2016 ASEE Annual Conference & Exposition

Modern online learning materials may include built-in questions that are used for some of a class' homework points. To encourage learning, question solutions may be easily available to students. We sought to determine to what extent students earnestly attempt to answer questions when solutions are available via a simple button click. An earnest attempt means to try answering a question at least once before viewing the solution. We analyzed data from 550 students in four classes, at a four-year public research university, a four-year public teaching college, and two community colleges. We found average earnestness was a rather high 84%. We also found that 89% of students earnestly attempted 60%-100% of questions, with 73% earnestly attempting 80%-100%. Only 1% of students blatantly "cheat the system" by earnestly less than 20% of questions. Thus, the heartening conclusion is that students will take advantage of a well-designed learning opportunity rather than just quickly earning points.

We noted that earnestness decreased as a course progressed, with analyses indicating the decrease being mostly due to tiredness or some other student factor, rather than increasing difficulty. We also found that analyzing per-question earnestness can help question authors find poorly-formulated questions.

In addition to providing results of our earnestness analysis, the paper also describes the style by which the questions are made -- how they complement text, teach rather than test, take only a small amount of time each, always include explanations, strive to occasionally "trick" students to explicitly dispel common misconceptions, avoid drill/kill and instead each teach a unique concept, create questions where alternative right answers are less likely to be marked wrong, and more. Via such design, students seem to discover that the questions are worth their time and effort, and thus most students earnestly try. We also discuss processes that can hurt earnestness, such as assigning excessive work. We describe our philosophy of not limiting student attempts for such learning questions, to create a safe learning environment (whereas other activities may indeed benefit from limits).