This study deals with the production of false memories in the DRM paradigm and the use of a distinctiveness heuristic to prevent the production of false memories.  The main idea behind the distinctiveness heuristic is that if your memory for targets is especially distinctive you should be less likely to accept false memories because the false memories will seem relatively sparse by comparison.

Schacter's conducted a couple previous studies looking at the distinctiveness heuristic.  In these studies he and his colleagues found lower rates of false memories when items are presented as pictures than when items are presented as words (Israel & Schacter, 1997; Schacter, Israel & Racine, 1999).  If one assumes that pictures result in more distinctive memories than words then you can account for these findings by means of the distinctiveness heuristic.

The basic purpose of the present research is to show that the distinctiveness pattern generalizes beyond the picture/word distinction.  This is important because its always possible that some other characteristic of pictures is what produces the effect not their memorial distinctiveness per se.

The manipulation that they decided to use was to compare cases in which the subjects say the target words out loud during study with cases in which they heard the words.

Participants were visually presented with sixteen DRM lists at a rate of one word every 1.5 seconds.

• In the hear condition subjects also heard the word 400 msec after the onset of the visual presentation.
• In the say condition subjects read the words out loud as they were being presented

• Targets: Items that had actually been presented at study
• False Targets: The non-presented word to which the study items were related (i.e. critical lures)
• Target Controls: Targets from non-presented lists (i.e. distractors)
• False Target Controls: Critical lures from non-presented lists. (i.e. distractor lures)

• Target vs. Target Control: This is just the traditional comparison between hits and false alarms.  It provides an indication of how much more common it was for subjects to select targets than for subjects to select completely unrelated lures.  Note that sensitivity here is well above chance and that there is no significant difference between the Hear and Say conditions.
• Target vs. False Target: This one compares memory for targets and critical lures.  In the Hear condition A' = .50.  This indicates that subjects are just as likely to select critical lures as they are targets (Go back to table 1 and you'll see that's true).  In the Say condition A' = .69.  This indicates that subjects were somewhat more likely to select targets than critical lures.
• False Targets vs. False Targets Controls: This one compares recognition of critical lures with recognition of critical lures from non-presented lists.  In both conditions A' is significantly greater than chance (i.e. .50).  This indicates that selection of the critical lures isn't just a function of random guessing.  The A' score for the Hear condition was greater than the A' score for the Say condition.

So in this experiment subjects were presented with some lists where they heard the words and some lists where they were asked to read the words.
 

	Targets	Target Controls	False Targets	False Target Controls
Hear	.70	.12	.57	.14
Say	.79	.12	.58	.14

 

A'	Target vs. Target Control	Target vs. False Target	False Target vs False Target Control
Hear	.86	.62	.76
Say	.88	.68	.77

Its quite clear in these results that there is no longer a difference in the rate of false memories for the Hear  and Say conditions.  So these results are in agreement with what the authors predicted.

This is interesting because a number of possible false memory supression mechanisms have recently been proposed (this diverges a little from Dodson's discussion, but its worth pursuing):

• Distinctiveness Heuristic: Expectations about the likely distinctiveness of items leads to rejection of false memories when distintiveness expectations are high (Israel & Schacter, 1997; Schacter, Israel & Racine, 1999).
• Recollection Rejection (Recall to Reject): Conscious recollection of target leads to the rejection of related lures (I know I didn't have a hotdog at the game because I remember having pizza at the game). (Brainerd, et. al, in press; Brainerd, et. al, 1995; Rotello, et. al, 2000; Rotello & Heit, 2000; 1999)
• Production Recollection: I know the critical lure wasn't presented because I remember generating the critical lure at time of study (Bredart, 2000)

So an interesting direction for false memory research right now has been to understand not just the mechanism through which false memories are produced but also the mechanisms through which they are suppressed.

The authors also take up the issue of whether a pure similarity account could explain their data.  They argue that it would be difficult for a pure familiarity account to explain why the distinctiveness pattern disappears in the within subjects manipulation.