Research ReportResisting false recognition: An ERP study of lure discrimination

- ERPs were used to investigate true and false recollection of pictures.
- A discrimination task emphasized explicit identification of similar lures.
- Parietal old/new effects suggest recollection of recognized not rejected lures.
- Late old/new effects in better performers suggest post-retrieval processing.
- Data support diagnostic rather than disqualifying monitoring.

There is keen interest in what enables rememberers to differentiate true from false memories and which strategies are likely to be the most effective. This study measured electrical brain activity while healthy young adults performed a mnemonic discrimination task, deciding whether color pictures had been studied, were similar to studied pictures (lures), or were new. Between 500 and 800 ms post-stimulus, event-related potentials (ERPs) for correctly recognized studied pictures and falsely recognized lures compared to those for correctly rejected novel items had a left centroparietal scalp distribution. This was typical of the parietal old/new effect associated with recollection, and in line with previous evidence that similar lures may elicit false or phantom recollection as opposed to just familiarity. There was no evidence of a parietal effect for correctly rejected lures as would be expected if recall-to-reject was used. The ERP old/new effects for lures also varied with individual differences in performance. Parietal effects for falsely recognized lures were larger in better performers, who successfully rejected a greater number of lures as “similar”. The better performers also showed more pronounced right frontocentral old/new effects between 800 and 1100 ms for correctly rejected and falsely recognized similar lures. The enhancement of false recollection in better performers implies false recognition of lures occurred only when more specific information was recovered about the study episodic. Together, the findings suggest reliance on recollection to decide that items were studied, supported by post-retrieval processing.