[2] Does retrieving episodic information influence the updating of emotional context?

Background

Episodic memory is your ability to recall specific episodes (including words or pictures) or events from your past. Prior research indicates that the ability to retrieve specific, episodic information from memory is integral to daily functioning. This research project will explore how retrieval of a single piece of episodic information (in this case, positive and negative words) influences access to other positive and negative episodic information. We will also explore whether this process varies with current mood.

Research Questions / Hypotheses

An interaction is expected between trial valence and depression scores in predicting the proportion of correct trials on the ERA task. Specifically, a stronger (negative) association will be observed between depression scores and accuracy on positive trials, compared to the association between depression scores and accuracy on negative trials.

Participants

211 participants completed the study. Those under 18 were excluded or with insufficient English language proficiency.

Methods

Participants were taken to the Qualtrics survey via the REP platform. Informed consent was provided before completing the task sequence, which comprised all 60 ERA trials, the Beck Depression Inventory-II (BDI-II), and the ERA task recall block. The task battery took approximately 45 minutes to complete.

Results

Hypothesis 1—correlation. Depression scores will be negatively associated with the proportion of correct responses for positively valanced trials on the ERA task. A bivariate correlation will be run between BDI score (0 to 63) and the proportion correct on positively valanced ERA trials. Our hypothesis will be supported if the p-value is <.05 and the correlation is in a negative direction.

Hypothesis 2—interaction. An interaction is expected between trial valence and depression scores in predicting the proportion of correct trials on the ERA task. Specifically, a stronger association will be observed between depression scores and accuracy on positive trials, compared to the association between depression scores and accuracy on negative trials. A linear mixed effects model will be used to examine hypothesis 2. The model structure is as follows: Proportion correct ~ BDI * Trial Valence + ESL + (1/Subject ID) Fixed Effects = BDI (between-subjects), Trial Valence (within-subjects; excluding neutral trials) Random Effect = Subject ID Covariate = ESL (English Second Language) Hypothesis 2 will be supported if the interaction term between BDI and Trial Valence (positive and negative) is significant in a manner compatible with the description provided. We intend to reproduce these outcomes through separate and combined analyses with data from study 1.a if the ERA trials need to be adapted on the basis of valence ratings.

Implications

This study was a pre-registered replication of our originally piloted study in 2023. Our findings were not replicated, however, some exploratory analyses have revealed interested trends which we may follow-up in further studies. We plan to publish this research, alongside our original pilot data. This work may also be presented at a conference as part of my PhD.