I'm looking for links/references to research which have looked into determining the optimum number of choices to include per question on a multiple choice quiz. My question is directed at STEM disciplines.
I am not sure if I understood the subject of our question correctly.
First, what is the meaning of 'Optimum' here? Is it a number within which such a test would best show the students' proficiency, so one can best assess the students' skills?
Second, the optimal number of Multiple Choice (MC) questions will depend on the subject. There is a difference in writing MC tests for humanities, and for mathematical sciences, where often to provide a correct answer one needs to perform some analytical reasoning.
I my math classes I never offer a multiple choice tests.
The review of literature suggests that MCQs with 3 options provide a similar quality of test as that with 4- or 5-option MCQs. It is suggested that MCQs with 3 options should be preferred. Here are interesting documents on the topic.
Backer's answer is interesting as my understanding was that tests with 4 or fewer answers lead to too many statistical anomalies if people guess answers. I think the 3 answer stats are possibly good assuming there is no guessing - so possibly use three if you were to impose negative marking and 5 otherwise?
I am rescinding my previous comment as I was rushing with my answer. Even though I am not using multiple choice tests/quizzes, after consulting with my coworkers, a general answer that people agreed on was 4.