AtlasSession 4L10: Sample size calculationsp.3–5, 17–18
Why sample size matters
- 6 earlier topics
- Alpha error
- Tests, p-values, significance
- CIs and p-values
- Beta error and power
- Why sample size matters
- Next: Continuous outcomes
Fig. 1
Collagen–PRP vs standard dressing for diabetic foot ulcers: difference in wound-area reduction at 4 weeks (points), one bar per trial
Fig. 1. Virtual trials with 10 per group and a true difference of 10 points. None run yet.
- Trials so far
- 0
- Showed the real difference
- 0 of 0 (no trials yet: press Play or +1)
- Missed it
- 0 (no trials yet: press Play or +1)
- Patients needed
- 20
Sampling Simulator · First look
Level 1 · First look · Lecture 10 · p.3–5, 17–18
Explain in plain words why a study needs the right number of participants, not too few and not too many.
Explain what a sample size calculation guarantees (false positive and false negative errors at or below the desired level), why it is done at the design stage, and the cost of too few or too many participants.
Tasting soup (the step you are on)
One spoonful can mislead, a few more settle it, and nobody needs to drink the pot.
Like tasting soup, each bar is one spoonful: a small trial with 10 patients per group. One can mislead, so here are 0; nobody needs to drink the whole pot.