Effects of practice on speed of information processing in children and adults: Age sensitivity and age invariance.

In 2 experiments, children and adults were exposed to 4 different information-processing tasks. Consistent with the global trend hypothesis, age-sensitive linear relations were observed between child and adult latencies, and 10- and 11 -year-olds were approximately 1.7 and 1.6 times slower than 19-year-olds as predicted by R. Kail's (1991) growth function. In Experiment 1, the relation between child and adult latencies did not change over 4 sessions of practice, implying that practice has equivalent effects on corresponding processing steps in children and adults. In both experiments, an age-invariant linear relation between dispersion and central tendency was observed, indicating that children's greater within-subject variability is entirely due to their slower speed of processing. A large number of studies have shown that children perform more slowly than adults on a wide variety of information-p rocessing tasks (e.g., Elliot, 1970; Fairweather & Hutt, 1978; Marmor, 1977; R. J. Sternberg & Nigro, 1980). Recent experimental evidence suggests that these age-related differences in cognitive processing speed are primarily quantitative and not qualitative in nature (Hale, 1990; Kail, 1986b, 1988). Kail has determined the growth functions for the duration of a number of cognitive processing components, and he has shown that these functions are similar both in form and in their rates of decrease with age. These findings suggest that all (or most) basic cognitive processes are maturing at the same rate, a view that Hale (1990) termed the global trend hypothesis. Hale (1990) used an analytic approach that focuses on the whole response latency (i.e., the sum of the component processes) rather than decomposing the latency into separate components. This approach involves plotting the performance of children on multiple tasks as a function of the performance of adults on the same tasks. Using this technique, Hale observed precise linear relations between the latencies of children and young adults (^s > .98), and the slopes of these relations decreased as a function of the age of the child group. A recent meta-analysis conducted by Kail (1991) combined