Implicit learning in memory rehabilitation: a meta-analysis on errorless learning and vanishing cues methods.

The objective of this study was to present a quantitative review on the treatment effects of memory rehabilitation techniques based on intact implicit learning capacity in amnesic patients, that is, errorless learning and the method of vanishing cues. English-language journal articles focusing on these rehabilitation techniques were examined using MedLine (1966-2002) or PsychInfo (1887-2002), as well as additional papers listed in the references of these articles. Studies had to meet the following inclusion criteria: (1) original data were reported, (2) memory rehabilitation was studied in memory-impaired patients, (3) a control intervention was included, (4) exact scores were listed for both intervention conditions, or the exact statistics were presented. Studies were classified on the basis of the to-be-learned material and the method of intervention (errorless learning, vanishing cues, control intervention), patient characteristics were determined, and the tasks that were used were taken into account. Effect sizes and variances were computed for each individual study compared to control treatment using within-group statistics. A "large" and statistically significant ES was found for errorless learning treatment, but no significant ES was demonstrated for the vanishing cues method. The results of the present study show that the errorless learning technique is effective in amnesic patients. The effects on the vanishing cues method are only small (and nonsignificant).