J. Abedi et J. Bruno, CONCURRENT VALIDITY OF THE INFORMATION REFERENCED TESTING FORMAT USING MCW-APM SCORING METHODS, Journal of computer-based instruction, 20(1), 1993, pp. 21-25
The literature has reported many problems associated with the conventi
onal forced-choice (R-W) test scoring system. The impact of guessing o
n assessment is among the most serious problems cited in the literatur
e. A Modified Confidence Weighted-Admissible Probability Measurement (
MCW-APM) scoring systems attempts to eliminate the impact of guessing
on assessment by allowing the measurement of partial information. Sinc
e this procedure is amenable to optical scoring it is especially suita
ble for use in computer-based instruction measurement and mass testing
by school organizations. The research literature, however, strongly s
uggests that two dimensional, confidence weighted scoring systems sacr
ifice test validity for test reliability. This study empirically exami
nes a reproducible, two dimensional, test scoring procedure (MCW-APM)
and finds that concurrent validity is enhanced - especially for diffic
ult items or when tests are used with low and medium attaining student
s. The findings of this study strongly suggest that the reproducible n
ature of a scoring system used on a test might impact in a positive ma
nner on both the validity and reliability of a test.