Processing speed, attention, and intelligence: Effects of spatial attention on decision time in high and low IQ subjects

Mental-speed theories of intelligence (Spearman, The abilities of man, 1927; Lemmon, Archives of Psychology, 15, 5–38, 1927–1928; Jensen, Journal of Social and Biological Structures, 3, 103–122, 1980; Jensen, Speed of information processing and intelligence, 1987) are challenged by the finding that, in some methods, bright subjects do not appear to process information at an increased speed. The present experiment was designed to explore new RT measures which might relate more strongly to IQ while possessing a clear theoretical relationship to biological theories of intelligence. One direct measure of information processing speed is the rate at which choice reaction time (RT) increases with increasing stimulus information content (RT slope). While RT slope is a critical variable in mental-speed models of intelligence, the results of several experiments, since Jensen proposed his model, suggest that slope does not correlate strongly with intelligence. We investigated the possibility that spatial attention affects the relationship between RT and IQ in the traditional Jensen paradigm. Thirty-five subjects performed an RT task with two and four choice stimuli presented in narrow and wide spatial configurations. Correlations between Ravenʹs Advanced Progressive Matrices (APM) scores and RT were in line with previous reports of a moderate negative relationship between IQ and RT. However, a much stronger effect was found in the relative speed in the narrow and wide spatial display conditions. The wide minus narrow RT difference in the two-choice condition correlated 0.71 with APM scores. This result suggests that the processing speed advantage of high IQ subjects is best revealed under optimal attentional conditions. It is suggested that this result supports the speed of processing model of intelligence. Paradoxically, the traditional Jensen paradigm confounds the high choice orders with increased spatial attention demands, thus artifactually increasing the RT slope in high, but not in low, IQ subjects. Explicit control of spatial attention requirements indicates that individual differences in intelligence are substantially underpinned by differences in the speed of information processing mechanisms which, in high IQ subjects, require focused attention if they are to be deployed adequately.