Biological and Social Aspects of Intelligence

The Institute’s thirty-second annual symposium was held on 21 & 22 September 1995. The following brief synopses of the first six papers can only hint at the breadth and depth of material covered. The rest of the papers will be reported in future Newsletters. All thirteen papers will be published in full as a single volume in due course.

Does general intelligence ("g") as measured by standard IQ tests matter? Some commentators argue that other mental, moral and social factors influence life chances so strongly that the residual difference that general intelligence makes is unimportant. Do these views stand up to scrutiny?

Those who depend upon predicting employment success for a living - the educational psychologists who devise selection programmes for employee recruitment - appear to have little doubt that IQ tests are good predictors of performance at work. The critics of IQ tend to ignore this work and argue that differences in success can largely be attributed to the other factors discussed below.

Mental factors. Creativity and personality are among the postulated factors to be taken into account. But all attempts to measure creativity lead to tests that correlate strongly with standard IQ tests and so are not measuring something different at all. However, it has to be acknowledged that g is a better predictor of success for those with low IQs than for more intelligent people who may be more selective in which talents they develop.

Moral factors. Conscientiousness and criminality feature here. However, there is no measure available to show the benefit of hard work other than as a means of exploiting intelligence. Criminality is easier to measure and low IQ is an extremely good predictor with the mean IQ of convicted criminals in the US in the range 85 to 90. Rates of crime among blacks are no higher than would be predicted on the basis of their average IQ.

Social factors. Length of schooling is a social factor thought by some to influence later intelligence. However, IQ at the start of schooling is a twenty-fold better predicter of IQ at the end of schooling than the number of years of education in between. Some critics of IQ think that parental socioeconomic status is the most important predictor of success in life but recent studies such as those cited in Herrnstein & Murray’s The Bell Curve have shown that this is fallacious and that IQ is much more important, especially at the lower end of the range.

So, consideration of critiques of IQ has pinpointed areas of human affairs where g is even more important than previously recognised. It is clear that all research into social problems ought to take IQ into account. It is time for neglect of the role of intelligence as a crucial social factor to end.

There are two directions in which the study of g may be extended:

• "upwards" to look at the effect on success in real life

• "downwards" to seek independently measurable components of intelligence at the physiological level.

A candidate for an acceptable component of intelligence should meet the following criteria:

1. A validated physiological process

2. High correlation with IQ

3. Individual differences in the laboratory should not be merely the result of being clever - they should be a cause, not a consequence, of differences in intelligence

4. A mechanistic explanation for the association between psychometric intelligence and the laboratory tests should be proved

Figure 1

A task which has been widely studied is that of inspection time. Subjects are presented with either figure A or figure B (see Figure 1) which is quickly masked by figure C. The length of exposure to A or B necessary to accurately discriminate between them varies from person to person and is inversely correlated with IQ (r = -0.4 to -0.5). The task has the added advantage that it does not involve motor speed and so success should depend solely on quick thinking.

This research is criticised by some who argue that the ability to discriminate quickly is not a cause but a consequence of intelligence, perhaps because the more intelligent are more motivated or because they are better able to focus attention. One approach to resolving these issues may be to break down the process even further into its components and the measurement of evoked potentials, the electrical signals generated in the brain while the task is being performed, is now being undertaken.

Intelligence cannot be measured, only inferred from performance on certain tests, including IQ tests, It is all too easy to read into the results of such tests far more than the simple measurements of time and mistakes will support.

Interestingly, performance at almost all tasks - including purely motor tasks - shows some correlation with IQ. The effect of practice merits some study. In general, poor performers improve more with practice than good performers. Comparison of mean performance levels can be misleading. Subjects are normally asked to repeat a task many times and the effect of practice is to eliminate the "poor" end of an individual’s performance distribution. This has a marked effect on the mean but may obscure the fact that there has been little or no improvement in the level of the individual’s best scores, only in their relative frequency.

It is also of note that performance distributions are highly skewed, making the mean a very poor measure. People with low IQ scores show much more variability than the more intelligent. Perhaps this variability is a more interesting measure than the mean.

UNICEF programmes in developing countries have looked in some depth at survival of babies and health of mothers but there has been very little research relating to children of school age. With more children than ever before going to school, it is important to know whether health impacts on ability to benefit from education.

A major health problem in developing countries is infection by parasites such as intestinal worms. The degree of infection is often greatest during school-age years. The speaker’s research in Jamaica, involving double blind clinical trials, showed a major improvement at fluency tests in the treated group compared with the group given only a placebo. However, these results have been hard to replicate. This raises questions such as:

• Are effects of some infections permanent?

• Is longer treatment required?

• Are studies flawed because of inadequate testing conditions?

• Are inappropriate tests being used?

• Should groups be classified by severity of infection rather than just as infected or not infected?

It is hard to know how long someone has been infected; some longitudinal studies might help. We need to understand the mechanism of recovery - if effects of infection are permanent then we must focus on prevention rather than recovery. If educational conditions are poor, does health status make any difference?

The brain accounts for 30% of basal metabolic rate. It uses glucose almost exclusively and needs a continuous supply - the glucose reservoir in the brain would last only a few minutes. The following table shows the effects of various levels of blood glucose:

Only the range 55 to 180 mg/dl will be of concern here.

Regional differences in glucose metabolism are limited in the resting brain. It takes 40 minutes after restoration of glucose levels following deprivation for brain function to return to normal. Various experiments with glucose and placebo drinks show that glucose improves performance in tasks such as story recall and verbal fluency tests. Glucose tolerance (the ability to bring down blood glucose levels quickly after administration) seems to be related to cognitive function, especially memory-related tasks.

Other experiments show that glucose demand is related to intensive use of particular brain areas. Thus simultaneous performance of two spacial or two verbal tasks is affected by glucose administration but performance of one spacial and one verbal task is not.

Observations indicating the role of glucose in brain function include the following:

1. Intraventricular glucose enhances memory

2. Intraventricular glucose increases acidic Fibroblast Growth Factor which facilitates memory

3. Glucose increases synthesis and release of acetylcholine

4. Glucose facilitates active areas of the brain

These observations, and others showing that peripheral glucose levels have little effect, suggest that glucose acts directly on the brain to enhance memory and cognitive function.

An opportunity arose in the 1980s to study the effect of early diet at two UK centres where pre-term infants were given different feeds, either on their own or as a supplement to expressed breast milk. The choice whether or not to express milk was the mother’s, but within these groups the choice of alternative or supplementary feed was randomised. There were then eight different combinations of feed, with four randomised pairings, as shown in figure 2.

The intervention was applied for a maximum of 4 weeks. The children were assessed at 18 months and again at 7½ to 8 years.

The only significant difference within a randomised pairing was a major advantage for PTF of TF. When assessed over the group as a whole, the advantage applied to males only; for a selected subgroup who received the regime for longer, the male advantage was even bigger and there was a female advantage too.

There was also an advantage in the groups with expressed breast milk over those without, but since allocation was not random but on the basis of maternal choice it is possible that the difference is due to other aspects of parenting that correlate with this choice rather than to the special composition of expressed breast milk. However, the differences persisted even when adjustments were made for social class. There was no difference between banked breast milk and PTF, although the latter contains twice as many nutrients.

If the composition of fresh milk is a factor, long chain polyunsaturated fatty acids are a component that might be responsible. Diets supplemented with these have been shown to affect retinal development and improve visual acuity. However, human milk composition is still poorly understood.

Whatever the reason, it is clear that early diet really does matter.