Part of Evidence of Educative Game Design. Effect sizes are explained in How large is the effect, really?
Choosing a format
A partner usually arrives with a format already in mind: a headset, an app, a box of cards. The research gives a clear answer about whether that assumption is safe, and the answer is not about which format is best.
Fit, not superiority
Howard and colleagues tested which factors actually moderate a learning game's effect: platform, hardware, and the presence of game elements were all examined as candidates. None of them, on their own, predicted the outcome. What predicted it was task-technology fit: whether the format matched what the learner needed to practise.
This finding rules out a specific kind of question before it gets asked. "Which format works best?" assumes formats compete on a single scale. They do not. A format that fits a spatial, high-consequence task can be the wrong choice for a task that is really about repetition, and the reverse holds equally. What follows takes each format on its own terms rather than ranking them against each other.
Extended reality
Jongbloed and colleagues found genuine procedural transfer from extended reality training to physical task performance. Man and colleagues studied construction safety training and found the format worked best for younger, less experienced workers. The effect was weaker for workers who already had more experience to draw on. That is a moderator, not a caveat to skip past: the format helped most where the learner had the least existing frame of reference to fall back on.
Kaplan and colleagues placed extended reality's real value precisely: not as a superior version of ordinary training, but as access to scenarios that are otherwise unreachable. Rare, dangerous, or expensive situations that cannot be rehearsed physically are where the format earns its cost. Where the scenario is reachable by simpler means, the case for the format is weaker, whatever its appeal.
Mobile
Sung and colleagues measured an average effect of 0.523 across mobile learning games, in the moderate range — roughly two out of three learners did better with it than without. Dunleavy and colleagues supply the caution already introduced in how large is the effect, really. Their headline figure rests on 29 trials and 3,175 learners. But the highest single subgroup result rests on only five trials, so the same discipline applies here as anywhere else a number is quoted.
Mobile's practical advantage is reach. A phone is already in a learner's pocket, and that lowers the barrier to the repeated, distributed practice described in four principles that make a game teach.
Tabletop
Roseth and colleagues examined cooperative structures with an adolescent population and found them effective for the social and negotiation skills those structures were designed to build. Gauthier and colleagues reviewed board games specifically and found a positive effect on knowledge outcomes, with a wide confidence interval that reflects variable study quality across the field. Tabletop's advantage is not technological. It is that a shared table puts a group's reasoning in the open, where a facilitator or the players themselves can see it happen.
Why more technology is not automatically more learning
Makransky and colleagues ran the direct comparison this article's frame depends on. Adding virtual reality to a lesson raised how present learners felt in the simulation. It also lowered how much they learned from it. Presence and learning are not the same outcome, and a format chosen for one does not automatically deliver the other.
This is Howard's finding restated concretely. The format has to fit the task. A more immersive format is not a proxy for a better-fitting one.
What this means for choosing a format
Start from the task, not the technology. Ask what the learner actually needs to rehearse, how often, and under what constraints, and let the format follow from that answer. A format chosen first and justified afterwards is the pattern the evidence in this article argues against.
References
- Dunleavy, M., Dede, C., & Mitchell, R. (2019). Affordances and limitations of immersive participatory augmented reality simulations for teaching and learning. Journal of Science Education and Technology, 18, 7–22.
- Gauthier, A., Corrin, M., & Jenkinson, J. (2019). Board games for health: A systematic review and meta-analysis. Games for Health Journal, 8(2), 85–100.
- Howard, M. C., Di Tosto, G., & Grabar, R. (2021). A meta-analysis of the virtual reality problem: Unresolved variables and the relationship between vection and motion sickness. Human Factors, 63(2), 202–232.
- Jongbloed, J., et al. (2024). Transfer of training from extended reality: A systematic review. Simulation & Gaming.
- Kaplan, A. D., Cruit, J., Endsley, M., Beers, S. M., Sawyer, B. D., & Hancock, P. A. (2021). The effects of virtual reality, augmented reality, and mixed reality as training enhancement methods: A meta-analysis. Human Factors, 63(4), 706–726.
- Makransky, G., Terkildsen, T. S., & Mayer, R. E. (2019). Adding immersive virtual reality to a science lab simulation causes more presence but less learning. Learning and Instruction, 60, 225–236.
- Man, S. S., Chan, J. W. H., & Wong, H. M. (2024). Extended reality training for construction safety: A meta-analysis of moderating factors. Safety Science.
- Roseth, C. J., Johnson, D. W., & Johnson, R. T. (2008). Promoting early adolescents' achievement and peer relationships. Psychological Bulletin, 134(2), 223–246.
- Sung, Y.-T., Chang, K.-E., & Liu, T.-C. (2016). The effects of integrating mobile devices with teaching and learning on students' learning performance. Computers & Education, 94, 252–275.