Table 1. Potential for the two- and three-condition experimental design to detect relationships between placebo and experimental conditions, and true baseline values.
| Likely outcome |
2 condition design |
3 condition design |
| (a) Exp > Pla |
Yes |
Yes |
| (b) Exp = Pla |
Yes |
Yes |
| (c) Pla > Exp |
Yes |
Yes |
| (d) Exp = Pla = Bas |
No |
Yes |
| (e) Exp > (Pla = Bas) |
No |
Yes |
| (f) (Exp = Pla) > Bas |
No |
Yes |
| (g) Exp > Pla > Bas |
No |
Yes |
| (h) Pla > (Exp = Bas) |
No |
Yes |
| (i) Pla > Exp > Bas |
No |
Yes |
| (j) Exp > (Bas > Pla) |
No |
Yes |
Abreviations: Bas = Baseline level; Pla = Placebo condition; Exp = experimental condition. The symbol ‘=’ indicates no significant differences between; the symbol ‘>’ indicates that the condition to the left was associated with significantly better performance than the condition to the right.
