In 2006, Asher Koriat and Robert Bjork asked participants to estimate how much they had learned immediately after studying new material. The predictions were systematically wrong — not randomly, but consistently overestimated. The mechanism is the fluency effect: when processing information feels easy (the material is on screen, the interface is intuitive, feedback is immediate and positive), the brain interprets that ease as a signal of deep learning. The feeling of having learned and actual learning diverge.
This is especially relevant for digital educational games, because many are designed to maximise exactly what produces the fluency feeling: friendly interfaces, immediate positive feedback, smooth difficulty progression, rewarding animations. A game can deliver an excellent user experience and superficial learning simultaneously. The question parents rarely ask — and research suggests is the most important — is not "is my child enjoying it?" but "what exactly is the game demanding of them cognitively?"
Recognition versus recall: the distinction that matters
Memory research distinguishes two fundamentally different processes: recognition (identifying the correct answer among presented options) and recall (generating the answer without external cues). Recall is consistently harder — and that difficulty has value. Practising recall produces stronger, more transferable memory traces than practising recognition.
In Hangman, the player must recall letters from memory — there is no option list, no letter bank on screen. Each attempt is pure recall. In Crossword, each word must be generated from a definition with no additional cues.
Educational Quiz uses a four-option format — recognition — but with a mechanism that makes it more effective than it appears: the incorrect options are plausible, forcing active elimination reasoning. The player who rules out "Jupiter" as the smallest planet is exercising knowledge, not just spotting the correct answer by mechanical exclusion.
Cognitive load: the parameter that separates challenge from entertainment
John Sweller's cognitive load theory (1988) identifies three types: intrinsic load (material complexity), extraneous load (caused by poor design) and germane load (effort of building new cognitive schemas). Genuine learning requires germane load — the effort of constructing new knowledge structures.
A game with an overloaded interface, constant sound effects and excessive animations may have such high extraneous load that it competes with the educational content for the child's attention. Math Adventure, Chess and Checkers have low-extraneous-load interfaces: attention goes to the problem, not to navigating the interface. Super Zoo uses visual stimulation as part of the memory challenge — not as decoration competing with it.
Four criteria for evaluating any educational game
- Does the game require the child to generate the answer, or only choose from options? In Hangman and Crossword, there is no recognition shortcut — the answer must come from memory.
- Is there a real cost for error? In Chess, every poor move permanently changes the strategic position within that game. Games that do not penalise errors incentivise random guessing, which consolidates nothing.
- Does the skill the game exercises transfer to offline contexts? If a child excels in the game but cannot apply the same skill in a different context, the game is training interface mechanics, not cognitive skills.
- Is the difficulty progression real or cosmetic? Requiring more points to advance is not cognitive progression. Increasing the complexity of reasoning required — more chained operations, less common words, mazes with more false paths, a chess opponent that analyses more moves ahead — is.
The role of analogue games as a counterpoint
Physical and analogue games exercise dimensions no screen fully replicates: physical manipulation of objects, negotiation with a person present, reading facial expressions, and tolerance for rules that cannot be reconfigured. A digital Checkers game is a legitimate exercise in strategic thinking — but a physical board game adds social and motor dimensions that the digital version does not cover.
Technology dependency in education is not a screen quantity problem: it is an exclusivity problem. The tool that cannot be set aside is no longer a tool — it has become the entire environment. And no single environment, however good, develops the breadth of skills that alternating between modalities produces.