Liepelt, Fabienne-Renée (2025) Affective Forecasting Accuracy Across Different Interval Lengths and Affective Valences. Bachelor thesis, Psychology.
|
Text
Bachelor-Thesis-F.-Liepelt-S5104491.pdf Download (803kB) | Preview |
A thesis is an aptitude test for students. The approval of the thesis is proof that the student has sufficient research and reporting skills to graduate but does not guarantee the quality of the research and the results of the research as such, and the thesis is therefore not necessarily suitable to be used as an academic source to refer to. If you would like to know more about the research discussed in this thesis and any publications based on it, to which you could refer, please contact the supervisor mentioned.
Abstract
Background: Affective forecasting, predicting future emotional states, significantly influences decision-making. However, humans tend to be inaccurate in forecasting their future affect because they overestimate the duration and intensity of their future emotions (impact bias). The impact of emotional valence and different interval lengths on affective forecasting accuracy, as well as differences between human and statistical model forecasts have been researched separately leaving unclear how these variables might interact. Aims: This study investigated whether human affective forecasting accuracy decreases with longer time intervals, whether this effect varies by emotional valence, and how human accuracy compares to Kalman filter predictions across 3-hour and 6-hour intervals. Methods: We conducted a 14-day experience sampling study in which participants responded to five daily prompts measuring current and predicted positive and negative affect. Absolute prediction errors were analyzed via multilevel models, comparing human and model forecasts. Results: Human prediction accuracy decreases when the interval size increases. However, this effect is independent of the emotional valence, contradicting previous research. Moreover, there was no significant difference between the Kalman filter and human forecasting accuracy. Conclusions: While human affective forecasts are more accurate for the near than the distant future, which applies equally to predicting positive and negative affect, the Kalman filter does not perform significantly differently than humans. However, differences in operationalization of interval size as differences in event specificity (specific event vs daily life forecasts) may limit the generalizability of our findings which could be a starting point for future research. Keywords: affective forecast, experience sampling, statistical model, interval size
| Item Type: | Thesis (Bachelor) |
|---|---|
| Supervisor name: | Petersen, F.J. |
| Degree programme: | Psychology |
| Differentiation route/Track: | None [Bachelor Psychology] |
| Date Deposited: | 23 Jul 2025 07:34 |
| Last Modified: | 23 Jul 2025 07:34 |
| URI: | http://gmwpublic.studenttheses.ub.rug.nl/id/eprint/5656 |
Actions (login required)
 |
View Item |