Premium
Invalidity of indirect and direct measures of attitude toward cheating
Author(s) -
Freeman Linton C.,
Ataöv Türröz
Publication year - 1960
Publication title -
journal of personality
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.082
H-Index - 144
eISSN - 1467-6494
pISSN - 0022-3506
DOI - 10.1111/j.1467-6494.1960.tb01631.x
Subject(s) - cheating , citation , psychology , philosophy , social psychology , library science , computer science
Tests for the indirect assessment of attitudes have gained some currency in the literature of social psychology. They have been used primarily as indexes to behavior in situations where social pressures have been assumed to inhibit truthful response to direct questions. In such situations investigators have frequently assumed that re sponses to indirect questions provide more valid behavioral indexes than responses to direct questions. A few studies of the correlation between direct and indirect questions have been made (Campbell, 1953; Parish & Campbell, 1953 ; Proshansky, 1943; Rankin & Camp bell, 1955; Robinson & Rohde, 1946; and Sanford & Rosenstock, 1952), but although Campbell has questioned the valiity of indirect questions as indexes to behavior, no systematic study of this ques tion has been reported. The present study was designed in an at tempt to examine the validity of several types of indirect questions and a direct question against the criterion of overt behavior. PROCEDURE In attempting to test the validity of indirect questions it was considered desirable to select an area of behavior in which social pressures are known to operate, yet one which would be amenable to direct observation. Cheating on examinations by students is such an area; it was chosen for this study. 5's were a group of 38 freshmen and sophomores enrolled in introductory sociology at Syracuse University. Actual cheating on an examination was measured by a device similar to the one suggested by Hartshorne and May (1928). Fifteen ambiguous ques tions were introduced into the students' regular mid-term examination. Five of these were true-false, five were fill-in, and five were multiple choice. After the tests were administered, responses to these fifteen questions were recorded and the least popular answers were chosen as "correct." A key was prepared for the whole examination, and at the next class period the 5s were asked to score their own tests. The instructor was called from the room and a student was drafted to read the "correct" answers. The papers were col lected and the presence of a changed answer constituted an instance of actual cheating. ^ The second author is now at Orta Dogu Teknik Universitesi, Ankara, Turkey.