Handle incorrect completions by giving full credit for correct or as-good-as-correct answers and part credit for answers that give sense but are not strictly correct¶

Strategy · All strategies
Evidence · no claims cited

Description¶

The article suggests two methods of handling accuracy in scoring: give no credit for what is not correct, or give full credit for correct or as-good-as-correct answers and "part credit for what gives sense but is not strictly correct." The authors believe the second is fairer, especially where papers contain many errors, since part-credit proportion can be assigned by individual judgment of accuracy.

Design Implications¶

Context¶

Requirements¶

• Credit in neither case is given for elisions left unfilled or filled so as to have no connection with the rest of the thought.

Constraints¶

• For the study's own papers there were so few errors that the arbitrary-deduction method was adopted; the authors note that with some papers results would be very different and the second method should then be selected.

Target Learners¶

• students taking completion tests

Target Learning Goals¶

• fair scoring of partially correct completion-test responses

Related Strategies¶

• Score sentence tests by summing each correctly translated sentence's scale value rather than raw number correct

Examples¶

-

Key Sources¶

• Walter R. Miles and Julian E. Butterworth. (1912). A Tentative Standardization of a Completion Test. The Journal of Educational Psychology, pp. 329-336. https://www.google.com/search?q=%22A+Tentative+Standardization+of+a+Completion+Test%22