Twenty-minute tutoring slots often drift into note-reading because it is low-friction and feels safe; the cause is that rereading gives an illusion of fluency without testing retrieval. Duffspin fixes that by turning the slot into a timed loop: prompt, pause, pair-check, rewrite. The tutor gets visible evidence of what the student can recall before any explanation is offered.
A common failure is a prompt bank that is too broad; broad questions hide the exact gap and waste time on vague answers. Keep prompts narrow enough to answer in 20–30 seconds, then mix definitions, labelled diagrams, and short “why” questions. For a model of how the routine is introduced before a session, the duffspin guide is a useful reference.
The problem in many drills is that the tutor answers too early; the cause is silence feeling unproductive. Duffspin makes silence part of the method. Run each round for four minutes.
Repeat with a new prompt. The rewrite is the point: it converts feedback into a second retrieval attempt, which is more useful than a nod of agreement.
Without a written close, the session can end on the tutor’s summary rather than the student’s recall. Choose the two prompts the student missed most often and ask for written answers with no notes. Use those answers to set homework: one similar application question and one explanation question. That keeps the next session focused on the actual gap, not a general topic.
If the student freezes, the cause is usually a prompt that combines too many steps. Split it into two smaller prompts and keep the 20-second pause. If accuracy is high but speed is low, keep the content and shorten the pause to 15 seconds. The routine stays the same; only the prompt size and timer change.