Response latencies measured on IQ tests: dimensionality of speed indices and the relationship between speed and level

It has been suggested that correlations observed between psychometric intelligence and simple reaction-time (RT) tasks support mental speed models of intelligence. It has also been argued that when simple and complex RT tasks are contrasted, RT on complex tasks offers better estimates of intelligence (complexity hypothesis). Some points on the continuum of task complexity seem to have been neglected in the research on mental speed and cognitive ability, however. In particular, latencies measured on individually administered IQ tests have been little studied. The present study provides evidence for a relatively strong correlation between psychometric intelligence and latencies measured on IQ-test items in a sample of 131 15- to 17-year-old adolescents. In contrast with results obtained for groups of children, no age difference in latencies was found. This is consistent with more general studies of the evolution of speed in childhood and adolescence [Kail, R. (1991). Psychological Bulletin, 109, 490]. Also contrary to results obtained with children, the speed indices clustered into distinct speed factors. The central role of task complexity in assessing speed, development, and intelligence is discussed, as well as the advantage of adopting a clear operational definition of speed and level when assessing intellectual performance.