Boerendonk, Marieke (2025) Modelling Affective Forecasts: Comparing the Kalman Filter to the Multilevel Autoregressive Model in an Ecological Momentary Assessment Study. Bachelor thesis, Psychology.
|
Text
BachelorThesisMariekeBoerendonkS5233224.pdf Download (1MB) | Preview |
Abstract
Making the wrong decision can be very upsetting, but could you have anticipated how you would later feel about this decision? The current study investigated affective forecasting accuracy of two different statistical models. More specifically, the predictive accuracy of the multilevel autoregressive model was compared to that of the Kalman filter. Participants (N = 30) rated and predicted their affect for 14 days in an ecological momentary assessment study, and provided point and interval predictions at the current time, three hours ahead, and six hours ahead. Consequently, one-step predictions (three hours ahead) were computed by the multilevel autoregressive model and the Kalman filter. We hypothesized that, compared to the Kalman filter, the multilevel autoregressive model will perform better due to its extensive use in affect dynamics and its potential to capture the development of emotions. Moreover, the multilevel autoregressive model was expected to specifically outperform the Kalman filter for negative affect (NA), as emotional inertia is suggested to be more strongly related to NA than positive affect (PA). Linear mixed models showed a significant difference between the statistical models in NA, where the multilevel autoregressive model outperformed the Kalman filter. However, no significant difference was found for PA. These findings are important when we consider the biases that humans are subjected to in making affective forecasts. This study might serve as a foundation for bridging statistical models with real-world affective forecasting, as further research could investigate how combining human judgments with statistical forecasts could enhance the overall accuracy. Keywords: affective forecasting, EMA, multilevel AR model, Kalman filter
| Item Type: | Thesis (Bachelor) |
|---|---|
| Supervisor name: | Petersen, F.J. |
| Degree programme: | Psychology |
| Differentiation route: | None [Bachelor Psychology] |
| Date Deposited: | 23 Jul 2025 08:20 |
| Last Modified: | 23 Jul 2025 08:20 |
| URI: | http://gmwpublic.studenttheses.ub.rug.nl/id/eprint/5679 |
Actions (login required)
 |
View Item |