Table 5.
Results of moderator analyses for nominal scaled variables.
| Variable | ρ | SDρ | 95% CI | V% | K | N |
|---|---|---|---|---|---|---|
| IMPLICIT FORMAT | ||||||
| Paper and Pencil | 0.136 | 0.189 | (0.070, 0.202) | 32.15 | 39 | 4439 |
| PC | 0.081 | 0.140 | – | 50.16 | 4 | 312 |
| Oral | 0.248 | 0 | – | 100 | 4 | 268 |
| EXPLICIT CONSTRUCT | ||||||
| Motives | 0.135 | 0.175 | (0.077, 0.193) | 35.24 | 48 | 5364 |
| Goals | 0.094 | 0 | (0.002, 0.186) | 100 | 9 | 910 |
| Wishesa | – | – | – | – | 1 | 85 |
| ORDER OF ADMINISTRATION | ||||||
| Implicit measure first | 0.149 | 0.187 | (0.084, 0.214) | 32.38 | 41 | 4198 |
| Explicit measure first | 0.086 | 0.027 | (−0.074, 0.246) | 95.85 | 7 | 798 |
| Varying order | 0.025 | 0 | – | 100 | 2 | 180 |
| INFORMATION SOURCE | ||||||
| Book/chapter | 0.283 | 0 | – | – | 1 | 78 |
| Dissertation | 0.043 | 0 | (−0.055, 0.141) | 100 | 9 | 785 |
| Journal | 0.144 | 0.182 | (0.033, 0.255) | 33.32 | 41 | 4758 |
| Unpublished | 0.138 | 0.071 | – | 77.08 | 3 | 263 |
Because there were no non-overlapping confidence intervals for any variable, we refrained from inserting subscripts indicating non-overlapping categories; ρ, mean population correlation; SDρ, estimated standard deviation of population correlations; CI, confidence interval (not computed for categories with less than five study correlations); V%, percentage variance accounted for by sampling error; K, number of independent study correlations; N, total sample size.
a
Computation of the population mean was not possible due to a weight of zero, the study correlation of the only study found was r = −0.006.