Table 6.
Properties | Methodological Frameworks | ||||
---|---|---|---|---|---|
Rank-o-grams | Scatter plota | Rank heat plot | Radar plot | Spie charts | |
Summary type | |||||
Numerical | ✗ | ✗ | ✗ | ✓ | ✓ |
Does not depend on ordering of outcomes | – | – | – | ✗ | ✓ |
Visual | ✓ | ✓ | ✓ | ✓ | ✓ |
Single figure | ✗ | ✗ | ✓ | ✗ | ✗ |
Outcome measures | |||||
Same units | |||||
Absolute effects | ✗ | ✓ | ✓ | ✓ | ✓ |
P (best) | ✓ | ✓ | ✓ | ✓ | ✓ |
P-score | ✗ | ✓ | ✓ | ✓ | ✓ |
SUCRA | ✗ | ✓ | ✓ | ✓ | ✓ |
All ranking probabilities | ✓ | ✗ | ✗ | ✗ | ✗ |
Mean or median rank | ✗ | ✓ | ✓ | ✓ | ✓ |
Different units | |||||
Relative effects | ✗ | ✓ | ✗ | ✗ | ✗ |
Outcome preferences | |||||
Straightforward ability to weight outcomes | ✗ | ✗ | ✗ | ✗ | ✓ |
Other considerations | |||||
Can include > 2 outcomes | ✓ | ✗ | ✓ | ✓ | ✓ |
Can accommodate missing outcomes | ✓ | ✗ | ✓ | ✗ | ✗ |
aScatter plots of two outcomes [35]