Measuring forgetting: a critical review of accelerated long-term forgetting studies.
Bottom Line: The memory difficulties associated with ALF can however cause considerable distress to patients.Studies of ALF in patients with epilepsy have so far demonstrated mixed results, which may reflect differences in methodology.These issues can also help to explain some of the mixed findings in studies of ALF and inform the design of standardised tests for assessing ALF in clinical practice.
Affiliation: Clinical Neuropsychology Services, Royal Hallamshire Hospital, Sheffield, UK; Department of Psychology, University of Sheffield, Sheffield, UK.Show MeSH
Related in: MedlinePlus
Mentions: Learning to criterion involves repeatedly presenting material until a criterion (e.g., 100% accuracy on two successive trials) is reached. Bell (2006) argued that this method of matching learning poses the risk of the material being over-learnt; leading to the possibility that early forgetting is masked by ceiling effects. Overlearning is the continued learning of stimuli beyond the criterion of one perfect trial (Krueger, 1929). In his early study of overlearning, Krueger (1929) gave participants either just enough trials to recall a word-list flawlessly, or twice this number of trials (i.e., 100% overlearning; see Fig. 1). When tested after 1-day and 27-day delays, participants in the overlearning condition forgot fewer words than those in the normal learning condition. This reduction of forgetting rates caused by overlearning has been identified across a range of studies but evidence suggests that it is short-lived (Driskell, Willis, & Copper, 1992) and may have greatest effect on forgetting over the first 24 h after learning with less effect over delays of 2–28 days (Driskell et al., 1992). It follows that forgetting studies which use paradigms prone to overlearning may underestimate forgetting over delays up to 24 h. Given that criterion levels are often set at a level which exceeds perfect recall on one trial, overlearning is indeed a risk inherent to this approach. In these cases, forgetting over long delays may also have been apparent over shorter, 30-min delays, yet this was obscured by overlearning. A simple alternative is to set the criterion to a level lower than 100%. A recent study used a criterion of 80% on a word-list, which matched groups without ceiling effects (Muhlert et al., 2010).
Affiliation: Clinical Neuropsychology Services, Royal Hallamshire Hospital, Sheffield, UK; Department of Psychology, University of Sheffield, Sheffield, UK.