AtlasSession 3L9: Variations in study designp.25–26
Cross-over
- 3 earlier topics
- Types of distribution
- Normal vs non-normal
- Choosing a statistical test
- Parallel and multi-arm trials
- Cross-over
Fig. 1
Transtibial amputees, silicone vs gel socket liner in random order: peak interface pressure (kPa) after 2 weeks on each
Fig. 1. 24 people, 12 per order, try both silicone liner and gel liner; now in period 2: the groups have swapped, AB on gel liner, BA on silicone liner; each person's two readings are joined.
- Stage
- swapped
- People
- 24
- Readings per person
- 2
Trial Design Board · First look
Level 1 · First look · Lecture 9 · p.25–26
See why letting each person try both options sharpens the comparison.
Describe a cross-over trial and explain why comparing each person with themselves reduces the sample size needed.
Try both shoes (the step you are on)
Each runner tests both brands, so differences between runners' feet cancel out.
Like runners who each test both brands of shoe, so differences between their feet cancel out, each of these 24 amputees will try both liners.