No document available.
Abstract :
[en] Alzheimer disease (AD) is characterized by an early impairment of explicit memory processes associated to a preservation of implicit memory processes (Fleischman & Gabrieli 1998). Due to the role of explicit memory in the suppression of errors during learning, AD patients tend to reproduce automatically (implicitly) errors that occurred during a previous learning (Baddeley & Wilson, 1994). Consequently, errorless learning should be more efficient than a classical “trial-and-error” procedure for AD patients. Indeed, errorless learning decreases the involvement of (impaired) explicit memory by avoiding the interference caused by the production of errors (Bier et al., 2002). The present study investigates the automatic post-learning error production in mild AD patients and matched control subjects by using a word stem completion task (Adam et al., 2005) in conditions of both errorless and trial-and-error learning. Results showed a lower word stem completion performance in mild AD than control subjects, but a similar performance in the patients’ group for the two learning conditions. Moreover, in the trial-and-error procedure, the errors consisted mainly in erroneous responses already produced during the learning phase. In addition, correlation analyses indicate that the ability to suppress errors in the trial-and-error learning condition in mild AD patients is subtended by the efficiency of episodic memory processes, but not by inhibitory abilities. These results suggest that the errorless procedure improves the quality of learning of mild AD patients (production of fewer errors) but do not influence the learning rate per se.