The first time I ever thought about test formats causing minor grade inflation was in a college economics class. I remember speaking to the professor about how guesswork within a multiple-choice exam likely produced higher scores. A student, I recall saying, would likely score much lower if the questions were presented in an open response format.
Looking back, I see that I was wasting my professor’s time (as most of my comments did). He had a well-earned reputation for being a tough grader that I see in hindsight was mostly because all of his tests were open response. It was very common to get an exam back from him with quarter-point deductions scattered across an essay like sprinkles on a doughnut. A response that likely would have been ‘correct’ in a multiple-choice format usually got 85 to 90% of full marks in the essay format.
My guess is I thought of this angle because I was simultaneously studying statistics. One day, I was probably sitting around thinking about probabilities when I realized that the math suggested multiple-choice exams inflated grades. Interestingly enough, I remember most of my statistics exams being multiple-choice. I suppose this explains why even though I was far better at economics than I was at statistics, my grades from those classes were generally the same.
