Computational modeling of pessimistic future views in individuals with depressive symptoms: dynamics in affective forecasts
Final Report Abstract
Affective forecasting—the ability to simulate how one will feel in the future—is a basic psychological function that guides daily life decision-making. However, human affective forecasts are known to be biased and inaccurate in general. Importantly, systematic biases have been found among individuals with depressive symptoms – that is, those individuals tend to predict high levels of negative affect for a given future moment (i.e., negative affective forecasts), and such a negative forecast seems to persist over time and to constitute a pessimistic view of the future and hopelessness. This project investigated how a negative affective forecast persists – and how a negative forecast is not updated even when a positive event is anticipated. Specifically, the objectives of the current project were: (a) to model the updating process for affective forecasts in both laboratory (WP 1) and daily-life settings (WP 2) using a computational approach and (b) to systematically test individual differences in these forecasting processes linked to depressive symptoms (WP 1 and 2). WP 1 used the film-clip task, where participants (a community sample) repeatedly rated their momentary levels of positive and negative affect while watching emotional film clips. We found that, in general, people tend to assume that the current affective state would remain unchanged for at least the next few minutes even in the presence of time-varying emotional stimuli. This present-focused forecasting prevents individuals from learning from their forecasting errors, as they are not entirely likely to consider what they forecasted in the past when making a new forecast. These forecasting errors were, however, not associated with depressive symptoms. Therefore, the present-focused forecasting style is a general heuristic that does not directly inform depressive cognition. WP 2 used the experience sampling method (ESM) to investigate the dynamics in affective forecasting in daily life. We were particularly interested in how individuals with depressive symptoms maintain negative forecasts even though they have opportunities to update their forecasts, for example, when pleasant events or activities are anticipated. We asked individuals (a community sample) to rate the current levels of negative affect 10 times each day for seven days. At the same time, participants were asked to predict the levels of affect that they would experience for the next few hours and to report the activities that they were most and least looking forward to. Results replicated the present-focused forecasting style and showed that people with higher levels of depressive symptoms make more negative affective forecasts. Furthermore, we found that individuals with depressive symptoms tend to underestimate the impact of an anticipated pleasant activity on improving negative affect. This underestimation was also predictive of depressive symptoms at the six-month follow-up. These findings suggest that individuals with depressive symptoms maintain a belief that negative affect continues regardless of anticipated pleasant activities, and this belief may function as a vulnerability factor for depression. A "surprise” in the project outcome was that the participants’ present-focused forecasting style was suboptimal, given that the participants made larger errors than did statistically optimal forecasts based on a Kalman filter. This means that statistical forecasts offer comparatively good performance and may therefore open-up new clinical implementations, such as modifying pessimistic future views by providing vulnerable individuals with online, machine-based affective forecasts.
Publications
-
Effects of Induced Mindfulness at Night on Repetitive Negative Thinking: Ecological Momentary Assessment Study. JMIR Mental Health, 10 (2023, 7, 19), e44365.
Sommerhoff, Amanda; Ehring, Thomas & Takano, Keisuke
