Table 3.
Results of the mixed ANOVAs taking treatment and occasion as inter-groups and within-group factors.
| Statement | Treatment |
Occasion |
Interaction |
dchange (95%CI) | |||
|---|---|---|---|---|---|---|---|
| F | p | F | p | F | p | ||
| (1) Treatment logic | 1.22 | 0.276 | 2.81 | 0.101 | 0.14 | 0.714 | −0.01 (−0.62, 0.60) |
| (2) Treatment satisfaction | 0.18 | 0.675 | 9.75 | 0.003 | 0.79 | 0.378 | −0.05 (−0.68, 0.58) |
| (3) Recommending to others | 0.63 | 0.432 | 15.27 | <0.001 | 0.15 | 0.704 | 0.19 (−0.47, 0.85) |
| (4) Usefulness for other disorders | 1.26 | 0.267 | 4.15 | 0.048 | 0.63 | 0.430 | 0.31 (−0.31, 0.93) |
| (5) Usefulness for the patient | 0.01 | 0.925 | 4.72 | 0.035 | 0.14 | 0.713 | 0.29 (−0.33, 0.91) |
| (6) Unpleasantness∗ | 0.03 | 0.865 | 4.10 | 0.049 | 0.18 | 0.672 | −0.11 (−0.72, 0.50) |
| Total score | 0.31 | 0.583 | 13.68 | 0.001 | 0.06 | 0.809 | 0.26 (−0.40, 0.92) |
F = F statistic for testing the significance of the treatment, the occasion, and the treatment × occasion interaction; all F-tests had 1 and 42 degrees of freedom. p = probability level associated to the F-test. dchange = standardized mean change between the two treatments (positive dchange values indicated a larger increase from expectations to satisfaction in EMMA than in TAU). 95%CI: 95% confidence limits for dchange. ∗Item 6 is answered in reverse, and so it has been recoded to follow the same scoring criteria as the other five items on the scale.