An Argument Opposing Jensen on Test Bias: The Psychological Aspects
Summaries of Jensen’s own work in test bias, as well as an exceptionally thorough review of the literature on this topic, were gathered together in his book, Bias in Mental Testing (1980a). In this work he identified an argument which he called the ‘Egalitarian Fallacy’: ‘An unbiased test should reveal reliable differences between individuals, but it should not show differences between the average scores of different racial or social groups in the population or between the sexes’ (1981, p. 129). This argument might be simplified as follows: ‘The groups are not different in the abilities being measured. Hence, if score differences between groups occur, the tests must be biased.’ Jensen’s problem with this argument lies, of course, with the premise that the groups are not different. Most serious scholars would agree that genuine differences between many groups do exist, although causes for the differences are a matter of considerable controversy and are discussed elsewhere in this volume. In the American culture it would be surprising indeed if obvious differences between racial and ethnic groups in economic advantage and opportunity for learning and advancement had no impact at all on the development of mental abilities.