contextually-meaningful scenarios, and feedback” (p 4). Other articles discuss the richness of eLearning, such as the flipped classroom and mobile learning.

Overall, “realistic decision making and authentic tasks, providing feedback on these activities, and spreading repetitions of these activities over time” (p 10) are hugely beneficial for learning.

Now, please open your ears widely and listen cause what we’re about to say is important.

Learning effectiveness depends on the learning methods, NOT on the ‘delivery’ methods (classroom or eLearning). Classroom instruction can ALSO use these beneficial learning methods and eLearning can FAIL to use these most effective learning methods (and that, as we all know, happens too often).

Comparing Learning Methods in eLearning

From the studies that Thalheimer reviewed for this topic, it becomes clear that there is an enormous variety in eLearning approaches and contexts. For example, solving linear algebra problems or making sense of cerebral haemorrhages. This is critical because, especially in the corporate learning space, the feeling is that eLearnings are just boring, ineffective page turners. Thalheimer’s research shows that eLearning can be used for much more than information dumps or learning trivial tasks (e.g., How to Clean Your Desk on a Friday kind of stuff). At the same time, it becomes clear from Thalheimer’s research that many eLearning applications don’t utilise research-recommended learning methods so therefore the potential is … underutilised.

Other Research Relevant to eLearning

Thalheimer also explores a random sampling of research that focuses on things like simulations, simulation games, feedback, animations, digital games, and so forth. Unfortunately, he doesn’t explain how he defines eLearning. One could say that simulations and digital games can be part of eLearning as well.  The research on these ‘other’ types of learning tech show that they investigate “learning methods that tend to go beyond those we typically use in classroom instruction” (p 24). Well, we could use them in classroom instruction but we generally don’t. There’s a lot of variability in the research results but we can conclude that learning technologies seem to be, overall, beneficial in supporting learning, though, and this is important, NOT necessarily better than non-tech interventions.

This is What You Need to Remember at All Times!

  • eLearning and classroom instruction produce similar learning results when learning methods are constant.
  • eLearning tends to outperform traditional classroom instruction when no specific effort is made to hold learning methods constant (suggesting that eLearning solutions more easily invite to use more effective learning methods somehow while classroom instruction still relies heavily on relatively ineffective lectures).
  • What matters for learning effectiveness are the learning methods; NOT the modality (e.g., eLearning or classroom).
  • Blended learning tends to outperform classroom learning and eLearning (probably because the eLearning part of the ‘blend’ often uses more effective learning methods).

We are especially excited about this type of research because it helps us to help you to identify and solve your learning challenges. Don’t hesitate to contact us with any questions!

We’ll be waiting for your call.


Clark, R.E., (1983). Reconsidering research on learning from media. Review of Educational Research, 53(4), 445-459. Retrieved from https://www.researchgate.net/publication/247357883_Reconsidering_Research_on_Learning_from_Media

Thalheimer, W., (2017). Does eLearning Work? What the Scientific Research Says! Retrieved from http://willthalheimer.typepad.com/files/does-elearning-work-full-research-report-final.pdf

Subscribe to get updates on Learnovate