Most students of psychology are likely to be familiar already with these terms, but for the sake of readers of this book who are specialists in one of the Arts rather than psychologists, I offer a brief and simple explanation of these terms which are occasionally used in the text.
Suppose we have two orders of merit, e.g. one in an intelligence test and one in a teat of drawing ability, made on 100 boys. Now we may be able to see at a glance that some very intelligent boys do badly in the drawing test. But it may be impossible to tell oven by careful inspection, whether the order in intelligence resembles that of yet another test, in say, pitch discrimination, more or less than it resembled the order in drawing.
Statisticians, however, have devised methods by which we can indicate by a figure (called the coefficient of correlation) how far such orders resemble one another. We may then find that the resemblance between intelligence and drawing is expressed by a correlation of 0·4, while that between intelligence and pitch discrimination works out at 0·25, showing clearly that for these pairs of orders the resemblance is definitely less for intelligence and pitch discrimination.
There is no need for the general reader to be able to calculate such correlations.1 But he should seek to build up an idea as to what is a high correlation and what a low, and how much resemblance they indicate. If two orders are exactly the same the correlation works out at +1. If two orders are exactly the opposite to one another____________________