The effect of contingency test instruction on locus of control and person reliability

Abstract

Recently, the topic of person reliability, the consistency of a test taker, has begun to draw attention from people in the field of educational and psychological measurement. Several different methods have been proposed to produce person reliability indices (Sato, 1975; Harnish & Linn, 1981; Levine & Rubin, 1979; Ayabe & Heim, 1987). Researchers have applied person reliability indices to identify students with unique learning characteristics (Sato, 1975), to discriminate, among students with different instructional backgrounds (Harnish & Linn, 1981), and to detect aberrant patterns of standard-setting judgments (Jaeger, 1988). More recently, Shishido (1991) applied person reliability to a college Japanese language course placement examination to ferret out students who might intentionally miss (sandbag) some of the test items. Most of the studies on person reliability have focused on the introduction of new person reliability indices and their applications. A study associating person reliability with dispositional characteristics might be able to shed light on our understanding and application of person reliability. The purpose of this study is, first, to relate person reliability to locus of control, a selected dispositional characteristic, and second, to examine the effect of the test instructions on locus of control and person reliability.