Abstract
The author, Chairman of the Sociology Department in the University of California at Santa Barbara, cites the dangers of applying significance tests to a single correlation when the goal is to generalize in time and area.
Get full access to this article
View all access options for this article.