Table 3.
Reliability of Attention Bias Variability (ABV) index as a function of outlier handling method
|
Study 1 (N=29; 320 trials)
Adult Social Anxiety Disorder |
Study 2 (N=15; 160 trials)
Adult Social Anxiety Disorder |
Study 3 (N=28; 48 trials)
Pediatric Healthy Volunteers |
|||||||
|---|---|---|---|---|---|---|---|---|---|
|
| |||||||||
| Method/measure | ICC (single measure) |
ICC/alpha (average of all measures) |
p | ICC (single measure) |
ICC/alpha (average of all measures) |
p | ICC (single measure) |
ICC/alpha (average of all measures) |
p |
|
| |||||||||
| Rescale (Windsorize) RTs | |||||||||
| Attention Bias Variability Index | .39 | .88 | <.001 | .14 | .57 | .006 | .15 | .47 | .01 |
|
| |||||||||
| Delete RTs <300 & >2500 & ±3 SDs from individual’s session mean |
|||||||||
| Attention Bias Variability Index | .63 | .95 | <.001 | .13 | .54 | .02 | .21 | .57 | .001 |
|
| |||||||||
| Delete RTs <300 & >3000 & ±2 SDs from individual’s session mean |
|||||||||
| Attention Bias Variability Index | .65 | .96 | <.001 | .22 | .69 | .001 | .30 | .68 | <.001 |
Note: RTs=reaction times. Significant ICC’s obtained in every analysis. Method yielding the highest overall reliability for each study is highlighted in pink.