In short
- Meta-analyses of digital vocabulary games report medium to large gains, mostly against different activities rather than the same content taught without a game.
- The advantage of games shrinks on delayed tests, and few studies show that games keep words better than an equally active task.
- The strongest evidence is for the mechanism inside a good game, retrieving and using a word, not for any particular format.
- Judge a game by what each pupil has to do with the word, and by whether the word comes back on another day.
Word games are easy to defend in a staff meeting. Pupils enjoy them, the room is busy, and the words are on the board. Whether those words are still there a month later is a different question. This article looks at what studies and meta-analyses have found about games and vocabulary, who took part, and what the games were compared with. It ends with a checklist for telling a game that builds retention from one that fills time.
Do word games improve vocabulary retention?
They can, but the evidence is weaker than the enthusiasm. Three meta-analyses of digital games report positive effects on second-language vocabulary, which their authors describe as medium to large. Most of the studies compared a game with a different activity, not the same content taught another way, and tested learners straight after. Where retention was tested later, the advantage of games was smaller and less certain.
The better-supported claim is about mechanism. Games help when they make learners retrieve a word from memory and do something with it. Research on retrieval and on task involvement backs that claim well. Research on specific classroom formats, such as bingo, word searches or guessing games, is thin.
What the studies found
The table lists the main reviews, then three experiments that explain why some games work. Effect sizes are given as the authors reported them. Dixon and colleagues (2022) describe an effect of about 0.5 as small to medium for this field, and about 1.0 as large.
| Study | What was pooled or tested | Learners | Main finding | Caution |
|---|---|---|---|---|
| Chen, Tseng and Hsiao (2018) | 10 studies of digital vocabulary games, 2002 to 2014 | Mixed ages and first languages | Overall effect about 0.78; adventure games gained more than other games | Small pool of studies |
| Tsai and Tsai (2018) | 26 studies of digital vocabulary games, 2001 to 2017 | Several educational levels | Large effect (d about 0.99) for games against other activities, in 10 studies | Only two studies gave the comparison group the same content |
| Dixon, Dixon and Jordan (2022) | Studies of publicly available digital games, all language outcomes, up to January 2020 | Learners of English, German, Italian, Japanese and Spanish | Small to medium effect against comparison groups (d = 0.50); 0.59 for vocabulary | No clear advantage over comparison groups on delayed tests |
| Sailer and Homner (2020) | Gamification: game elements added to ordinary teaching | Mostly university students | Learning outcomes g = 0.49, holding up in the more rigorous studies | Not specific to language, and not about full games |
| Hulstijn and Laufer (2001) | Ten new words met by reading, reading plus gap-fill, or writing a composition | Adult EFL learners in two countries | Retention highest after the composition, lowest after reading alone | The tasks took different amounts of time |
| Karpicke and Roediger (2008) | Foreign-language word pairs: repeated study against repeated testing | University students | Repeated testing had a large effect a week later; repeated study had none | A lab task, not a game |
| Webb (2007) | New words met 1, 3, 7 or 10 times in context | 121 Japanese learners of English | Each extra meeting improved at least one aspect of knowing the word | Full knowledge may need more than ten meetings |
Read the first four rows together and the picture is consistent. Games and game elements tend to beat the activities they were compared with. The last three suggest why. The tasks that kept words best made learners retrieve them, work with them, and meet them again.
Who took part in these studies?
Much of this evidence comes from older learners. Sailer and Homner (2020) found that most samples measuring knowledge gains came from higher education; only three came from schools. Karpicke and Roediger (2008) tested university students, and Hulstijn and Laufer (2001) tested adults.
There is some reassurance for teachers of children. Chen, Tseng and Hsiao (2018) found that the effect of vocabulary games did not vary with learners' age or first language. Their pool held ten studies, so that result is a hint rather than a settled answer.
All three language meta-analyses in the table were about digital games. The card, board and team games that fill most classrooms sit outside them, so the figures do not transfer to those games directly. For paper games, the mechanism is the better guide.
What were the games compared with?
