Table 8.
Numbers of participants required for various designs when d = .4, .5, and .6 and the data are analyzed with Bayesian statistics (BF > 10). The numbers of the d = .4 column are the default numbers to use. The higher values of d require dependent variables with a reliability of .8 at least. Therefore, authors using these estimates must present evidence about the reliability of their variables. This can easily be done by calculating the ICC1 and ICC2 values discussed above.
| Bayesian analysis (BF > 10) | |||
|---|---|---|---|
| d = .4 | d = .5 | d = .6 | |
| 1 variable between-groups | |||
| • 2 levels | 380 | 240 | 170 |
| • 2 levels, null hypothesis | 2400 | 2400 | 2400 |
| • 3 levels (I = II > III) | 690 | 450 | 300 |
| • 3 levels (I > II > III) | 2850 | 1800 | 1200 |
| 1 variable within-groups | |||
| • 2 levels | 100 | 65 | 45 |
| • 2 levels, null hypothesis | 720 | 720 | 720 |
| • 3 levels (I = II > III) | 125 | 80 | 55 |
| • 3 levels (I > II > III) | 540 | 340 | 240 |
| Correlation | 370 | 230 | 160 |
| 2 × 2 repeated measures | |||
| • Main effect one variable | 52 | 32 | 23 |
| • Interaction (d v. 0) | 210 | 130 | 85 |
| 2 × 2 split-plot | |||
| • Main effect between | |||
| ◦ r = .5 | 290 | 190 | 130 |
| ◦ r = .9 | 360 | 220 | 160 |
| • Main effect repeated-measure | 100 | 66 | 46 |
| • Interaction (d v. 0) | 390 | 250 | 170 |
| • ANCOVA | |||
| ◦ rrep_measure = .5 | 300 | 190 | 130 |
| ◦ rrep_measure = .9 | 380 | 230 | 170 |