Table 2.
Group Differences on Task Performance Measuresa
| Measure | All Participants, Mean (SD)
|
t Value (df = 141) | P Value | No MDD or Anxiety Disorder, Mean (SD)
|
t Value (df = 57) | P Value | Lifetime MDD or Anxiety Disorder, Mean (SD)
|
t Value (df = 82) | P Value | |||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| LR (n = 60) | HR (n = 83) | LR Participants (n = 37) | HR Participants (n = 22) | LR Participants (n = 23) | HR Participants (n = 61) | |||||||
| RT, ms | ||||||||||||
|
| ||||||||||||
| Incongruent | 674.5 (6.4) | 664.8 (5.1) | −1.2 | .24 | 672.5 (9.1) | 640.8 (11.8) | −2.13 | .04 | 669.0 (9.3) | 668.5 (5.6) | −0.04 | .96 |
|
| ||||||||||||
| Congruent | 494.8 (6.4) | 492.0 (5.1) | −0.88 | .38 | 498.5 (9.1) | 482.4 (11.8) | −1.09 | .28 | 480.5 (9.3) | 490.7 (5.6) | 0.96 | .34 |
|
| ||||||||||||
| Interference RT effectb | 179.7 (3.0) | 172.8 (2.4) | NA | .07 | 174.0 (3.7) | 158.4 (4.5) | NA | .01 | 188.5 (4.8) | 177.8 (2.9) | NA | .23 |
|
| ||||||||||||
| Accuracy, % | ||||||||||||
|
| ||||||||||||
| Incongruent | 85.9 (0.48) | 88.6 (0.4) | 4.45 | <.001 | 85.6 (0.6) | 86.5 (0.8) | 0.9 | .36 | 85.2 (0.74) | 89.4 (0.45) | 4.9 | <.001 |
|
| ||||||||||||
| Congruent | 100.0 (0.48) | 99.2 (0.4) | −1.36 | .18 | 99.3 (0.6) | 98.5 (0.8) | −0.9 | .37 | 99.9 (0.74) | 99.4 (0.45) | −0.56 | .58 |
|
| ||||||||||||
| Interference accuracy effectc | 14.1 (0.65) | 10.5 (0.5) | NA | <.001 | 13.7 (0.89) | 12.0 (1.0) | NA | .80 | 14.7 (1.0) | 10.0 (0.6) | NA | .08 |
Abbreviations: HR, high-risk; LR, low-risk; MDD, major depressive disorder; NA, not available; RT, reaction time.
Values are calculated using the least squares mean values of the mixed model: RT (or accuracy) = age + diagnosis + sex + congruence + run + run × congruence + diagnosis × congruence + diagnosis × run × congruence over 10 runs. The reduced RT interference effects were derived from those in the HR group with no lifetime illness (ie, those who were resilient), whereas the improved accuracy in responding within the HR group were derived from those HR participants who had a lifetime history of illness. Some t values are not available because, in the mixed model, we used the least squares mean values for the diagnosis × congruence term to obtain the t value when congruence was fixed. The P values for interference RT or accuracy effects are simply the P values for the diagnosis × congruence term, and no t values are available.
The mean RT for incongruent trials minus the mean RT for congruent trials.
The accuracy during incongruent trials minus the accuracy during congruent trials, where accuracy is defined as percentage correct trial responses.