Do English Learning Games Actually Improve Vocabulary?

Hero image: Story Book - English App, App Store listing.

Do English Learning Games Actually Improve Vocabulary?

Digital games built around English vocabulary do produce measurable gains, but the size of the gain depends heavily on what the game asks a learner to do, not on whether it has points, levels, or badges attached. A 2022 meta-analysis of 29 studies found a medium-sized improvement in vocabulary scores immediately after game-based practice (Dixon, Dixon & Jordan, 2022). Much of that advantage over non-game instruction disappears once learners are retested weeks later, though. What survives is narrower: game mechanics that force a learner to recall or produce a word, rather than simply recognize it, are the ones the research ties to vocabulary gains that outlast the session.

That distinction, between a mechanic that forces real retrieval practice and one that only decorates a session with rewards, is the difference this post works through. A points counter or a level-up animation measures engagement with the app, not acquisition of the word. The research below looks specifically at what happens to vocabulary scores when the game element is added, removed, or changed, which is a different question from whether learners enjoy playing or keep opening the app.

What Does the Research Actually Show About Game-Based Vocabulary Learning?

Two peer-reviewed meta-analyses, pooling dozens of individual experiments, converge on the same broad conclusion: digital games produce a real, medium-to-large vocabulary gain compared with non-game instruction, though the exact size varies by how the comparison is set up.

Daniel Dixon, Tülay Dixon, and Eric Jordan analyzed 29 experimental studies of digital game-based language learning published through January 2020 (Dixon, Dixon & Jordan, 2022). Between groups, where one group of learners played a digital game and a comparable group received non-game instruction, the weighted effect size was d = 0.50, a small-to-medium advantage for the game group. Comparing the same learners before and after playing, without a separate control group, produced a larger effect of d = 0.95. An earlier, separate meta-analysis of 10 studies by Meng-Hua Chen, Wen-Ta Tseng, and Tsung-Yuan Hsiao found a comparable medium effect of d = 0.78 for digital games against traditional instruction (Chen, Tseng & Hsiao, 2018).

Neither meta-analysis found that a game's genre or design mattered less than whether it was marketed as an educational product. Chen, Tseng, and Hsiao's moderator analysis found the effect on vocabulary varied significantly with game design (Q = 5.857, p = 0.016) but not with learner age (Q = 0.906, p = 0.341) or language background (Q = 0.0001, p = 0.994), meaning the design of the game predicted the outcome far better than who was playing it.

Does the Vocabulary Boost From Games Last, or Does It Fade?

Most of a game's advantage over non-game instruction shrinks once researchers test learners again weeks later, which is the single most important caveat in this research. Dixon, Dixon, and Jordan's meta-analysis compared delayed posttest scores between game-taught and conventionally taught groups and found a weighted effect of only d = 0.09, a gap so small its 95% confidence interval (-0.62 to 1.03) crosses zero (Dixon, Dixon & Jordan, 2022). Compared against their own immediate posttest scores, the same game-taught learners lost ground: the delayed-versus-immediate contrast was d = -0.69.

That pattern does not mean the games failed to teach anything. Compared against their own pretest scores, the delayed posttest still showed a large effect (d = 1.36), so a real share of new vocabulary survived. What disappeared was the game group's edge over the group that studied without a game, not the vocabulary itself. Applied linguist Jonathon Reinhardt has raised a related concern about apps designed specifically as language-learning games: they risk becoming "repetitive and superficial tasks in which the learning objectives are too obvious," which quietly reduces the retrieval challenge that makes the practice stick (Reinhardt, 2019, as quoted in Dixon, Dixon & Jordan, 2022, p. 2).

The practical reading is not that games stop working after one session. It is that the extra motivation a game supplies has to translate into more retrieval practice, not just more time in the app, or the vocabulary gain converges toward whatever a learner would have retained without the game at all.

Do Points, Levels, and Badges Teach Vocabulary by Themselves?

