Notebook AtlasSURG0002 · Research methods: go to the course map

AtlasSession 3L9: Variations in study designp.3–6, 10, 17–18, 38–39

Parallel and multi-arm trials

  1. 2 earlier topics
  2. Distribution of data
  3. Types of distribution
  4. Normal vs non-normal
  5. Choosing a statistical test
  6. Parallel and multi-arm trials
  7. Next: Factorial trials and interactions

Fig. 1

Rat critical-size femoral defect, bone volume (mm³) at 8 weeks: rats needed per design

Fig. 1. One study with 3 groups of 150 sharing one empty defect group: 450 rats in all.

Rats in all
450
Groups
3
Rats per group
150

Trial Design Board · First look

Design design
3

Level 1 · First look · Lecture 9 · p.3–6, 10, 17–18, 38–39

See how sharing one control group tests several treatments with fewer animals.

Compare two-arm and multi-arm parallel trials on total sample size, and decide whether a multiple-testing adjustment is needed using the p.6 algorithm.

  1. Cooking contest (the step you are on)

    Three new recipes judged against one shared standard dish, not three separate contests.

    Like new recipes judged against one shared standard dish, not in separate contests, two new treatments here share one empty-defect group: 3 groups, 450 rats in all.

step 1/3