Dokument: The Role of Situation Models in the Continued Influence Effect
| Titel: | The Role of Situation Models in the Continued Influence Effect | |||||||
| URL für Lesezeichen: | https://docserv.uni-duesseldorf.de/servlets/DocumentServlet?id=74222 | |||||||
| URN (NBN): | urn:nbn:de:hbz:061-20261116-115136-4 | |||||||
| Kollektion: | Dissertationen | |||||||
| Sprache: | Englisch | |||||||
| Dokumententyp: | Wissenschaftliche Abschlussarbeiten » Dissertation | |||||||
| Medientyp: | Text | |||||||
| Autor: | Hey, Carolin [Autor] | |||||||
| Dateien: |
| |||||||
| Beitragende: | Prof. Dr. Bayen, Ute J. [Gutachter] PD. Dr. Raoul Bell [Gutachter] | |||||||
| Dewey Dezimal-Klassifikation: | 100 Philosophie und Psychologie » 150 Psychologie | |||||||
| Beschreibung: | Information that is later proven to be false can continue to influence memory, beliefs, and inferences even after it has been retracted. This phenomenon is known as the Continued Influence Effect (CIE; e.g., Walter & Tukachinsky, 2020; Wilkes & Leatherbarrow, 1988). Despite extensive research demonstrating its robustness, there is ongoing debate about the underlying cognitive mechanisms. Two prominent accounts propose different cognitive processes to underly the effect (e.g., Ecker et al., 2022; Lewandowsky et al., 2012; Sanderson & Ecker, 2020): The retrieval-failure account attributes the CIE to failures during memory retrieval, whereas the model-updating account locates its causes in encoding processes, specifically in the construction and updating of situation models. Critically, direct empirical evidence for the model-updating account has been lacking.
The present dissertation addressed this gap by pursuing three objectives. First, I developed and validated a method for assessing situation models within a standard CIE paradigm. Second, I identified source credibility as a manipulation that affects situation-model construction and updating as well as the CIE, enabling experimental investigation of their relationship. Third, I used this manipulation to test predictions derived from the model-updating account against those of the retrieval-failure account. In Experiment 1, I developed a question-based assessment of participants’ situation models. The assessment was validated by experimentally manipulating agreement with the original information and demonstrating that it reliably differentiated between levels of agreement and predicted subsequent CIE responses. Experiments 2.1 and 2.2 established source credibility as a suitable manipulation to test the model-updating account. Informant credibility and retractor credibility were manipulated orthogonally. Across both experiments, they exerted independent and antagonistic effects on the CIE: High informant credibility led to a stronger CIE, whereas high retractor credibility led to a weaker CIE. Importantly, these effects did not interact, indicating that credibility was applied selectively to the individual critical statements. This pattern is consistent with the assumption that source credibility can independently influence the integration of the original information into the situation model and its persistence therein following the retraction. In Experiments 3.1 and 3.2, I used credibility-disclosure timing to disentangle encoding and retrieval processes and to test the assumption that the CIE arises during encoding. In Experiment 3.1, source credibility affected the CIE regardless of credibility disclosure timing, suggesting a possible influence at retrieval. However, this pattern may reflect retrospective integration of credibility information or short-term activation at immediate retrieval. Experiment 3.2 addressed these issues by separating credibility disclosure and situation-model construction and introducing a delay before the CIE test. Under these conditions, source credibility affected the CIE only when it was available during encoding; when disclosed afterward, it had no effect on either the situation model or the CIE. Finally, I used structural equation modeling to directly test the relationships between situation models and the CIE. The results showed that the situation model held after retraction consistently predicted the CIE and that source-credibility effects were mediated through the situation models when present. Taken together, the findings provide direct empirical support for the model-updating account of the CIE. They demonstrate that the effect arises from the way information is integrated into and maintained within situation models, rather than from failures during retrieval. At the same time, some findings suggest that retrieval processes may still contribute under certain conditions. Overall, this dissertation advances the understanding of how misinformation continues to influence cognition despite retraction and clarifies the mechanisms underlying the CIE. | |||||||
| Lizenz: | ![]() Dieses Werk ist lizenziert unter einer Creative Commons Namensnennung 4.0 International Lizenz | |||||||
| Fachbereich / Einrichtung: | Mathematisch- Naturwissenschaftliche Fakultät » WE Psychologie » Mathematische und Kognitive Psychologie | |||||||
| Dokument erstellt am: | 16.11.2026 | |||||||
| Dateien geändert am: | 16.11.2026 | |||||||
| Promotionsantrag am: | 26.05.2026 | |||||||
| Datum der Promotion: | 29.06.2026 |

