Abstract
Recently Saykin et at. (1995), departing from their results in a sample of normal individuals, concluded that, "Demographic factors infrequently account for more than 10% of the variance for many neuropsychological teat scores" (p. 79). In this short article it is pointed out that their results have to be interpreted with extreme caution, because the education and age ranges they used are not valid to analyze the education and age effects on neuropsychological test performance. Educational effect does not represent a linear effect; rather, it represents a kind of negatively accelerated curve, and age effect becomes evident only in a population older than the sample they have selected.