Points, levels, and badges are motivation tools, not vocabulary-teaching mechanisms, and the research on general gamification treats them that way. A peer-reviewed 2020 systematic review of 46 empirical studies on gamification in education found consistent evidence that game elements raise motivation and engagement, but flagged the link between specific game mechanics and actual learning outcomes as inconsistent across studies, varying with duration, task difficulty, and how the mechanic was implemented (Zainuddin, Chu, Shujahat & Perera, 2020). In other words, a badge reliably makes a learner want to keep going. Whether that extra time in the app produces vocabulary a learner did not have before depends entirely on what the badge was awarded for.

A badge earned for opening the app every day rewards the same behavior a plain streak counter rewards, covered separately in how to build an English learning habit that lasts. A badge earned for correctly recalling ten new words in a row rewards something closer to the retrieval practice the vocabulary meta-analyses above measured. Both use identical UI: a counter, an animation, a small reward. Only one of them is doing anything a vocabulary test would detect.

This is the practical filter a learner or a parent can apply to any English learning game inside an app: ask what specific action triggers the reward, not how satisfying the reward feels. An app that cannot answer that question clearly is optimizing for session length, which is a legitimate business goal but a different one from building vocabulary. The same filter applies to a game built around comics or short video clips rather than text, since comics and graphic novels and captioned video can supply the same context-driven word encounters a narrative game does.

Which Game Design Choices Produce Bigger Vocabulary Gains?

Game genre and structure move the size of the vocabulary effect more than almost any other variable the research has tested, including a game's educational branding. The comparison below pulls the moderator results reported across the two meta-analyses discussed above.

Games built around adventure or narrative tasks, where a word has to be understood or produced to progress through a story, produced roughly double the vocabulary effect of games without that structure (Dixon, Dixon & Jordan, 2022, citing Chen, Tseng & Hsiao, 2018). Supplementary material, such as a teacher reviewing what a learner got wrong in the game, also roughly doubled the measured effect. Games designed specifically for entertainment rather than for language instruction performed at least as well as, and in some comparisons better than, games built explicitly for L2 education, though the entertainment-game studies varied widely enough that the two categories' confidence intervals still overlap.

How Should You Judge Whether a Learning Game Is Actually Building Vocabulary?

A learner evaluating a vocabulary game inside an English learning app can apply the same distinctions the research above draws between a strong and a weak design, without needing to read the studies themselves.

None of these signals require special expertise to check. Opening a vocabulary game inside an app and asking "what did I just have to do to earn that reward" answers the first two rows directly. Storybook, from Dobi Apps, builds its reading games around this principle: its App Store listing describes rewards, levels, and achievements that unlock as a learner progresses through actual books, comics, and stories, tying the game layer to material a learner is already reading rather than to a stand-alone drill. A learner who wants the same underlying vocabulary-in-context principle in audio form can pair Storybook with Dobi Apps' Audiobook, which organizes its library by CEFR level from A1 to C2. A game that skips level matching entirely, throwing the same difficulty at every player, works against the same comprehensible input principle that makes leveled reading effective: a word surrounded by too many other unfamiliar words gives a player nothing to guess its meaning from, game or no game.

How Do Vocabulary Games Fit Alongside Spaced Repetition and Reading?

A vocabulary game and a spaced repetition system solve two different parts of the same problem, and neither one replaces the other. A game is typically where a learner first meets a word inside some task with a consequence attached: choose the wrong word and a character in the story does not progress, or a puzzle does not solve. That first, engaging encounter is exactly what the moderate-to-large effect sizes in the research above are measuring.

What a game rarely does well is bring that same word back for review at the moment it is about to be forgotten. That scheduling problem is what spaced repetition is built to solve, reviewing a word at lengthening intervals rather than scoring it once and moving on. A learner who wants both benefits gets more from combining a game that introduces new words in context with a spaced review system that brings them back on a schedule than from expecting either tool to do the other's job. The same logic applies to extensive reading through stories: a story supplies repeated, low-stakes exposure to a word across different contexts, which a points system inside a reading app can track and reward without being the actual mechanism doing the teaching.

A learner comparing tools across a broader set of goals, not vocabulary alone, can start from the top 5 apps for learning English in 2026, which covers where a gamified app such as Duolingo fits next to a story-based app, a pronunciation coach, and a spaced repetition tool. Rereading a familiar story or replaying a slower audiobook chapter are two other low-effort ways to get the same repeated exposure a game provides, without a single point awarded for it.