Table 2.
Dependent measure | Means (SD) |
Univariate effect of internal validity |
||||||
---|---|---|---|---|---|---|---|---|
Valid | No control | Confound | Experimenter bias | F | df | p | η 2 | |
Expert evidence quality | 4.55 (1.07)a | 3.86 (1.11)a | 4.29 (0.90) | 4.13 (1.19) | 4.01 | 3, 239 | 0.02 | 0.07 |
Expert credibility | 5.52 (0.92)b | 5.01 (1.04)b | 5.37 (0.99) | 5.17 (0.96) | 2.95 | 3, 239 | 0.05 | 0.06 |
Plaintiff credibility | 3.89 (0.89) | 3.79 (0.98) | 3.74 (0.98) | 3.88 (1.07) | 0.38 | 3, 239 | 0.50 | 0.03 |
Notes: Differences between means sharing the same superscript within each row were statistically significant at p ≤ .05
Expert evidence quality, expert credibility, plaintiff credibility were all composite variables
η2 = partial η2