Poster Session A, Saturday, March 23, 1:30 pm - 3:30 pm, Pacific Concourse
An electroencephalographic investigation of trial-by-trial updating of knowledge structures
Franziska Richter1; 1Leiden University
Knowledge structures such as memory schemas efficiently store large amounts of information without the cost of memorizing every detail of previous experiences. The usefulness of schemas, however, critically depends on their adaptability: how flexibly a schema can be updated according to changing environmental conditions. Yet, how schemas may be updated when conflicting information is encountered is not well understood, as it is difficult to track the dynamic updating of knowledge structures with traditional memory measures. Here, combining a continuous-report paradigm with electroencephalogram (EEG) recordings, I investigate the trial-by-trial relationship between event-related potentials (P300), schema-based prediction errors (PEs), and their effect on schema-updating. Twenty-five participants completed a 3-day experiment: on day 1 they learned object-location associations by making predictions and receiving feedback about object locations on a circular dial. This way participants formed location-schemas for four object categories (animals, clothes, food, furniture). On day 2, participants studied new objects, but unbeknownst to them the location schema for two of the categories changed to a new, inconsistent location. On day 3, participants’ memory for all items was tested. EEG was recorded on day 2, while participants received feedback on objects that were either consistent or inconsistent with the pre-established schema. Behaviourally, PE at day 2 predicted schema updating on day 3. Higher P300 potentials were observed in the inconsistent compared to the consistent condition and PE size correlated with P300 amplitude. The results demonstrate that memory schemas, similar to more short-lived belief structures, are updated via prediction-based learning.
Topic Area: LONG-TERM MEMORY: Episodic