Do musicians have better executive function?
When I started my PhD on language and music in memory I spent a good few months, as you do, reading everything I could get my hands on to do with my new topic. I wanted to know all about the evidence for language and music processing overlap. Fairly quickly I came upon “music transfer effects” –typically correlation studies that test musicians and nonmusicians on cognitive tasks to see if there is a group difference. If the groups are well matched and the musicians do better on the task then the temptation is to suggest that the music training improved cognitive function in this area.
There are several pitfalls to this type of experiment:
1) Matching: Unfortunately not many studies go to the necessary lengths to ensure that the samples are well matched. If you want a guide on know how to do it really well then look at Glen Schellenberg’s IQ studies – you have to have A LOT of background information!
2) Attribution: How do you know that the musical experience of the musicians is the causal factor? Maybe the musicians started out with the cognitive difference you have found (e.g. better memory function) and that contributed to their success in music in the first place?
3) Confounding variables: How do you know that some other factor that you haven’t measured is not actually the one leading your effect?
4) The bottom draw problem: If one study shows a positive effect then there is likely to be at least one that failed to replicate it sitting in someone’s draw somewhere because they can’t get it published. Take it from one who has been burned: always pilot effects yourself before relying on them as gold for your own experiments!
Having said all that, there are some very good studies that have shown correlations between musical training and improvements on a variety of cognitive skills – and luckily for me during my PhD a lot of them concern music and language and/or music and memory! In my paper you can see a fairly up to date summary of this evidence.
Since the finding of a relationship between musical training and higher IQ, there has been an idea kicking around that music training might improve executive function (i.e. It is a candidate for a ‘third variable’). Executive function covers many things but is basically the ability to manage and allocate attention resources, and it is a crucial component of working memory (linked to higher cognitive skills such as the ability to reason, learn and understand).
However, when I went to ICMPC last year Glen Schellenberg gave a very convincing talk about his research where he had apparently found no evidence that musical training leads to better executive function. You can see a blog I wrote about it here. So I thought that was the end of that story – nice idea but no place in reality.
So this week I was surprised to see a paper looking at musical training and working memory. But I read on, intrigued as it is an EEG study. The authors exploit a well established link between working memory and an ERP component known as the P300. Both the latency and amplitude of the P300 have been related to working memory function in different tasks: higher amplitude = an easier working memory task and an earlier latency = better performance.
Tasks: The authors chose some standard phonological, visuospatial and executive function working memory tests, as well as an auditory and a visual odd ball task (to look directly at the P300).
Findings: Group accounted for a modest but significant average of around 35% of the variance in the behavioural tasks – meaning musicians did do a bit better on every task. In the odd ball tasks musicians showed larger P300 components compared to nonmusicians, especially at medial sites (5.62mv vs. 3.49mv) and earlier latencies (362.23ms vs. 403.63ms). The same patterns were seen in visual odd ball task.
Conclusions: The musicians seemed to do better at everything! This is the first paper to look concurrently at behavioural and ERP components of working memory in musicians and nonmusicians and the authors claim to have found evidence to support the theory that musical training improves aspects of executive function. But remember my earlier warnings:
1) Matching: the groups in this paper appear to be only matched on age! The authors themselves mention the possibility that maybe they should have taken measures of IQ – a definite in my book. So my advice is that the group differences should be taken with a pinch of salt.
2) Attribution: It is only a correlation – we know nothing about these peoples abilities before they started musical training.
3) Confounding variables: There are no measures relating to possible confounding variables like IQ, attention or perceptual skills (auditory or visual).
So while it is good to see this paper combining techniques in such a clever way, I have some concerns about the general nature of the conclusions. But please don’t take my word for it – these are only my opinions! Read it for yourself and see what you think. The whole idea of science is to replicate, improve, and as part of the process to move closer to an answer.
For the meantime, in my mind, the jury on musical training and executive function is still out.
Paper – George, E.M & Coch, D (2011 – in press) Musical training and working memort: An ERP study. Neuropsychologia. Access here
