Finally – some sun on this little island!
I am enjoying this delightful early spring day with a small decaf coffee (some of you may know I am not able to consume caffeine) and a faceful of bright sunshine. What a difference it makes. The daffodils are coming out in York!
Today’s blog takes its title from a new paper by Reyna Gordon and colleagues (USA) that I am sure will be of interest to many of you. The paper in question is a meta-analysis, a powerful review of multiple studies.
Learning to read makes heavy demands on the process of phonological awareness. That is, the ability to focus attention on individual phonological units within syllables/words, and to sound them out effectively in your mind.
It is one of the early crucial stages of alphabetic decoding.
Good basic auditory perception is a necessary foundation for phonological awareness and we know already that musical training is associated with improvements to auditory perception skills (Shahin et al., 2003; Kraus et al., 2014). This and other links between perception skills, neural changes and language skills are detailed in Patel’s OPERA framework hypothesis (2014)
However, this relationship between music and reading it is not direct. It offers us insights into potential mechanisms to effect, why music training might improve literacy, but when we add up all the evidence to date is it really true that introducing musical training can help boost reading skills?
First – two notes on the need for caution.
1 – Measures of music aptitude have been found to account for over 40% of the variance in reading performance in typically developing 8-13 year old children who have little to no musical training. This means we need to account not only for musical training but the baseline musical abilities of children. Hence, a lot of studies that just compare musicians and non-musicians without decent controls need to be taken with a pinch of salt.
2 – Another important point is that, yes, it is exciting to show brain changes in response to musical training. However, these findings need to be translated into real behavioural outcomes that matter in the classroom. If a studies outcomes can’t be expressed in this way then they are minimally useful to those who are seeking to support the development of effective strategies for impacting on academic achievement.
The present paper’s meta analysis can add in a number of ways to the ongoing debate:
1 – It focuses on studies that show measurable behavioural outcomes
2- It focuses on studies that have control groups
3 – It helps to identify common attributes of the different musical training paradigms that might be effective
4- It helps to account for variability across results, so it can factor out elements such as the age of children and the quantity (amount) of training received
The first step in any meta analysis is a good trawl of the literature. The present authors searched 12 of these databases with key terms to uncover 4855 related articles. That is a lot.
These papers were reduced to 178 by imposing selection criteria (only peer reviewed journal articles in English). A second stage then saw studies excluded if they did not have a control group, did not assess reading pre- and post- intervention, did not provide equal reading training across groups, or did not provide sufficient data to allow the authors to extract the effect size.
That sweep left 12 articles and 13 individual studies (one paper had two separate studies).
Participants in the studies ranged in age from 4.53 – 9.33 years. Range of musical training quantity involved in each intervention varied hugely from 3 to 90 hours. Nearly all the studies featured singing, 9 featured rhythm training, and smaller collections featured visual notation training, rhyming, or clapping/ marching.
IQ was reported as equivalent in 9 out of 13 studies and socio-economic status only in 6 out of 13. Only 6 of the studies used ‘true’ random assignment. Control groups varied, some doing phonological training (3), some doing sport/ arts (3), and some using a ‘no treatment’ alternative (6), which is less powerful than a true matched control. These figures go to show how few papers apply a good gold standard of decent control to their studies, something we should really encourage in future work.
The final analysis of musical training influence on all phonological awareness scores revealed a modest effect size of 0.20, a small but significant difference between children who had musical training vs. those who do not.
- When broken down, the effects of music on rhyming ability was non-significant. The authors model suggested that a minimum of 40 hours of musical training would be necessary for an effect to emerge for this particular reading related outcome.
- There were only weak trends towards significance for other phonological outcomes and reading fluency when measured individually. In these cases there were no measurable impacts of a child’s age or the amount of training received.
Although these results might seem a little disappointing at first glance there is good reason to be excited. Meta-analysis results like these are powerful and having any effect stand up to this amount of close scrutiny is important and interesting.
Gordon and colleagues have reported a modest effect of musical training on the broad category of phonological awareness. The rest of the statistics probably fall a little flat due to a lack of power. We need more well done large studies to assess the true impacts of different sub-skills within the category of phonological awareness.
The finding regarding minimum hours is worth following up on. Many psychology studies report 100’s if not 1’000s of hours of training are necessary to have measurable impacts. Here we have evidence that, yes, one or two lessons won’t do much, but a good year (assuming the usual pattern of lessons once per week) will likely foster a positive change in rhyming ability. None of the studies that showed an improvement in rhyming as a result of musical training added any rhyming training to their intervention and instead, most strongly featured rhythmic work. The curious case of the power of rhythm continues…
All in all the present authors describe their results as ‘somewhat inconclusive’ with regards the effect of music education on reading skills. I like a cautious response, especially when it is backed by data. The underlying problem is the variability in the way that reading is tested and the way that musical interventions are run. Without more consistency across both factors it is difficult to get a clear picture. The hints are there but the strong evidence has yet to emerge.
I enjoyed the following quote provided by the authors:
“...in the field of skill learning, transfer of learning from the trained task to even other very similar tasks is generally the exception rather than the rule”
(Green & Bavelier, 2008)
That idea is consistent with what I have observed in psychology over the years, especially in my field of memory. Memory experts tend to misplace their keys just as much as the rest of us…but boy, can they count cards!
A cautious approach to the question of musical training transfer is a welcome voice in the field. We don’t want to run before we can walk and make strong claims without: 1) the evidence to back them up; or 2) the important knowledge regarding what aspects and how much training are necessary. This is what educators really need to know and what will make the case of music in education maximally impactful.
I will end with the words of the present authors – I couldn’t agree more.
“The present finding converges with the hypothesis that music supports phonological awareness; further study is needed to determine if intensive and long-term musical training can enhance reading fluency via improvements to auditory skills, phonological awareness, and rhyming in particular”