Nature or nurture? Heredity or environment?

[Paul McKeown]

Some people pick up chess/math/programming effortlessly

It is folly to assume that chess, mathematics and programming are a single skill, or use the same neural mechanisms. More plausibly one might have said that "some people pick up Spanish/Yoruba/Mandarin effortlessly". Or "some people pick up "chess/lacemaking/photography effortlessly" might make about as much sense.

[Brian Towers]

Brian I am not sure a control group was needed as it correlated results in various puzzles (as a measure of chess skill) with practice undertaken and IQ.

Alan, this is the part I find problematic:

The present study investigated the chess skill of 57 young chess players

There were no youngsters who were non-chess players at the start. The chess players were, no doubt, of different ages and different stages of their chess development, having enjoyed different standards of chess "practice".

A rigorous study would take a much larger number of non-chess playing children and train them up over a period of time giving different groups different amounts of training and forbidding them to play or train on their own before trying to make comparisons within age and ability (IQ or whatever) groups.

Going back to my previous objections. Consider when, or if, children master differential calculus. With the possible exception of somebody like John Stuart Mill, I don't think there has ever been a case of somebody mastering it by the age of 7. By the age of 10 there are several examples. In the 80's I think there was a 12 year old girl who made history by going up to Oxford to study maths, thereby smashing John Nunn's record as a 15 year old of being the youngest Oxford undergraduate since Cardinal Wolsey. I believe there was also the case of an 11 year old boy who started a maths degree as an external student at London University. Most people who are going to master differential calculus do so in the age range 16-18. A lot of people are never going to master it because they just aren't clever enough. Nobody would think it sensible to compare the differential calculus ability of different age groups of children.

I think chess is similar, although perhaps less extreme. A lot of 7 or 8 year olds can learn to play chess but I still don't think it makes sense to try and compare the chess playing ability of 8 year olds with 11 year olds with 14 year olds.

Character also plays an important part. From personal experience, I taught my 3 boys to play chess and comparing their abilities at age 11 is quite striking. The two older boys are very bright, the youngest much closer to average. The oldest is fiercely competitive and was easily the best at 11. He declined my invitation to attend my club's junior section after school because that would have interfered with his judo at which he alter went on to finish 3rd in the national championships for his age and weight category.

The middle one, while competitive socially, is completely non-competitive in sporting terms. My one abiding memory of his competitive instincts when we used to send him to the local swimming club is of him stopping in the middle of a race to go back and chat to a pretty girl he noticed on the side of the pool [slaps forehead in frustration]. On the other hand he appears to have no difficulty finding girlfriends despite currently having the face of a smallpox victim.

Bottom line on the character front: different kids are going to get much different benefits from the same amount of "practice" despite similar levels of intelligence.

The youngest, currently 11, while not as good as his brothers were at the same age, beats his more intelligent friends (based on the assumption that there better marks at school indicate higher IQ) on similar amounts of "practice". His secret? A small amount of his practice is with me teaching him basic stuff like KQvK, KRRvK, KRvK, KPvK, Nimzowitch 101 which means he knows how to win a lot more winning positions than his mates (and so also knows what to aim for) and, hopefully, he get his pieces out of the box a bit more efficiently.

So, the bottom line on that is that not all practice is equal.

[Arshad Ali]

Some people pick up chess/math/programming effortlessly

It is folly to assume that chess, mathematics and programming are a single skill, or use the same neural mechanisms. More plausibly one might have said that "some people pick up Spanish/Yoruba/Mandarin effortlessly". Or "some people pick up "chess/lacemaking/photography effortlessly" might make about as much sense.

Mea culpa -- I meant chess or math or programming. Didn't mean they're a single skill, et cetera.

[Alan Kennedy]

Um, dunno about that. He's showing that IQ test results are not correlated with chess prowess. That doesn't mean practice is a bigger component.

Arshad that is exactly what he is saying that practice is the biggest component particularly from the elite subsample. To quote "Although practice had the most influence on chess skill.

A rigorous study would take a much larger number of non-chess playing children and train them up over a period of time giving different groups different amounts of training and forbidding them to play or train on their own before trying to make comparisons within age and ability (IQ or whatever) groups.

I disagree. Merim is clearly looking at the factors influencing chess skill regardless of the starting baseline assessment. From a practical viewpoint I would very much doubt that Oxford University would have granted him a PhD on the basis of a non "rigorous" study and in any event i would suggest the study is a lot more rigorous than the anecdotal evidence and opinion presented on this discussion topic.

[Brian Towers]

Merim is clearly looking at the factors influencing chess skill regardless of the starting baseline assessment. From a practical viewpoint I would very much doubt that Oxford University would have granted him a PhD on the basis of a non "rigorous" study and in any event i would suggest the study is a lot more rigorous than the anecdotal evidence and opinion presented on this discussion topic.

I recently did some work for a pair of academics writing paper(s) on a cross-disciplinary subject. It was fascinating listening to the professor of molecular biology and the professor of linguistics at one stage compare the relative levels of rigour required for publication in their two disciplines.

After that I would not automatically expect high levels of rigour from social science papers regardless of the source.

[Michael Farthing]

I disagree. Merim is clearly looking at the factors influencing chess skill regardless of the starting baseline assessment. From a practical viewpoint I would very much doubt that Oxford University would have granted him a PhD on the basis of a non "rigorous" study and in any event i would suggest the study is a lot more rigorous than the anecdotal evidence and opinion presented on this discussion topic.

I find it very hard to see how any meaningful results can be achieved from a sample size of 57, and even less so from 23. In one of the source quotations earlier in the thread (referring to the 23 elite sub-group) there was an attempt to explain away a negative correlation of 'intelligence' with chess improvement by reference to a negative correlation between intelligence and effort. This would only have any validity if a proper multi-variate correlation had been carried out. On a sample of 23? I just can't buy it.

[Alan Kennedy]

There is a whole raft of statistic theory about whether a sample size is meaningful or not it depends on your population. Given that in my profession (accountancy) sample sizes are between 30-60 in order to achieve 95% certainty some comfort can be drawn. In any event I am not sure that Merim was trying to extrapolate his results to the population as a whole but merely to to say in the study undertaken we found that.... I would repeat that the study is a lot more meaningful than the anecdotal evidence and opinion above. As regards whether a proper multi-variate correlation was carried out I cannot imagine, given that psychology is in a large part about statistics, that it was not. However to be certain you will have to read either the PhD thesis or the published papers (or to write to Merim or Peter McLeod). Both are well respected Psychologists See for example https://chessprogramming.wikispaces.com/Peter+McLeod

[Alan Kennedy]

Since posting the above I have found an 39 page abstract that is available to download including details of the statistical analysis for those interested. Personally, I am losing the will to live. http://core.ac.uk/download/pdf/333373.pdf