People are generally skilled at using a confidence scale to rate the strength of their memories over a wide range. Specifically, low-confidence recognition decisions are often associated with close-to-chance accuracy, whereas high-confidence recognition decisions can be associated with close-to-perfect accuracy. However, using a 20-point rating scale, the authors found that the ability to scale memory strength had its limitations in that a high proportion of list items received the highest rating of 20. Efforts to induce participants to differentiate between these strong memories using emphatic instructions and alternative scales were not successful. Remember/know judgments indicated that these strong and hard-to-scale memories were often based on familiarity (not just recollection). Providing error feedback on a plurals discrimination task finally produced a high-confidence criterion shift. The authors suggest that the ability to scale strong (and almost perfectly accurate) memories may be limited because of the absence of differential error feedback for very strong memories in the past (the kind of differential error feedback that may account for the memory-scaling expertise that participants otherwise exhibit).Keywords: recognition memory, confidence and accuracy, signal-detection theory, feedbackMemories vary in strength, and people are generally quite adept at using a numerical confidence scale to indicate how strong different memories are. This ability is most apparent in studies of recognition memory, in which accuracy is typically strongly related to confidence (Ratcliff & Murdock, 1976). Mickes, Wixted, and Wais (2007) showed that the strong relationship between confidence and accuracy extends over a wider range than had been previously investigated. In that study, participants completed a standard recognition memory test during which targets and lures were presented one at a time, and participants were asked to rate each item using a 20-point scale. A rating of 1 indicated 100% certainty that the item was new, and a rating of 20 indicated 100% certainty that the item was old. Intermediate ratings indicated varying degrees of lesser certainty, with ratings of 10 and 11 indicating complete uncertainty that the item was new (10) or old (11). As shown in Figure 1, participants were very accurate when confidence was high (i.e., for ratings of 1 or 20), and accuracy declined in continuous fashion to chance levels as confidence decreased to complete uncertainty (toward the middle of the scale). Thus, as a general rule, participants are able to use a confidence scale to provide valid ratings of the strength of their memories. The fact that participants exhibit this expertise without being trained by the experimenter is a key consideration for the memory scaling issue that is the main focus of this article. Figure 2a illustrates one interpretation of these results in terms of the standard unequal-variance signal-detection model. According to this account, items vary in memory strength, with the mean and varianc...