This is the question to ask of any study. Tsai and Tsai (2018) sorted their 26 studies by design. In 10 studies, games against other activities gave a large effect. In another 10, a game with a feature added or changed beat the plain game by a medium effect. Only two studies gave the comparison group the same content by conventional means, and there the effect was medium to large.
A game that beats no extra practice tells you little about whether it beats a good worksheet on the same words. Sailer and Homner (2020) saw a similar pattern for gamification. Effects on knowledge were larger against passive teaching (g = 0.81, two samples) than against active teaching (g = 0.47, eight samples).
Definitions also differ. Dixon and colleagues (2022) found that only three of Tsai and Tsai's 26 studies met their own definition of a game. One included study used digital crosswords and tile-moving puzzles. Two reviews of games can be reviews of quite different things. Our article on gamification in the language classroom looks at gamification on its own.
Do the gains last?
Retention is the point of the question, and here the evidence is weakest. Dixon and colleagues (2022) found smaller effects on delayed tests than on immediate ones. On delayed tests, games showed no clear advantage over comparison groups, because the confidence interval crossed zero.
Learners did keep much of what they gained: delayed scores were still well above pretest scores. But that is true of many kinds of teaching. What the studies cannot yet show is that a game keeps words better than an equally active task without a game.
One session rarely makes a word stick, in a game or anywhere else. Webb (2007) found that each extra meeting with a word, up to ten, improved at least one aspect of knowing it. He suggested full knowledge may need more than ten. Our article on how many times a learner must meet a word sets out the studies.
Why some games work and others do not
If the game studies are thin, the research on how words are learned is not. Karpicke and Roediger (2008) found that recalling foreign-language words, not restudying them, drove memory a week later. A game that makes each player pull a word from memory is a retrieval task with a score attached.
Hulstijn and Laufer (2001) found that the task matters too. Learners kept more words when a task created a need for each word, a search for its meaning or form, and an evaluation of how it fits. Writing a composition with the words beat a gap-fill, which beat reading alone.
Dixon and colleagues (2022) add a practical point. Gains were about twice as large when teachers added supplementary material around the game, although the confidence intervals overlapped. The game is one part of the lesson, not the lesson.
They also found that games made for entertainment did better than games made for language learning, though again the intervals overlapped. Being designed for teaching does not, on its own, make a game effective.
Put together, the honest conclusion is this. Games help vocabulary when they make learners retrieve and use words, and when the words come back over time. The evidence for any specific format, on its own, is thin.
How to tell a useful game from a time-filler
Before a game goes into your routine, ask six questions. A game that earns five or six yeses is doing vocabulary work. One that earns one or two is a warm-up or a reward, which is fine if you count it as one.
- Have pupils met these words before? A game is a poor first meeting. The words should come from something the class has read.
- Does every pupil retrieve, or only the fastest? Count how many pupils produce a word in a round, not how loud the room is.
- Is the word hidden? Producing a word from a clue is retrieval. Spotting it in a list or a grid is recognition, a lighter task.
- Does the game need the meaning? Many letter games can be solved without knowing what the word means.
- Does anyone use the word? Explaining it, judging whether it fits, or putting it in a sentence adds the evaluation Hulstijn and Laufer (2001) describe.
- Will the word come back? A game played once on a list nobody sees again is a time-filler, however lively.
Classroom example
Two games use the same eight words from a story about a flood. Game A is a word search with the words printed under the grid. Pupils have met the words, but nothing is hidden, meaning is never needed and nobody uses the words. It earns one yes. Game B is back to the board with the same words, then one sentence per pupil. It earns five yeses, and a sixth if the words return in next week's quiz.
Our catalogue of memory games sorts ten classroom games by level and by whether they ask for recognition or recall. Do word searches teach vocabulary? shows how to rescue the weakest format.
A simple check you can run in your own class
You do not need a research design to learn something about your own games. This is a rough check, not a study.
- Pick 16 words from the next story and split them into two sets of eight of similar difficulty.
