Friday, October 10, 2008

Computer games - astounding improvements in numeracy

Here's to you Mr Robertson
Brain Training did more for e-learning than any government campaign or product. It took e-learning mainstream.
The wonderful Derek Robertson has been using this sort of stuff in schools for ages but we now have an excellent piece of research from Learning and Teaching Scotland.

  • 600 pupils from 32 schools
  • 20 minutes at start of class for nine weeks
  • control group did normal class stuff
  • pupils tested at start and end of study
Brain Training group:
  • 50% better test scores than control!
  • 13.5 minutes to do test, control 18.5 minutes
  • more improvement in less able kids
  • no difference between boys and girls
  • reduced absences
  • reduced lateness
These results are outstanding. If replicable, they have huge implications in terms of a potential solution for our low numeracy standards.
While Derek is wrong in claiming that this is the, 'first independent, academic evidence that this type of computer game could improve attainment when used in an educational context', it's a damn fine piece of work.
Get these things into primary schools now! Better still, simply encourage parents to buy them for their kids. Perhaps we can see the politicians and educational establishment stop whinging about poor numeracy and doing something simple to solve the problem.


Anonymous said...

Derek's also got a great blog which will kep you up-to-date with some of the fantastic things he's doing through the Consalarium in Dundee. Link: Hot Milky Drink

Donald Clark said...

Thanks Neil - like this blog - so it's in my links bar.

Mark Frank said...

"50% better test scores than control!"

Not true. The mean improvement in test scores was 50% higher - a very different thing.

I am sure this is a great study but, as Robertson says, it needs to be properly published and peer reviewed before inflicting yet another initiative on primary schools. Until that happens we do not know for sure how good the study is. For example:

What was the selection and randomisation process? Were schools simply told - use this method - with no option? It seems unlikely. But if not, then there is scope for self-selection bias.

The results are "statistically significant" but there is no more detail than this - what test was used, what P value, confidence intervals, effect size etc? I don't blame him for not supplying this stuff - but until we can see it we don't actually know much.

Studies like this are obviously at risk from the Hawthorne effect.

You would expect the lowest performing students to improve anyhow just through regression to the mean. I wonder how the lowest performing in the control group did compared to the lowest performing in the trial group.

Donald Clark said...

Fair comment. I used the word 'replicable' as I agree that this study needs more scrutiny and is the sort of study that would really benefit from follow up research,

As you say, Derek has deliberately held back the detail until it has been peer reviewed and published. This is a sensible approach.

On the 50% figure, I wasn't trying to hype it up - just trying to be brief for the blog. As you say, the mean improvement is indeed the measured improvement. This is shown on a graph in the released report.

Just as impressive is the mean reduction in time taken to complete the task. here's the link to the short-form report that gives some detail on methodology and results.

Here's the link:

Mark Frank said...

I think the key is not whether the studies are replicable, but whether the results are sustainable. The Hawthorne effect looms large. I would guess almost any study where the kids were given something unusual to do for 9 weeks would lead to some improvement.

Anyhow this is all really interesting. Thanks.

Donald Clark said...

Derek worked with David Miller to set up this trial, David teaches 'Research methodology' at the University of Dundee, so he was keen to do it properly. Derek is also honest about it being a 'small scale intervention'. This gives me some faith that an evidence-based approach is being followed.

Peer review may uncover some of the weaknesses, but this was a trial well worth doing, as it is likely to lead to follow up research around some key hypotheses. Thanks for the comments.

Donald Clark said...

Well done to Derek for winning the award at Handheld learning yesterday. He certainly put a few of the traditional researchers' noses out of joint. Well deserved as he had the balls to do something that was timely, relevant and could actually lead to change.

Anonymous said...

Our first small scale intervention has recently been peer reviewed and accepted for the British Journal of Education.

My colleague and friend Dr David Miller is working with me on my projects and he is guiding me through the finer details of the range of tests that we can apply through SPSS etc as well as helping to develop my understanding and ability to use a range of procedures for gathering qualititive data as well as quantitative.

Apologies for any sense of hubris in my claims for this being the first intervention of this kind. I am aware of the research that has been done do far but don't think that this type of approach has been done before...keep me right on this if you will.

Our extended trial was a randomised control trial. We also attempted to stratify the group base on free school meal allocation.

We are confident that our methodology and results and the independent integrity of what we have done can allow us to make claims for generalisability in terms of gains a school can expect by following a similar approach. We are also putting our extended trial up for peer review hence the reason for not sharing out fuller details. I'm sure that colleagues can appreciate my reasons for this.

Your comments are most welcome. I care about helping young people learn. I have a number of other projects currently ongoing and in the planning stage and am hugely motivated by what appears to be a hugely positive approach to help young people achieve far more of their potential than maybe they are at present.

I wish that other ICT based initiatives had as much scrutiny and enquiry aimed at them, (as my work has from certain quarters), prior to the huge investments that have been previously made.

Donald Clark said...

Excellent news. I look forward to blogging the full article. More power to your elbow! And well done on the award.

Mark Frank said...


I also look forward to reading the full paper. I can't find the British Journal of Education through Google or my Open University Library access. Do you mean the British Journal of Educational Studies or possibly the British Journal of Educational Technology?

Unknown said...


Thanks for the information. I would like to know more about it.


Luwow Goldman