Citation

Sturges HA. Am. J. Sociol. 1927; 33(2): 269-275.

Copyright

(Copyright © 1927, University of Chicago Press)

DOI

10.1086/214386

PMID

unavailable

Abstract

Of the many possible use of the methods of correlation in connection with experimental investigations of attitude change, the present report suggests four: the use of correlation methods (1) to indicate the extent to which changeability is a general or specific characteristics of personality; (2) to test materials used in attitude tests in order to select those which give the most valid and consistent results; (3) to ascertain the degree of reliance which may be placed on attitude test scores from persons voting only once; and (4) to measure the diagnostic and prognostic and prognostic value of preliminary tests in connection with courses of study.