- Read the story as usual. Practise set A with your chosen game twice over the next fortnight. Give set B no extra practice.
- Three weeks after the reading, give a short quiz with a gapped sentence for each word and no word list.
- Compare the scores for the two sets. Repeat with the next story before you draw any conclusion.
If set A does no better, the game is adding little beyond the reading. If it does better, you have shown that extra practice helps, which you expected. The more useful comparison is two different games on two sets of words.
How iRead handles this
iRead's four word games are regenerated every 24 hours from the vocabulary of the story the class is reading, with no teacher setup. Every game clear and quiz answer feeds one progress record per word. A clear without hints counts double, and a word cleared on two different days is marked Mastered. Teachers see retention per word, per game and per class.
The core principle: A game builds retention when each pupil has to retrieve a word they have already met and do something with it, and when the word comes back on another day.
See it in iRead: Four word games are regenerated every 24 hours from the story your class is reading. A word cleared on two different days is marked Mastered, and teachers see retention per word.
See the vocabulary gamesKey takeaway
Word games are a reasonable way to practise vocabulary, not a proven shortcut. Keep the ones that make every pupil retrieve and use words, and bring those words back.
Frequently asked questions
Do word games help vocabulary?
Usually, if the game makes learners retrieve words they have already met. Meta-analyses of digital vocabulary games report medium to large gains, but most compared games with weaker activities and tested straight away. Evidence that games keep words better than an equally active task without a game is limited.
What does research say about games for vocabulary retention?
Immediate gains are well documented; lasting gains less so. Dixon and colleagues (2022) found that the advantage of games shrank on delayed tests and was no longer clearly above comparison groups. Research on retrieval explains why some games work: recalling a word strengthens memory more than seeing it again.
Is gamification the same as game-based learning?
No. Gamification adds game elements, such as competition or a story frame, to an ordinary task. Game-based learning uses a full game. Sailer and Homner (2020) found positive effects of gamification on learning, mostly in studies with university students. Studies of vocabulary games are a separate body of work.
Do word searches count as vocabulary practice?
A word search with a printed word list is mainly recognition: pupils match letters and need not know what the words mean. It suits beginners and early revision. Replace the list with definitions or gapped sentences and add a sentence-writing task, and it becomes real vocabulary practice.
How long should a vocabulary game last?
We suggest five to fifteen minutes. A game is worth its time while most pupils are still retrieving words, and that stops once the same few pupils answer. Several short games on the same words across a week spread the retrieval over more days. No study has fixed an ideal length.
Sources
- Tsai, Y.-L., & Tsai, C.-C. (2018). Digital game-based second-language vocabulary learning and conditions of research designs: A meta-analysis study. Computers & Education, 125, 345-357. doi.org/10.1016/j.compedu.2018.06.020
- Chen, M.-H., Tseng, W.-T., & Hsiao, T.-Y. (2018). The effectiveness of digital game-based vocabulary learning: A framework-based view of meta-analysis. British Journal of Educational Technology, 49(1), 69-77. doi.org/10.1111/bjet.12526
- Dixon, D. H., Dixon, T., & Jordan, E. (2022). Second language (L2) gains through digital game-based language learning (DGBLL): A meta-analysis. Language Learning & Technology, 26(1), 1-25. hdl.handle.net/10125/73464
- Sailer, M., & Homner, L. (2020). The gamification of learning: A meta-analysis. Educational Psychology Review, 32(1), 77-112. doi.org/10.1007/s10648-019-09498-w
- Hulstijn, J. H., & Laufer, B. (2001). Some empirical evidence for the involvement load hypothesis in vocabulary acquisition. Language Learning, 51(3), 539-558. doi.org/10.1111/0023-8333.00164
- Karpicke, J. D., & Roediger, H. L. (2008). The critical importance of retrieval for learning. Science, 319(5865), 966-968. doi.org/10.1126/science.1152408
- Webb, S. (2007). The effects of repetition on vocabulary knowledge. Applied Linguistics, 28(1), 46-65. doi.org/10.1093/applin/aml048