- Poster presentation
- Open Access
EEG correlates of successful semantic encoding during wakefulness
© Frey et al.; licensee BioMed Central Ltd. 2008
- Published: 17 April 2008
- Left Hemisphere
- Stimulus Onset
- Word Pair
- Verbal Memory
- Brain Potential
Event-related brain potential (ERP) studies have demonstrated that encoding of subsequently recalled items elicit a more positive response than non-recalled items, an effect known as the “difference due to subsequent memory” (Dm-effect) . We studied whether specific ERP characteristics during declarative learning of unrelated word pairs correlate with subsequent successful memory recall.
Thirteen healthy young males (21–28 y) learned 154 unrelated word pairs, which were recalled later. EEG (F3, F4, C3, C4, P3, P4, O1, O2) was continuously recorded during encoding and recall. Stimulus-locked ERPs during the encoding session were averaged separately for successfully recalled and non-recalled word pairs and then subjected to a four way rANOVA with factors ‘condition’ (recall vs. not-recalled), ‘laterality’ (left vs. right), ‘antero-Posterior-site’ (frontal vs. occipital), and ‘time’ (0.2-1.9 s after stimulus onset).
Results disclosed a main effect for the factors ‘condition’, ‘antero-Posterior-site’ and ‘time’ (p < 0.05), and interaction effects for ‘condition’x‘time’ (p<0.05), ‘laterality’x‘time‘ (p < 0.001), ‘antero-Posterior-site’x‘time’ (p < 0.01) and ‘laterality’x‘antero-Posterior-site’x‘time’ (p < 0.01). Post-hoc analyses revealed significantly higher ERPs between 0.35 and 1.1 s after stimulus onset for successfully recalled than non-recalled word pairs. Furthermore, negative correlations were found between mean ERP-activity in the 0.35-1.1-s range in the left hemisphere at learning and the number of subsequently recalled word pairs (r = −0.59 to −0.72; p < 0.05).
Our data indicate a Dm-effect within a wide time window of the ERP in a verbal memory task, and that less positive mean ERP-activity in left cortical areas relates to enhanced success during subsequent semantic recall.
We thank Claudia Renz, Marie-France Dattler, Giovanni Balestrieri for their excellent help in data acquisition. We also thank all volunteers for their compliance. This research was supported by the Swiss National Foundation Grant START 320000-108108/1.
This article is published under license to BioMed Central Ltd.