How Selective True-False Questions Reward Student Recognition

True and False (T/F) questions seldom undergo research in computer science education; hence, little exists to document improvements to T/F questions in CSE. We analyze the effects of Selective True and False (STF) questions. In STF, the student answers a subset of the question pool, rather than answering all of it. Students choose questions they believe they can answer, and avoid questions which they deem difficult. Therefore, an intelligent selection of questions leads to a lower probability ofguessing. This rewards students with knowledge (as it allows them to recognize easier/harder problems) while remaining indifferent to students with total ignorance. In other words, STF lessens the probability of a weaker student achieving a higher grade than a stronger student via guessing. Our data suggests that stronger students are able to recognize and answer questions with success (achieve a score in the top half of the class). Explicitly, the number of times a question is selected versus student success among students which selected that question has a correlation of 0.73. Furthermore, examining the success of the student's particular set of selected questions rather than the success of the student themselves, one finds an even stronger correlation versus the number of times a question is selected, i.e.\ a correlation of 0.99.