Does answering 55% of the questions in an exam automatically make you pass? No… in fact, passing or failing a test depends on the cut-off score applicable to the test in question. The cut-off score is the boundary between a pass and a fail. But how is the cut-off score determined and which (test) aspects influence this cut-off score?
We distinguish four pillars for the determination of the cut-off score:
The cut-off score is based on an absolute cut-off score.
The absolute cut-off score is a cut-off score that is determined in advance. This cut-off score is at a fixed point, independent of the scores of the group of students. The cut-off score determined afterwards is the relative cut-off score. The relative cut-off score takes into account the scores of the group. The absolute cut-off score is the most commonly used. The relative cut-off score is useful when you do not know how a test ‘will do’, for example, with a new course or test form.
In a test with closed questions, there is always a chance that a student will get the question right, even if he does not know the answer. This is called the guessing chance. The size of the guessing chance depends on the number of answer options. Suppose a test has a total of 60 questions with 1 point each and 4 answer options. The guessing probability is then 25% (1 in 4). This means that a student can obtain 15 points (¼ of 60), by guessing.
When determining the cut-off score, you usually take the height of the guessing chance into account. In practice, this means that the cut-off score is higher because the guessing chance is taken into account. In this case, the cut-off score is calculated as follows.
You can determine the cut-off score in a number of ways, while taking into account the difficulty level of the questions. A method that is often used is that of Angoff. In this method, different experts check all questions on their degree of difficulty. They assess what percentage of the candidates who would just manage to pass would get the question right. The percentages of the different experts are then added up and the average of these is the cut-off score.
A group of five teachers take the following steps.
Average percentages (extremes excluded):
Added up and averaged: 60% cut-off score
If a test is important for the practice of the profession and/or concerns a final examination, you use a high cut-off score. This also applies if the work involves a high risk of damage. The cut-off score for a medical exam is probably higher than for many other exams.
Of course, the opposite is also true. When a (formative) test is intended to measure progress that does not count towards the final result, you can lower the cut-off score.
You can also adjust the cut-off score on the basis of an expected pass rate. For example, if you expect that a test is very easy for the target group because all candidates are highly educated and the test only contains simple insight questions. In that case the expected pass rate is very high and you can choose to raise the cut-off score. If you expect the opposite, that the test will be very difficult for most candidates, you can lower the cut-off score.
Think carefully when determining a cut-off score. A fixed ‘percentage of questions correct’ or ‘number of points obtained’ is not always a good indicator. There are several things that determine whether the cut-off score is higher or lower. Coordinate this matter with colleagues and record the decisions made. Want to know more about this topic, please contact us or follow Optimum Assessment on LinkedIn: