|Galton Institute Home Page||September 1999 Newsletter Contents||Newsletter Index|
Sir Cyril Burt was, and probably still remains, the best known of all educational psychologists, deservedly so and he was the first of them all, having been elected as an Assistant Lecturer in Educational Psychology at Liverpool in 1908, under Sherrington. From the start he was involved in the statistical assessment of the results of IQ testing, and especially with “factor analysis” on the extent to which performance in different subjects was independent, and whether or not there was some general factor of intelligence underlying them all. The arguments here were technical and sometimes rather acrimonious but Burt, although his mathematics was largely self taught, was well able to hold his own in them. This work was developed in his last major book, The Factors of the Mind (1940), much of which however is now of little more than historical importance.
In 1913, at the age of 30, Burt was appointed as psychologist to the London County Council (the first they had ever had, and originally part-time) to be especially concerned with backward, maladjusted and delinquent children. He produced several influential Reports on these, and two highly regarded books, The Young Delinquent (1925) and The Backward Child (1933). But in the 1920s he started also to assess the performance of normal children in schools, collecting an enormous amount of data based upon teachers’ reports, and the Stanford-Binet intelligence tests, and so on. He finally left the LCC in 1932, on succeeding Spearman as Professor of Psychology at University College, London.
At that time it seemed obvious that a child’s performance at school and indeed later on in life was related to its social class, but it was not clear whether this was because better-off people were able to give their children a better start in life, with education and so on, or whether intelligence and educability were to some extent inherited genetically, with a higher average level in the upper classes. Burt thought that he had established the genetical heritability of intelligence quite firmly, by comparing pairs of twins brought up together with others who had been separated soon after birth and placed in different adoptive families. Monozygotic twins, which were genetically identical, were closely similar in both appearance and habits, with similar IQ scores, whether or not they had been separated. Non-identical twins, on the other hand, were no more alike than they were to their other brothers and sisters, and the IQ scores of those that had been separated sometimes diverged from one another significantly more as compared with dizygotic twins brought up together under what would presumably have been the same environmental conditions. But what was finally achieved, as measured by IQ testing, was also affected by education and early upbringing, where those with better-off parents had an obvious advantage. That explained the strong correlation between academic performance at school and social class. But although the professional classes might well on average have a higher level of genetically inherited intelligence, since those lower down in the social scale were much more numerous, in absolute numbers they would probably have as many and quite possibly more children of high potential intelligence capable of profiting from a good education, which for economic reasons they were denied.
In the 1920s there were a good many Grammar Schools providing excellent education up to and including university entrance standards. But these charged fees which, although much less than the so-called “Public” schools, were beyond the reach of working class children, except for the very few who were successful in being awarded scholarships to them. This was not only unfair, but it was obviously economically disadvantageous that full use was not being made of all the abilities of the whole population. Burt therefore believed that all children should be given the best education that they were capable of benefiting from, up to and beyond the statutory school leaving age of 14, raised to 16 in 1945. This meant that the most intelligent children were to be given free places in the Grammar Schools, and Burt was largely responsible for this being done. It did of course involve selecting a minority, of around 20%, who could cope with the highly demanding Grammar School courses, with the majority going to “Secondary Modern” schools better suited to their more modest abilities. But no selection methods are perfect, and a single hit-or-miss test at the age of 11+ will certainly miss some late developers who could have flourished in a Grammar School, and hit others of lower ability, destined to languish at the bottom when they could have done much better with a less academically orientated syllabus.
So “failure” at the age of 11+, which usually had the irrevocable effect of closing the door to later University entry and to most professional careers, caused a great deal of discontent to the many who had only just failed, and especially to those of their often articulate middle-class parents who could not afford to opt out of the State education system altogether.
Burt’s views on selective education were becoming increasingly unpopular in progressive circles even before the war, in which his department in University College was bombed and was evacuated to Aberystwyth. It returned to London only in 1945, to find a good deal of destruction and disruption of the records. Burt, who was knighted in 1946, continued to work there for another 20 years after his retirement in 1950, but he found himself increasingly alienated from the new generation of mostly egalitarian-minded teachers and educationalists. He had always had a rather difficult personality, not taking kindly to criticism, and this was made worse by increasing deafness and the onset of Ménière’s disease. But up until his death in 1971 at the age of 88 he was highly respected, even if only as a relic from the very different pre-war age of educational psychology, although there had been doubts about the validity of some of his later publications.
