Hi all,
I am working on making adjustments to scores on a test. There were two sections on the test: (1) Multiple Choice (MC), and (2) Short Answer (SA). There were 16 MC questions, and each was worth 3 points. There were 6 SA questions, and each was worth 8 points. Thus, each part of the exam was worth 48 points (the final scores were computed as percentages, since the points didn't add up to 100). Let's call this Weighting Scheme 1.
Some people showed that they understood the material very well, as evidenced by their impressive answers on the SA part of the test. However, many of these folks did very poorly on the MC part of the test. I was concerned that the MC portion of the test was simply too tricky, or poorly-designed. Thus, I recomputed the exam scores, this time counting the MC questions as being worth 1 point each, and leaving the SA questions as being worth 8 points each. This new weighting system led most people's percentage scores to go up, and it felt fairer to me. Let's call this Weighting Scheme 2.
However, there were a few people who had higher percentage scores with the Weighting Scheme 1. This indicates that, unlike most people, they seemed to have more of a knack for the MC questions than the SA questions. I am thinking about doing the inverse of Weighting Scheme 2 for this group. However, my limited math abilities are preventing me from knowing exactly how to proceed. I am wondering if anyone has any thoughts on this.
Hopefully I've explained this clearly, and it all makes sense. Any thoughts are very appreciated!!
Best,
Craig
I am working on making adjustments to scores on a test. There were two sections on the test: (1) Multiple Choice (MC), and (2) Short Answer (SA). There were 16 MC questions, and each was worth 3 points. There were 6 SA questions, and each was worth 8 points. Thus, each part of the exam was worth 48 points (the final scores were computed as percentages, since the points didn't add up to 100). Let's call this Weighting Scheme 1.
Some people showed that they understood the material very well, as evidenced by their impressive answers on the SA part of the test. However, many of these folks did very poorly on the MC part of the test. I was concerned that the MC portion of the test was simply too tricky, or poorly-designed. Thus, I recomputed the exam scores, this time counting the MC questions as being worth 1 point each, and leaving the SA questions as being worth 8 points each. This new weighting system led most people's percentage scores to go up, and it felt fairer to me. Let's call this Weighting Scheme 2.
However, there were a few people who had higher percentage scores with the Weighting Scheme 1. This indicates that, unlike most people, they seemed to have more of a knack for the MC questions than the SA questions. I am thinking about doing the inverse of Weighting Scheme 2 for this group. However, my limited math abilities are preventing me from knowing exactly how to proceed. I am wondering if anyone has any thoughts on this.
Hopefully I've explained this clearly, and it all makes sense. Any thoughts are very appreciated!!
Best,
Craig
