Preserved pattern completion priming for novel, abstract geometric shapes in amnesics of several aetiologies

Two experiments were conducted using a paradigm developed by Gabrieli et al., Neuropsychologia 28, 417-427, 1990, which assessed both indirect and direct memory performance in a completion task for novel abstract geometric patterns. The preferred method of scoring was the lines method, based on the number of correct and incorrect lines produced for each item. It was chosen because it is both the simplest and the most informative measure. Two methods of scoring were used in previous work, namely, the strict whole figure method and the lenient whole figure method (Gabrielli et al., 1990; Verfaelie et al., Brain Cognit. 18, 34-45, 1992). Therefore to facilitate comparisons between studies and to determine the characteristics of different scoring methods, results with all three measures were included. In Experiment 1, two different encoding strategies of naming and copying were used in order to explore the relationship between indirect and direct memory performance. Indirect memory performance in the naming condition was at baseline whereas in the copying condition it was significantly above baseline. Cued recall did not differ across encoding conditions but recognition was higher in the naming condition than the copying condition. In Experiment 2, an attempt was made to extend the findings of two studies, one with H.M. (Gabrielli et al., 1990) and one with nine Korsakoffs (Verfaelie et al., 1992), to a larger group of 14 amnesics of several aetiologies. Indirect memory performance was found to be equivalent for the amnesics and their matched controls, only when the lenient and the lines methods of scoring were used. Recognition and cued recall performance was impaired for the amnesics compared to the controls.