Big assumptions, poor data: this time James Flynn won’t change your ideas about IQ

    11 August 2016

    James Flynn’s career has always been about progress. As the first researcher who noticed that IQ scores were rising across the 20th century — well, perhaps not quite the first: as with Darwin and natural selection, others may have hinted at it, but he documented it most compellingly — he’s thought more deeply than anyone about what these intellectual changes mean for society. His new book, Does Your Family Make You Smarter?, is aimed at encouraging its readers to make another kind of progress: away from outmoded views about the immutability of IQ and towards an understanding that ‘all of us… have the capacity to choose to significantly enhance our cognitive performance’.

    It’s a big claim, but Flynn is no stranger to those. The generation-on-generation gain of about three IQ points per decade — a phenomenon we now know as the Flynn effect — seemed shocking when Flynn first described it in 1984, but it is now universally accepted by researchers (although its cause remains an enigma: a recent review lists, among others, education, exposure to technology, changes in the way we approach tests, pathogen stress, family size, and nutritional advances as potential explanations). Flynn frames this environmentally driven effect as the ultimate rebuke to those suffering from what he calls ‘post-twin pessimism’, the notion that twin studies have shown that the environment plays little role in shaping cognitive differences between people.

    Flynn notes that, despite increases in vocabulary and reasoning, his eponymous effect has failed to budge arithmetic skills — people are just as bad at maths as they always were. That’s a shame, because arithmetic is a great help in understanding the voluminous sums, tables, and appendices he provides in this book.

    His argument is that it’s statistically unlikely that exceptionally smart children will be born into homes where everyone else is as smart as them (after all, that’s why they’re ‘exceptional’). They can thus expect their IQ score to be dragged down, even just a little, by their family surroundings. The same thing works in reverse for children born with poorer intellectual abilities: they get a boost from their brighter family members. Flynn uses tables taken from IQ testing data to divine the extent of these compensatory or detrimental effects, which he argues can make the difference between being above or below the cut-off to enter university, thus having an important influence on people’s lives.

    To my mind, Flynn’s new method is simultaneously too complex and too simple. Too complex because it requires such a tottering pile of assumptions (one of which seems to be that your family can indeed make you smarter, which rather begs the question), and too simple because it feeds in no actual data except for IQ scores and the test-taker’s age: despite mentioning nature and nurture in the book’s subtitle, Flynn has no information on the specifics of each family’s environment (their social status or any parenting activities), or on genetics. His evidence is thus very indirect: why not rely instead on studies that actually measure these aspects?

    Such studies suggest that the family does affect IQ scores early in life, with the effects petering out as we age. But they also reveal complexities — like the idea of gene-environment correlation, where genes cause people to sort themselves (or be sorted) into corresponding environments, or gene-environment interaction, where different family environments alter the genetic effects on intelligence — that Flynn can’t hope to get at in unadorned IQ data.

    The second part of the book is an opinionated description of the many scientific theories of intelligence. Although some of the specifics are off — Flynn’s description of mental exercise ‘spraying’ the brain’s cortex with dopamine, thus thickening it, baffled a neuroscientist colleague when I read it to him — it’s a useful and wise discussion, and returns to Flynn’s theme of progress: using the history of physics as an analogy, he urges us not to let overly broad theories force us into one particular mode of thinking on intelligence, delaying scientific advances while researchers obsess over the precise predictions from a model that might not correspond well to reality. I heartily agree.

    However, we should also apply this reasoning to the new methods Flynn describes in the book. Instead of poring over old IQ data tables and extrapolating from them, the way to uncover any effects of the family that twin studies may have missed is to keep collecting ever more detailed samples and testing ever more nuanced hypotheses. Eventually, with proper progress in our data (as well as perhaps in our IQs), theory will follow.

    Stuart Ritchie is a postdoctoral research fellow in the psychology department at the University of Edinburgh. Follow him on Twitter: @StuartJRitchie