Learning in virtual realityEffects on performance, emotion and engagement

A textbook, a headset, and a video. Same content. Same seven minutes. Same three-dimensional model of a plant cell. Three ways to learn exactly one thing. One of them changed how people felt while they were doing it. That detail sounds like a footnote, but it might actually be the finding. By 2018, virtual reality had accumulated a lot of enthusiasm in educational circles and not nearly enough rigorous evidence. The pitch was compelling: virtual reality headsets deliver full three-dimensional immersion, replacing the physical world with a rendered environment that can respond to your movement, let you interact with objects, and engage multiple senses at once. For subjects like chemistry or cell biology, where spatial structure is the whole point, that sounded genuinely useful. There were early signals — Webster compared immersive virtual reality to traditional lecture methods in military training and found an average improvement of twenty-six percent for the virtual reality group versus eleven percent for the traditional group. But those studies often used small samples, loose designs, and short timelines. Nobody had cleanly isolated delivery format from content. That's what Allcoat and von Mühlenen set out to fix. To understand what they were testing, it helps to know Bloom's taxonomy — a hierarchy of cognitive skills that runs from the most basic, remembering factual information, up through understanding, applying, analyzing, evaluating, and finally creating. Different learning environments might support different levels of this hierarchy. Virtual reality's immersive, interactive qualities seemed like they might help with recall and comprehension, but the question was whether that advantage would hold up against a well-designed textbook when everything else was held equal. Ninety-nine first-year psychology students at the University of Warwick were randomly assigned to one of three conditions. The traditional group received textbook-style screenshots of the three-dimensional plant cell model alongside descriptive text in a PDF. The video group watched a two-dimensional recording matched to virtual reality footage, with the ability to play, pause, and rewind. The virtual reality group used an HTC Vive headset running the Lifeliqe Museum application, which let them rotate, resize, and highlight parts of a fully manipulable three-dimensional plant cell — the audio narrator was disabled so the text content stayed identical across all three groups. Every participant saw the same information. The only variable was the delivery format. The procedure ran in three phases: a pretest, a seven-minute learning period, and a post-test. The knowledge test used seventeen biology questions drawn from or modeled on a British AQA Biology A-Level exam, split into twelve questions targeting remembering and five targeting understanding. Mood was tracked using an adapted Differential Emotions Scale, covering nine emotion categories rated before and after the learning phase. Engagement was measured with the Web-Based Learning Tools Evaluation Scale, a thirteen-item instrument covering learning, design, and engagement subscales. Now for what they actually found. Across all participants, knowledge improved by twenty-three point two percentage points from pretest to post-test — a strong and significant overall effect. But the conditions diverged meaningfully. The virtual reality group moved from twenty-eight point one percent correct to fifty-six point five percent, a gain of twenty-eight point five points. The textbook group moved from twenty-five point three to fifty point two, a gain of twenty-four point nine points. The video group moved from twenty-seven point nine to forty-three point nine — a gain of only sixteen point one points. Post-hoc tests confirmed that virtual reality's post-test score was significantly higher than video's. The textbook also outperformed video. But here's what the headline obscures: virtual reality did not clearly beat the textbook on overall performance. The advantage was real but specific. When Allcoat and von Mühlenen broke the questions down by Bloom's taxonomy, the picture sharpened. On the twelve remembering items — straightforward factual recall — virtual reality scored fifty-three point one percent on the post-test, textbook scored forty-three point six percent, and video scored forty point six percent. Virtual reality was significantly better than both. On the five understanding items, virtual reality and textbook both outperformed video, but the difference between virtual reality and textbook was not statistically significant. In other words, virtual reality's specific edge was in memory, not comprehension. For higher-order understanding, a well-structured textbook kept pace. Confidence ratings followed the same shape. Virtual reality and textbook both produced larger confidence gains than video, with virtual reality jumping one point twelve on a five-point scale and textbook jumping one point eighteen — nearly identical. Video gained only zero point seventy-one. The passive format left people both less knowledgeable and less confident in what they knew. That's the cognitive story. Then there's the emotional one, and it's where the data get genuinely interesting. After the seven-minute learning phase, virtual reality participants showed a significant increase in positive emotions — rising from three point two to three point eight on the Differential Emotions Scale — and a significant decrease in negative emotions, falling from one point seven to one point three. In the textbook and video conditions, the opposite happened: positive emotions fell significantly in both groups. Negative emotions stayed flat in those two conditions — they didn't spike, but they didn't drop either. Only virtual reality moved the emotional needle in a positive direction. Engagement scores on the Web-Based Learning Tools Evaluation Scale reinforced this. Both the learning and engagement subscales were rated significantly higher in virtual reality than in the textbook condition, and the textbook rated higher than video on both. The qualitative responses — provided by about half the participants — put texture on the numbers. Video got words like "confusing" and "difficult to navigate." Textbook got "basic," "boring," and "bland." Virtual reality drew "engaging," "immersive," and "made learning more exciting," though some users noted it was "difficult to use at first." So the picture is this: virtual reality didn't dramatically outperform textbooks on raw knowledge scores. But it produced better factual recall, higher engagement, and — most distinctively — a positive emotional shift during the learning experience itself. That combination raises a real question. Is virtual reality helping people learn because it's cognitively superior? Or because it makes learning feel less like a chore? And in practical terms, does that distinction matter? Allcoat and von Mühlenen push on the mechanism. The key observation is that the video and virtual reality conditions shared essentially the same visual content — the same three-dimensional model rendered on screen. The video condition performed worst. That rules out visuals as the driver of virtual reality's advantage. What the video lacked, and virtual reality had, was interactivity and three-dimensional immersion. The paper frames this as an active versus passive learning distinction: virtual reality required participants to navigate and manipulate the model; video just asked them to watch. The textbook, despite being a flat medium, also required active engagement — reading, navigating text — which is likely why it held its own on understanding tasks. The authors are candid about what this study can't answer. The most significant open question is novelty. Participants encountering a virtual reality headset for the first time may have been energized simply by the newness of the experience. The positive mood shift, the higher engagement ratings — some portion of that could be the novelty effect rather than anything intrinsic to virtual reality as a learning format. Whether the gains persist, diminish, or even grow as users become familiar with the technology is something a single seven-minute session cannot tell you. Allcoat and von Mühlenen call explicitly for longitudinal studies to track whether these effects hold over time. What this study does establish is a useful baseline. When content is held constant and delivery format is the only variable, virtual reality produces better memory retention than both textbook and passive video, and a better emotional experience than either. That specific profile — stronger recall, better mood, higher engagement — points toward the contexts where virtual reality might matter most: not as a replacement for textbooks across the board, but as a format for situations where motivation and retention are the real barriers. Medical training, corporate onboarding, and abstract scientific concepts where spatial reasoning is essential — anywhere that engagement is what separates learning from forgetting. The textbook is not going anywhere. In a tightly controlled experiment with identical content, it matched virtual reality on comprehension and nearly matched it on recall. But the headset earned something the textbook couldn't: participants who walked out of the learning session in a better mood than they walked in. In education, that might matter more than any single test score. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.