ABSTRACT
The correlation between two variables is markedly affected by the range of the variables. For example, if we correlated height and weight for a group of persons who ranged from 5 feet, 6 inches, to 5 feet, 8 inches in height, we should find that the correlation is very low, as illustrated in Figure 1. It is, of course, unlikely that we should make such a peculiar selection of persons for the purpose of correlating height and weight. Hoivever, the effect would be similar if, for example, we were to correlate height and weight for pupils in the fifth grade, as compared with correlating height and weight for pupils in grades one to twelve. The correlation between mental age and chronological age will be much greater for a school population than for a given grade.
