Reconstructing the past: Category effects in estimation

Janellen Huttenlocher, Larry V. Hedges

Research output: Contribution to journalArticlepeer-review

3 Scopus citations

Abstract

This chapter is concerned with people's reconstructions of the past in cases where their memories are inexact. When people are unsure about what actually happened on some occasion, their reconstructions tend to be biased in that they are influenced by categorical or schematic information. It discusses that bias arises in reconstructing particular past episodes even if memory itself is unbiased, but is inexact. People may combine inexact but unbiased information about a particular episode with category information in estimating what actually happened. It has also been suggested that it may be rational to use category information in reconstructing inexactly remembered past episodes. That is, while the use of category information introduces bias, it may, on average, improve accuracy. A formal model is developed that makes precise predictions about the ways category information is used in estimation. The model has been tested for certain limited aspects of memory. The chapter considers the possibility of extending the model to make it applicable to other aspects of memory. The chapter also examines category effects in estimating time and location. Large biases are observed that are fully explained by positing that people used category information to adjust memories for particular episodes that were inexact, but unbiased. Using this adjustment process led to greater overall accuracy than the use of unadjusted memories only.

Original languageEnglish (US)
Pages (from-to)251-280
Number of pages30
JournalPsychology of Learning and Motivation - Advances in Research and Theory
Volume28
Issue numberC
DOIs
StatePublished - Jan 1 1992

ASJC Scopus subject areas

  • Social Psychology
  • Developmental and Educational Psychology

Fingerprint Dive into the research topics of 'Reconstructing the past: Category effects in estimation'. Together they form a unique fingerprint.

Cite this