Learning you can see
Why I teach abstract classes with paladins and necromancers, what research says about pictures, attention and games in learning, and which popular "facts" are simply not true.
Arcanum instead of Shape
When I teach abstract classes to students, I do not start with Shape, Circle and Square. I start with a world. It is called Arcanum and five races live in it: paladin, amazon, barbarian, necromancer and witch. Every hero is a Character, with a name, health and strength. Every hero can take a hit and every hero has an ability of their own. Polymorphism stops being a textbook word. It is the reason one loop can go through all the heroes and each can use their own ability, without the loop knowing who is who.
The goal is for the student to want to finish the story.
But does this work because of the story and the picture, or do I just like it? I decided to check. Here is what I found, and which of the “facts” we all repeat turned out to be myths.
A picture helps when it explains
Richard Mayer and his colleagues compare learning from text or speech alone with learning from text and a picture. In 13 out of 13 tests of the principle, people who learned from words and an illustration (or speech and animation) did better on tasks where they had to apply what they learned. The median effect is d = 1.35. For educational research that is huge.
But there is a catch. A picture helps when it shows the mechanism. When it is decoration, it hurts. A meta-analysis of 58 studies with more than 7,500 learners (Sundararajan and Adesope, 2020) shows that interesting but irrelevant material, such as a funny photo or an anecdote unrelated to the topic, reduces learning. So Arcanum works only as long as every race and every hit is an example of a concept. If I add a dragon that teaches nothing, it only distracts.
That is where the idea behind wannacode comes from: code is not described, you see how it works. The Java course has 3D scenes that show what happens in memory, and Arcanum is now a whole console game, lesson by lesson.
Attention is not 8 seconds
You have probably heard that our attention span is 8 seconds, shorter than a goldfish’s. The BBC traced where the number comes from. It is cited by a 2015 Microsoft Canada report, but the number is not from their research: the report takes it from a website called Statistic Brain, and the BBC could not find any source behind it. The psychologists the BBC spoke to say an “average attention span” makes no sense at all: attention depends on the task. And there is, by the way, no evidence that goldfish have particularly short memories.
The same goes for the rule that students switch off after 10 to 15 minutes of a lecture. Neil Bradbury traced the sources in 2016. The most cited of them barely discusses attention, and the primary data do not support such a limit. The most consistent finding in the literature is a different one: student attention varies most with the teacher, not with the format of the class.
I like that and it scares me at the same time. It means we cannot blame phones for everything.
Two more myths
Learning styles. The idea that there are “visual” and “auditory” learners and that each learns better when the material matches their style is everywhere. A review by Pashler and colleagues (2008) looked for the experiments that would prove it and found that there is practically no evidence. Pictures help everyone, not only the “visual” ones.
The learning pyramid. “We remember 10% of what we read, 20% of what we hear, 90% of what we do.” These percentages do not come from research. Subramony and colleagues (2014) show that the numbers appear in dozens of different versions, that they cannot be the result of a real experiment, and that they were pasted onto Edgar Dale’s cone, which has no percentages at all.
What really makes learning stick
If I had to pick one thing that works, it is trying to recall. Roediger and Karpicke (2006) had students read a text. Then some read it once more and others tried to recall it. After five minutes, those who reread remembered more. After a week it was the other way round: 56% against 42%. In a second experiment, students who read the text four times remembered 40% of it a week later. Those who read it once and tried three times to recall it remembered 61%.
A large review of study techniques (Dunlosky and colleagues, 2013) gives its highest rating to exactly two: practice testing and practice spread over time. Highlighting and rereading, the students’ favourites, are rated low.
And games?
Here the research is more careful than the marketing. The meta-analysis by Sailer and Homner (2020) shows a small to moderate positive effect of game elements on learning (g = 0.49), which holds in the rigorous experiments too. The effect on motivation, however, is no longer certain in the rigorous studies.
There are also cases where the game hurts. Hanus and Fox (2015) compare two identical courses. One has a leaderboard and mandatory badges. By the end of the semester its students are less motivated and less satisfied, and they score lower on the exam. Edward Deci, Richard Koestner and Richard Ryan (1999) explain why, with 128 studies behind them: an expected reward for something that is interesting anyway pushes the interest out. Feedback like “you are doing well”, on the other hand, strengthens it.
And one more thing that concerns me directly. Rodrigues and colleagues (2022) followed 756 students in their first programming course for 14 weeks. The effect of the game drops after the fourth week and partly comes back between the sixth and tenth, once they get used to it. The new shines first, then fades.
Why I am making a game anyway
Lumen Academy is a Learn to Win RPG. The idea is simple: points will not be bought with money but earned by learning, outside the game, with exercises and assignments. Learning makes the hero stronger.
I know about the reward trap. If the hero becomes the only reason to solve a task, the game will push out interest in the code itself. I do not know yet whether Lumen Academy will avoid it. The closed beta will show that, not my enthusiasm. That is exactly why it is closed.
But I know what I want from it, and it comes from the research, not from instinct: the game should be tied to the content, not decoration; competition should come with a team; progress should be visible; and we should measure knowledge, not points, long enough for the novelty effect to pass.
Arcanum taught me that people learn when they want to see what happens next. Now I want to prove it with numbers.
Sources
- Mayer, R. E. The multimedia principle. The Cambridge Handbook of Multimedia Learning.
- Sundararajan, N. & Adesope, O. (2020). Keep it coherent: a meta-analysis of the seductive details effect. Educational Psychology Review, 32, 707-734.
- BBC News (2017). Busting the attention span myth.
- Bradbury, N. A. (2016). Attention span during lectures: 8 seconds, 10 minutes, or more? Advances in Physiology Education, 40(4), 509-513.
- Pashler, H., McDaniel, M., Rohrer, D. & Bjork, R. (2008). Learning styles: concepts and evidence. Psychological Science in the Public Interest, 9(3), 105-119.
- Subramony, D. P., Molenda, M., Betrus, A. K. & Thalheimer, W. (2014). The mythical retention chart and the corruption of Dale’s Cone of Experience. Educational Technology, 54(6), 6-16.
- Roediger, H. L. & Karpicke, J. D. (2006). Test-enhanced learning. Psychological Science, 17(3), 249-255.
- Dunlosky, J. et al. (2013). Improving students’ learning with effective learning techniques. Psychological Science in the Public Interest, 14(1), 4-58.
- Sailer, M. & Homner, L. (2020). The gamification of learning: a meta-analysis. Educational Psychology Review, 32, 77-112.
- Hanus, M. D. & Fox, J. (2015). Assessing the effects of gamification in the classroom. Computers & Education, 80, 152-161.
- Deci, E. L., Koestner, R. & Ryan, R. M. (1999). A meta-analytic review of experiments examining the effects of extrinsic rewards on intrinsic motivation. Psychological Bulletin, 125(6), 627-668.
- Rodrigues, L. et al. (2022). Gamification suffers from the novelty effect but benefits from the familiarization effect. International Journal of Educational Technology in Higher Education, 19, 13.