Authors: JOHN J. WILD¹* and JONATHAN M. WILD²

Abstract

Learner feedback shapes how courses, curricula, and instructional technologies are designed, yet the value of that feedback rests on an untested premise. The premise is that student advice tracks the evidence of what works. We test this premise in a large course, linking 523 mid-term survey responses on course-design preferences to each learner’s complete courseware records and outcomes, and measuring how strongly each graded activity tracks examination performance across 1,659 learner-term records. Three findings emerge. First, learners hold definite views. The same broad pattern appears in every administration and both formats. Second, misalignment is concentrated. Learners credit passive review videos with learning far more often than retrieval-based quizzes, at 85 versus 60 percent, although quiz scores align far more closely with examination results. Learners who want fewer examinations outnumber those who want more by four to one. Yet learners do not reject retrieval itself. Only 9 of 315 open-text comments explicitly ask to cut quizzes or examinations. The dominant requests are for more worked examples and more time. Third, alignment is tied to performance. Learners who say quizzes help them learn score 7 points higher on examinations than those who do not, and a simple evidence-alignment index predicts examination performance after controlling for coursework. The results identify which parts of the learner voice carry information for course design and which parts need checking against evidence. Consumer-style responsiveness to student demand would harm the weakest students most, so evidence-tested listening is itself a public interest concern.

Keywords: Learner Preferences; Course Design; Retrieval Practice; Desirable Difficulties; Student Voice; Public Interest; Accounting Education.

Download PDF