Let me start by answering the question as posed – are digital games a valuable tool for learning? Valuable could mean a number of things; for example, are students more likely to come to school on days when digital games are used to teach?

Do students become interested in history because they play ‘Civilization’? Does playing digital games motivate the desire to create such games and to develop a range of creative and computational skills?

However, the most straightforward way to interpret "valuable" is to ask whether students can develop new skills or knowledge through interacting with digital games?

A number of reviews and meta-analyses (which combine statistical results from a number of different studies) have been conducted to answer this question. Research published in 1992[1] found 67 studies in the preceding 28 years of which the majority showed games to be equally effective as conventional instruction and a third found games more effective.

Two recent meta-analyses[2][3] with more developed criteria for inclusion have concluded that games are significantly more effective for learning and retention than conventional instruction. Clearly, it would seem the research has spoken and it has shown that games are indeed valuable.

But let's think a little more about the question. In using the word "tool" we might be reminded that hammers do not strike nails by themselves nor chisels carve wood. Analogously, it is by playing games that students have opportunities to learn.

Clearly a game downloaded but never opened cannot support learning but perhaps there are ways to play games that might be more valuable than others? Can we encourage learners to become mindful game-players? We do not think that students should learn solely by reading quietly by themselves in classroom. Instead, interactions with teachers and peers are crucial in helping learners to realise the potential of media.

Therefore, it's always worth remembering that in school contexts how digital games are integrated into the classroom will fundamentally influence their success3.

Finally, in the meta-analyses discussed above there were huge variations in the games researched and the outcomes in each study. Just like any other media, there will be more and less effective ways to design digital games for learning and some of these will be similar to other digital learning environments (e.g. are they suitably challenging, provide helpful feedback, etc.).

Jake Habgood and I[4] set out to test whether games that integrate learning content and fun are more effective that games which keep them separate (using games as a reward for "dull learning"). Children played a version of a specially designed maths game for two hours and participated in one lesson. The game, Zombie Division, helped children aged 8 to 11 practice division by slaying zombies wearing numbers with mathematical weapons.

We found that playing a game that integrated fun and learning was significantly more effective than playing the separate one (under strictly controlled conditions) and that children played the integrated games far longer when they free to choose.

However, what is even more striking was that children who played the separate version learnt no more than student who just studied the lesson (even though they practiced maths in the game for an extra two hours). And what is so worrying about this finding is there are far more of these types of game available for learners.

So to conclude, let's agree that indeed digital games can be a valuable tool for learning. However, a better question is to ask which games played in what ways and supported (where relevant) by which classroom practices create valuable opportunities for learning.

Shaaron is a Professor at the University of Nottingham and Deputy Director of the Learning Sciences Research Institute.

Nesta's games in education blog series seeks to encourage a wider discussion. To share your thoughts, ideas and questions, please comment below or tweet using #gamelearning