More of a pamphlet than a book :)

IQ tests to g-factor as physical fitness tests to general fitness. We have an intuitive understanding of fitness not as a concrete attribute in itself, but as a correlation between various physical attributes.

If you were recruiting people for physical labor then a general fitness score would still be a useful filter. On the other hand, if you were recruiting bench-pressers you would want to look at the results of that specific test.

Debate on whether g-factor is measuring a single attribute (eg processing speed or working memory) vs measuring multiple attributes that are each used in most of the tests.

Dynamic mutualism - theory that g-factor is made up of many attributes that correlate because they help each other develop eg larger working memory might make it easier to develop abstract reasoning skills.

Fluid intelligence declines with age. Speed-related tests decline the most dramatically. Possibly protected against by remaining physically active (I assume this evidence is correlational so far).

Deary 2013 gave 92yos the same IQ test they sat at 11, found r=0.54

IQ correlates positively with variety of measures of health, even when controlling for class and in countries with free health-care. Probably causality in both directions - healthier genes lead to better brains, smarter people make better life and health decisions.

IQ strongly predicts educational success (but education also increases IQ). Moderately predicts income, job performance (in complex jobs), success in work training programs, class (but class also affects IQ). Weakly predicts job performance (in simple jobs), creativity (as measured by eg alternate use tests), leadership

IQ negatively correlated with racism, sexism, support of authoritarian parenting and policing (no mention of how the first two were measured). Positively correlated with social and economic liberalism.

IQ mildly correlated with happiness (although notes that much of the research on this is debatable).

IQ correlated with short-sightnedness. Plausibly caused by more reading at a young age.

Possibly correlated with higher odds of bipolar disorder.

Heritability typically measured at ~0.5 across twin studies and adoption studies (although see ‘Intelligence and how to get it’ for arguments that this is biased upwards). Genomewide Complex Trait analysis finds similar numbers. Remainder seems mostly due to non-shared environment - surprisingly little effect from shared environment (possibly because there isn’t much variation in shared environment across rich nations to begin with?).

In countries with high inequality, heritability is higher in upper class families. Can we use this to ballpark the maximum effect of genetic vs environmental changes?

IQ is polygenic - small contributions from many genes rather than large contributions from few genes.

Correlations between IQ and brain volume around 0.24

Parieto-Frontal Integration Theory - theory that IQ depends partly on connection between frontal and parietal lobes. Support: Efficiency of white matter connections (as measured by Diffusion Tensor MRI) correlates with IQ. (Notable that white matter deteriorates with age.) fMRI shows high activity in front and parietal lobe during IQ tests, and also less activity in high-IQ subjects when working on the same problem - suggests more efficient processing.

‘Mozart effect’ is bunk.

Mixed results for n-back. Heavy criticism of original experiment - no time limits for IQ test, no active control group.

Evidence for IQ increases from treating parasitic worms, from banning leaded fuel and from treating iodine deficiency. Little success so far for dietary supplements in healthy children.

Brinch and Galloway 2012 - Norwegian government added 2 years to mandatory schooling with staggered rollout across counties. 7.4 point IQ gain. Agrees with previous uncontrolled studies.

Pre-school interventions. Some raised IQ temporarily. Reverted by teen years, but other positive effects remained.

Flynn Effect. Nettelback and Wilson 2004 gave same IQ test to schoolkids in Australia 20 years apart. Vocab tests improved but speed tests remained the same. Suggest rising scores (at least in Australian schools) are due to cultural changes rather than faster brains from eg better nutrition.

IQ scores rising faster in the developing world. Not slowing over time in developed world.

Same mean for men and woman, but higher variability in men. (But men have higher brain volume, and brain volume correlates with IQ. What’s up with that?). (Not a huge effect eg at IQ 140 we see 58% men vs 42% women). Women have higher mean verbal scores, men have higher mean spatial scores.

Differences in race are possible, but the area is so toxic that little good research is being done.

Repeats Pinkers thought experiment - a world where IQ was totally determined by environment would have less social mobility and be more susceptible to governmental social engineering.

List of links to books, papers etc. Comment on ‘Intelligence and how to get it’:

…reaches many conclusions that are in conflict with the behaviour-genetic consensus on, for instance, heritability.