| |
Abstract:
Abstract: Many cognitive neuroscience researchers have been
using Roediger and McDermott's 1995 associative-word paradigm to
create false memories in order to investigate the underlying brain
mechanisms of such a psychological state. We maintain that the
basic psychological state responsible for the strong false
recognition effect is not a "false memory" but a biased response
(Miller & Wolford, 1999). By including additional conditions,
signal detection analysis demonstrates that the results are
consistent with a criterion-shift model. Other researchers have
maintained that a storage-based model (Wickens & Hirshman,
1999; Wixted & Stretch, 1999) could also explain our results.
In this experiment, we used our modified version of Roediger and
McDermott's word paradigm. In addition, subjects were instructed to
use either liberal or conservative criteria when judging whether a
word was "old" or "new" depending on the color in which the word
was presented. A storage-based model predicts uniform effects
across item types depending on the instructions. Our
criterion-shift model predicts an effect for conservative
instructions but no effect for liberal instructions on critical
lures (items that subjects respond to very liberally), and it
predicts an effect for liberal instructions but no effect for
conservative instructions on unrelated items (items that subjects
respond to very conservatively). The results confirm a criterion
shift hypothesis.
|
