Remembering happy times instead of sticking to negative memories after social exclusion

Previous research shows that difficulty disengaging from negative (self-related) stimuli (i.e. negative self-referential processing; NSP) is a vulnerability factor for depression (Gotlib & Joormann, 2010) and contributes to its recurrence (LeMoult, Kircanski, Prasad, & Gotlib, 2017). The Emotional Reversal Learning Task (ERLT) was designed to investigate this, and we examined its construct validity by inducing social rejection, an etiological process of depression, within the ERLT model. We expected excluded participants to have difficulty disengaging from NSP.

We administered Cyberball to 130 participants randomly assigned to the excluded or included condition. Participants then completed the ERLT: They chose a valence option (positive or negative), retrieved a memory of the same valence, and then were rewarded or punished for their valence choice. For the first phase, retrieving a negative memory was probabilistically rewarded, and this action-outcome contingency was twice reversed during the task. We used Q-learning models to analyze learning rates.

Excluded participants had no more difficulty disengaging from NSP than included participants: Bayesian computational modeling identified no difference between conditions regarding learning that retrieving negative memories was punished. Exploratory analyses found that excluded participants learned the association between retrieving positive memories and reward quicker than included participants, however.

Doubts remain regarding whether participants fully understood action-outcome contingencies, and we did not explicitly check whether participants truly retrieved memories, which could have affected results.

We did not find support for the construct validity of the ERLT when using social exclusion to model depressogenic development within the ERLT.