What is the Angoff method?
The Angoff method sets a pass line by judging the questions, not by ranking the people who wrote them. A panel of subject experts pictures a candidate who is just good enough to enter the profession. For each question, each panellist estimates the chance that this candidate gets it right. Average those estimates and you have the score that candidate should earn, which becomes the basis of the pass line.
The method takes its name from William Angoff, whose 1971 chapter in Educational Measurement is the work ExAC's own technical report cites. It is widely used on licensing and certification exams. Canada's physician assistant certification exam, for one, uses a modified form of it.
The minimally competent candidate is the idea that makes it work. This is neither the average candidate nor a star. It is the person who has reached the minimum level of competence needed to practise, and no more. Panellists are not asking whether a question is hard in general. They are asking how this one person would do on it.
A worked example
Here is the arithmetic on a five-question exam with three panellists. The questions and numbers are invented to show the method. They do not come from any ExAC paper or panel.
| Question type | Panellists A, B, C | Average |
| Recall of a definition | 80%, 85%, 75% | 80% |
| One-step code lookup | 65%, 60%, 70% | 65% |
| Envelope detail with a flaw to spot | 70%, 60%, 65% | 65% |
| Contract scenario, two plausible answers | 45%, 55%, 50% | 50% |
| Multi-step cost calculation | 40%, 50%, 45% | 45% |
| Panel's estimate for this paper | Average of the five | 61% |
Now suppose next year's paper is harder, and the panel rates the same five slots at 70, 55, 55, 40 and 35 percent. The estimate drops to 51 percent. That is the point of the method: a harder paper gets a lower line, an easier paper gets a higher one, and the standard expected of the candidate stays where it was.