Academic journal article Memory & Cognition

Illusions of Competence during Study Can Be Remedied by Manipulations That Enhance Learners' Sensitivity to Retrieval Conditions at Test

Academic journal article Memory & Cognition

Illusions of Competence during Study Can Be Remedied by Manipulations That Enhance Learners' Sensitivity to Retrieval Conditions at Test

Article excerpt

Monitoring one's knowledge during study is susceptible to a foresight bias (Koriat & Bjork, 2005). Judgments of learning (JOLs) are inflated whenever information that is present at study and absent, but solicited, at test, such as the targets in cue-target paired associates, highlights aspects of cues that are less apparent when those cues are presented alone. The present findings demonstrate that foresight bias can be alleviated by study-test experience (Experiment 1), particularly test experience (Experiments 2 and 3), and by delaying JOLs after study (Experiment 4) and that both foresight bias and its alleviation have behavioral consequences, as measured by study time allocation (Experiment 5). Collectively, the findings suggest that overconfidence and misallocation of study time arise from a mismatch that is inherent to education-that the answer is present at study and absent at test-and that alleviating the problem requires creating conditions at study that sensitize learners to retrieval conditions at test.

Recently, we (Koriat & Bjork, 2005) described an illusion of competence that arises from what we termed a foresight bias. Such a bias, in our view, derives from an inherent discrepancy between the standard conditions of learning and the standard conditions of testing, a difference that seems innocuous but is the source of distorted judgments and overconfidence. On a typical memory test, people are presented with a question (e.g., "What is the capital of Belgium?") and are asked to produce the answer. In contrast, in the corresponding learning condition, both the question and the answer generally appear in conjunction (e.g., "The capital of Belgium is Brussels"), meaning that the assessment of one's future memory performance occurs in the presence of the answer. This difference has the potential of creating a perspective bias. That is, the learner, who needs to adopt the perspective of the examinee, may find it difficult to detach himself or herself from the perspective of the learner, because doing so requires discounting what he/she now knows. The failure to discount potential answers during learning results in a foresight bias-that is, unduly high predictions of one's future performance.

Several discussions in the literature have used the phrase curse of knowledge to capture the difficulty in discounting one's privileged knowledge and experience in judging what a more ignorant other knows or should know (Birch & Bloom, 2003; Camerer, Loewenstein, & Weber, 1989; Keysar & Henly, 2002; see Pronin, Puccio, & Ross, 2002, for a review). Participants, for example, tend to overestimate the effectiveness of their communications, assuming, more than is warranted, that the recipients will understand the intentions behind the message (Keysar & Henly, 2002; Newton, 1990). We argue that a similar curse of knowledge might underlie the unduly high expectations that students often hold about their future test performance (see Dunning, Johnson, Ehrlinger, & Kruger, 2003; Metcalfe, 1998).

How might such prediction inflation be remedied? In the next sections, we will describe our approach to that question; we will review some relevant findings, and we will lay out the theoretical background that motivates our investigation.

The Foresight Bias and Its Determinants

In our experiments, we used a simple paired-associates task to investigate the foresight bias hypothesis. Participants studied a list of paired associates and made judgments of learning (JOLs) after the study of each pair. Previous studies have established that a critical determinant of both JOLs and recall in this task is the degree of preexperimental association between the two words (see, e.g., Carroll, Nelson, & Kirwan, 1997; Connor, Dunlosky, & Hertzog, 1997; Dunlosky & Matvey, 2001; Koriat, 1997). As far as the foresight bias is concerned, however, a distinction must be drawn between two types of associative relations: a priori associations and a posteriori associations. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.