| |
Abstract:
Abstract: False memories were elicited using the conjunction
paradigm, in which subjects study compound words (e.g., blackbird,
airmail) and are tested on lures that contain elements of studied
words (e.g., blackmail). Using event-related fMRI procedures and
randomly-ordered trials, we found robust differences among the
three types of items tested: studied, conjunction, and unrelated
nonstudied. Performance-based comparisons (e.g., false alarms to
conjunction items compared to correct rejection of these items)
were not fruitful, however, due to a low probability of false
alarms (.27) to these items. For the comparison between conjunction
lures and studied items, differences were seen in multiple regions,
including L inferior and middle frontal cortex, and L inferior
parietal cortex (all conjunction>studied). The comparison
between studied and standard nonstudied items demonstrated
differences in bilateral anterior prefrontal cortex
(studied>nonstudied), inferior parietal cortex
(studied>nonstudied) and R extrastriate visual cortex
(nonstudied>studied). The studied>nonstudied differences
cannot be explained by calling upon time-on-task explanations, as
reaction times for nonstudied items exceeded those for studied
items. The results with respect to anterior prefrontal cortex
inform the debate over the role of this region in episodic memory
retrieval and support the conclusion that retrieval success
modulates activation levels in this region, even when
random-ordering of trial types is used (and therefore context
effects can be discounted).
|