However, on 29 November 1978 the Sunday Times printed a long article under the heading “Crucial data faked by eminent psychologist”, by an investigative journalist who had been advised by several members of the British Psychological Society (BPS) with axes to grind. In this Burt was accused of having fabricated and falsified his data on intelligence testing, to provide evidence for the genetical inheritance of educability. And in particular it was suggested that the three named “missing ladies”, who were supposed to have been responsible for collecting much of the pre-war data, were in fact fictitious since 20 or 30 years later they could not be found.
After this various other BPS members, who were politically opposed to Burt’s advocacy of selective education and the 11+ examination, hastened to come up with further examples of supposed fraud and falsification, culminating in the authorised biography by the late Prof. L S Hearnshaw published in 1979. And finally in 1980 the BPS organised a special Symposium “in which Burt’s deceptions should be seen in the wider context of scientific method in Psychology” in which his guilt was taken for granted, and with which it was hoped that the whole unhappy business would be closed and forgotten. This was followed by several books on fraud in science in which Burt, described as a “wicked old fraud” and a “senile liar” was pilloried as among the worst of the offenders. However, the question was re-opened by R B Joynson in The Burt Affair (1989), who argued persuasively that many of Burt’s supposed falsifications were not what they might have seemed to be. And the famous missing ladies have turned out not to be missing at all: the last of them, who had married and lived in Dublin, died there a few years ago but is survived by her grown-up children.
There is however no doubt that much of what Burt published after he had retired was muddled and badly presented, and would never have been printed had it not appeared in a journal of which he was editor, and with no proper reviewing system. But that is far from saying that there is evidence here of fraud. Some of the inconsistencies are so obvious that the data could scarcely have been intentionally falsified: even in old age, Burt could have made a better job than that had he intended to deceive!
Much has been made of the fact that there were suspicious increases in the numbers of separated identical twins which Burt claimed to have studied, from 15 pairs in 1943 to 21 in 1955, and finally a remarkable 55 pairs in 1966. The implication may have been that Burt had collected these after his retirement, when he would have had no access to school records, but he never actually said so, and it is now clear that the data were from the 1920s and 1930s, mislaid during the war and found only after his return from Aberystwyth, with no reason to think that they had been fabricated. Most unfortunately, after Burt’s death but before there had been any suggestion of dishonesty, most of Burt’s huge accumulation of papers were incinerated on the advice of an eminent educational psychologist, and these are known to have included several large boxes in which his secretary remembered finding some of the long-lost twins data sheets. Possibly this pre-war data would nowadays be regarded as of little value, but it would have been relevant to the question of dishonest fabrication. So, since Burt was in no way responsible for the destruction of the evidence after his death, he should be given the benefit of the doubt as to what it would have revealed.
But Burt’s method of comparing identical and non-identical twins separated and reared apart does represent a valid approach to the question of the genetical heritability of intelligence as measured by IQ scores. And this has now been repeated in the Minnesota study (Bouchard, T et al., Science 250, 223-8, 1990), on a sample of more than 100 twins and triplets collected after Burt’s death, which arrived at an estimate of about 70% for the genetical heritability of IQ scores, near enough to Burt’s earlier but now unverifiable figure of 77%.
The only proper conclusion is that the charges that Burt deliberately falsified his data cannot now be sustained, true though it is that much of what he published in old age was badly presented and perhaps even culpably careless. And the BPS, without explicitly repudiating their 1980 condemnation of all his work, on 24th February 1992 rather grudgingly issued a statement which concluded that “Council considers that it is now inappropriate for the Society as such to seek to express a fresh opinion about whether or not the allegations directed at Burt are true. Moreover, in the light of greater experience, the British Psychological Society no longer has a corporate view on the truth of allegations concerning Burt”.
The conclusion must therefore be that, since these rather disgraceful allegations of fraud and falsification as opposed to mere muddle in old age cannot be sustained, Burt’s reputation as the first and most influential of educational psychologists is now fully restored. And that furthermore selection in education, based upon the genetical heritability of intelligence as first advocated by Sir Cyril Burt, will provide the only satisfactory basis for educational policy in the future, whether or not it may still be regarded as elitist and politically incorrect by some teachers and educational psychologists.