This article was previously published in our January 2013 Newsletter
by Doug Smith, CEO
An often-asked question is, “We tell the students they need to get an 80% to pass and yet the test is cut at a 62%. How do we resolve this discrepancy?”
We spoke last week about how a cut score is determined for a test. As a reminder, it is the average of the cut score of each individual item on the test. Each individual item’s cut score is based on the probability that the student is not just guessing and is based on the number of good wrong answers. As not all questions are cut the same, not all tests have the same cut score. However, the published passing score of an 80% does remain constant. So what we do is attempt to equate your published passing score with the tests actual cut score.
Think of this as an example why. Imagine you are expecting any participant to run a mile in 6 minutes. That is your published cut and you might think this is consistent. Yet, if one course as your run up a steep grade for the full mile, and another course has you run down a low grade for a mile, clearly these courses are not consistent so the published cut of 6 minutes would not be consistent.
What we do is create a consistent, and therefore reliable, measuring process. The cut score of the test is determined based on the method described above. The published passing score is based on some decision made by the program. Achieve each means you pass so we equate the cut score with the published passing score. So using the 62% cut example and the 80% published passing score, if a student receives a raw score of a 62% we report to them that they passed with an 80%. On the other hand, if a student receives a raw score on this test of less than a 62% we report the student receives a properly pro-rated amount less than 80%. Similarly, if the receive a raw score of greater than 62% we report a pro-rated amount greater than an 80%.
Another way to provide for consistency would be to continually modify your tests until you achieved a cut score of 80% to equal the published passing score. This process is very time consuming and may result in some objectives not being assessed in order to achieve your goal. If you have selected the raw score method of testing, this should be the method you should use to create your tests in order to create a fair and reliable testing process.