AtlasSession 2L6: Hypothesis Testing and Statistical Testsp.59–60
Interpreting results
- 6 earlier topics
- Alpha error
- Tests, p-values, significance
- CIs and p-values
- Beta error and power
- Interpreting results
Fig. 1
Constant–Murley score at 12 months, collagen-patch-augmented minus standard rotator cuff repair (points)
Fig. 1. A number line with “no difference” at 0 and “worth it” at 10.00 points. The likely range 2.00 to 14.00 points: real, but maybe not worth it.
- likely range
- 2.00 to 14.00 points
- Worth it from
- 10.00 points
- Real? Worth it?
- real, but maybe not worth it
Two-Curve Test Visualiser · First look
Level 1 · First look · Lecture 6 · p.59–60
Read a result in plain words: is there a difference, and is it big enough to matter?
Interpret a result: reject H0 only when p is below the pre-set threshold and the groups differ only in the intervention, read the CI against the MCID, and prefer parametric tests for normal data.
Two questions, not one (the step you are on)
'is it real?' and 'is it worth it?', like a sale that saves you 1p.
Two questions, not one: is the gain real, and is it worth it? A sale that saves you 1p is real but not worth having. Drag the gain: a likely range of 2.00 to 14.00 points reads real, but maybe not worth it.