Four hours later, the participants returned for a final session in which they were told about the trickery. They were then asked again to provide "memory" and "belief" ratings for the different actions. The take-home finding is that for 25 per cent of the fake actions, the participants now reported significantly stronger memory scores than belief scores - in other words, their (false) memory of having performed the fake actions persisted even though they often no longer believed they'd performed the actions.
Clark and his team said that their findings raised ethical questions about memory research: "To the extent that debriefing might not always completely 'undo' the effects of suggestive manipulation, we might question the ethics of inducing false memories in experimental participants. Is it ethical for participants to leave research labs with remnants of non-believed false memory content in the forefront of their minds?"