Tests of the neural noise hypothesis of age-related cognitive change.

Three predictions were derived from the assumption that with increased age there is a decrease in the effective signal-to-noise ratio of neural representations of visually presented stimuli. Although the results from manipulations designed to examine the internal consistency of the predictions were quite positive, the predicted age differences failed to appear in two of the three dependent measures. Because the same pattern was found in two independent studies and estimates of measurement reliability and statistical power were moderately high, it was concluded that at least some versions of the neural-noise hypothesis of cognitive changes with age may be untenable and that more effort should be devoted to devising experimentally testable implications of the hypothesis.