Relative Effectiveness of the Conventional Formula Score

Abstract

An estimate of examinees' true knowledge in English vocabulary was obtained using a multiple-choice test asking them to mark the degree of confidence in each answer they made. A number of formula scores were checked against this estimate. The results suggest that the conventional scoring formula (S = R W ) was found to be in closest agreement with the obtained estimate, while the reliability of the number-right scores was highest. c~ IN 1957, MEAD AND SMITH (2) reported an investiga tion for the empirical verification of the conventional scor ing formula for true-false items, S = R ? W, where S is the corrected score, R is the number of right answers, and W is the number of wrong answers. They administered for that purpose a test in which subjects were requested to indicate on a three-level scale the degree of confidence with which they answered each item. Two years later, Davis (1) used the concept of partial knowledge to obtain an estimate of the subjects' true knowledge from the data obtained by Mead and Smith. In spite of the apparent importance of these types of tests, a review of the literature shows that their use has not been reported since then. It was thus the intent of this investigation to conduct a testing situation similar to that used by Mead and Smith, and to check the relative effectiveness of the conventional formula score using the concept of partial knowledge as defined by Davis.