The Oral Presentation That Ran Out of Time
Lena has a 10-minute oral presentation on her investigation into whether retrieval practice improves retention of chemistry formulae, followed by 5 minutes of questions. Her slides contain 14 bullet points per slide across 18 slides, including three dense data tables, and she plans to read them. At the 4-minute mark of a practice run she is still on slide 6 explaining her school's timetable structure. Her results - a within-subjects comparison of practice-test scores versus rereading scores across 40 students, with a paired analysis - appear on slides 13–15. In the question period a moderator asks why she used a within-subjects design and whether order effects were controlled. Lena answers that within-subjects is 'more scientific' and that all participants did retrieval practice first because 'it was easier to organise'.
1. Diagnose Lena's time-management problem and prescribe a restructured 10-minute plan.
Read the worked answer
Diagnosis. She is spending roughly 4 minutes on context that the audience does not need (timetable structure) and will reach her results with about 3 minutes left, forcing her to rush the section that carries the most marks. Reading dense bullets also slows delivery and prevents eye contact, and 18 slides in 10 minutes is unworkable at any reasonable pace.
Restructured plan.
- 0:00–1:00 - Question, why it matters, one-sentence finding preview.
- 1:00–3:00 - Design in brief: 40 students, within-subjects, retrieval vs rereading, what was measured and when. Cut timetable detail entirely.
- 3:00–4:00 - Method essentials only: materials, counterbalancing (or its absence), analysis approach.
- 4:00–6:30 - Results: one clean figure plus one summary table; state the effect, its size, and variability.
- 6:30–8:30 - Interpretation, limitations (order effects front and centre), implications.
- 8:30–10:00 - Conclusion and one slide on what she would do next.
Reduce to about 10 slides, no more than 5 short bullets each, and never read the slide.
2. Evaluate Lena's answers to the moderator's two questions and write a stronger answer she could have given.
Read the worked answer
Evaluation. 'More scientific' is not a justification; within-subjects designs have specific advantages (controls individual differences, greater power with small n) and specific costs (order and carryover effects). The answer shows she does not understand the trade-off. The second answer is worse: running retrieval first for every participant means practice, fatigue, and familiarity with the test format all favour whichever condition comes second, so the comparison is confounded. 'Easier to organise' is an administrative convenience that undermines the inference.
Stronger answer. 'I chose within-subjects because individual differences in chemistry aptitude are large relative to the expected effect, and a within-subjects design removes that variance, giving adequate power with 40 students. The cost is order effects. In my design, however, all participants completed retrieval practice before rereading, so practice and fatigue are not balanced across conditions - this is a genuine limitation and means the advantage I observed may be partly order. A better design would counterbalance the order across participants, or include a washout period between conditions. I would report the effect as provisional pending that control.'
3. Lena has three dense data tables on slides 13–15. Advise her on presenting this evidence for an oral audience and on handling a likely follow-up question about the effect size.
Read the worked answer
Presentation. Oral audiences cannot read a dense table while listening. She should reduce to one summary table with condition means, standard deviations, and the test statistic, plus one figure - for example a paired dot or bar chart showing each participant's two scores or the mean difference with a confidence interval. The full tables belong in the written report and can be offered as a handout or referenced verbally. Every number on a slide should be there to make one point she states aloud.
Effect size follow-up. She should be ready to say: 'The mean difference was X points (SD = Y), corresponding to a Cohen's d of Z, which is a [small/moderate/large] effect by conventional benchmarks; the 95% confidence interval was [a, b].' She should also note that the effect size is estimated under the confound of fixed order, so it may overstate the true benefit of retrieval practice. Being able to state the effect size, its interval, and its caveat converts a weak moment into evidence of understanding.