In the standard scoring procedure for multiple-choice exams, students must choose exactly one response as correct. Often students may be unable to identify the correct response, but can determine that some of the options are incorrect. This partial knowledge is not captured in the standard scoring format. The Coombs elimination procedure is an alternate scoring procedure designed to capture partial knowledge. This paper presents the results of a semester-long experiment where both scoring procedures were compared on four exams in an undergraduate macroeconomics course. Statistical analysis suggests that the Coombs procedure is a viable alternative to the standard scoring procedure. Implications for classroom instruction and future research are also presented.