tronger startle responses to the loudnoise during the fear-conditioned stimulus (CS1+) as compared withthe control stimulus (CS2–) reflects the fearful state of the participantelicited by CS1+.

CS1-R

ef

TestExtinctionAcquisition

Reconsolidation offear memory was manipulated by administration of propranolol(40 mg,n¼20), randomized and double-blind placebo controlled(n¼20) (seeSupplementary Methodsonline). For the additionalcontrol condition (n¼20), propranolol (40 mg) was administeredwithout memory reactivation.

eclarative knowledg

ear-relevant stimuli

neurobio-logical manipulations

Recently, itwas rediscovered that fear memory in animals is not necessarilypermanent but can change when retrieved3–5. The reactivation of aconsolidated (fear) memory can return it to a labile, supposedly proteinsynthesis–dependent state, a process that is referred to as reconsolida-tion4.

r=−.49,p<.01

Table 1

Recognition d-prime

WM taskAn operation span (OSPAN) task

For example, participants often assigncritical lures morerememberthanknowresponse

Recently, the relationship between working memory(WM) capacity and performance on the DRM task

A signal detection view, on the other hand, suggeststhat participants who have been forewarned may adopt amore conservative criterion during testing, which wouldlead to reduced false-memory rates

been conducted to explore variables that could potentiallyeliminate, or at least reduce, the false-memory effect

activation/monitoring theory

In fact, the intrusionand false-recognition rates for critical lures have often beenfound to be comparable to, or sometimes even exceed,recognition rates of the studied words

ANOVA of Schema Discrepancy by Emotional State Induction on the average occurrence of spatial memory bias,Study 2

Indeedpositive emotional stimuli presented at the topand negative emotional stimuli presented at thebottom of a monitor (stimulus congruent loca-tions) were expected to be more accuratelyretrieved than positive stimuli presented at thebottom and negative stimuli presented at the topof a monitor (stimulus incongruent locations)

reconstructions in confor-mity with the valence (positive vs negative) of thestimuli

An emotional state of alarm was induced inparticipants

ANOVA of Word Valence by Emotional State Induction on average number of accurately recognised words, Study 1.

A22 mixed-design ANOVA was run on theaverage number of correctly recognised words,with Emotional State Induction (Alarm Statecondition vs Control condition) as the between-participants factor, and Word Valence (Negativevs Neutral Words) as the within-participants fac-tor. A significant main effect of Word Valence wasfound to be significant,F(1, 82)29.73,MSE116.67,pB.001: Compared with Neutral Words,Negative Words were more accurately recognised.Moreover, the Emotional State Induction maineffect was significant,F(1, 82)4.51,MSE42.00,pB.01: Participants in the Alarm Statecondition recognised both Negative and NeutralWords better than participants in the Controlcondition (see Figure 2). The interaction effect

he experimental approach is the major strengthof the present work.

reconstructive factors

Now Print!

direct pathway

a considerable amount of FBMresearch has used three parameters as proxiesfor FBM accuracy: quantity of details, consis-tency, and confidence

have criticised thespecial encoding hypothesis, suggesting thatFBMs are subject to the forgetting curve, and toreconstructive and inferential processes, compar-able to those observed for ordinary autobiogra-phical memories

Across both experiments, strong cues led to an initial advantage in recall, but items recalled from weak cues were actually retained better over time, such that the advantage for strong cues was eliminated or reversed at the time of the final test

items learned through study were retained at approximately equal rates on the final test whether they were learned in the context of strong or weak cues

Both experiments revealed a robust testing effect.

Mean Relatedness Ratings

Design and Procedure

If elaborative activation is not as likely to occur during restudy, final test recall of items presented for restudy would not be expected to vary in either experiment as a function of cue–target relatedness.

Initial retrieval was expected to be higher for targets recalled from strong cues than from weak cues. However, because initial retrieval is less direct with weak cues, recalling a target from a weak cue is more likely to involve the activation of more elaborative information (e.g., Basket → Eggs → Flour → Bread) than is recalling a target from a strong cue (e.g., Toast → Bread).

Furthermore, the degree of elaboration during retrieval would seem to be greater under conditions in which the target is less accessible.

the act of retrieval r

Testing could therefore have a mediating benefit on retention by influencing how well learners process the material during their next opportunity to study it

Instead, it was best for items that were given a free-recall initial test, regardless of the type of final test.

has been observed for a wide variety of materials, including word lists (e.g., Carpenter & DeLosh, 2006), general knowledge facts (e.g., Carpenter, Pashler, Wixted, & Vul, 2008), foreign language vocabulary (e.g., Carrier & Pashler, 1992), and text passages (Agarwal, Karpicke, Kang, Roediger, & McDermott, 2008). The effect has also been demonstrated with face–name pairings (e.g., Carpenter & DeLosh, 2005; Landauer & Bjork, 1978)

One explanation that has been offered is based on the notion of transfer-appropriate processing

It may be hypothesizedthat beta-adrenergic blockade during reconsolidation may selectivelydisrupt the protein synthesis of the amygdalar fear memory, resultingin deconsolidation of the fear memory trace while leaving thedeclarative memory in the hippocampus untouched

our finding that a well-established retrieval technique for fear memories (reinstatement)failed to uncover any fear response suggests that the fear memorymay either be erased (storage theory) or may be unavailable as aresult of retrieval failure (retrieval theory)5

Placebo reactivation

Unconditionedstimuli

Placebo reactivation

In addition, declarative knowledge of the con-tingency between the conditioned stimulus and the unconditionedstimulus was measured

ifferential fear-conditioning procedure with fear-relevant stimuli.

Reconsolidation of fear memory can be influenced by neurobio-logical manipulations during or shortly after the reactivation period

Once emotional memory is established, itappears to last forever.

know

remember

Watson et al., 2005).

Experiment 2

recognition d' (a) and remember d' (b)

Fig. 1

WM capacity

recollective detail.

could potentially eliminate, or at least reduce, the false-memory effect. One such condition, which is important for the present study, is explicitly forewarning participants about of the nature of the DRM illusion

activation/monitoring theory

Deese/Roediger–McDermott

the primary objective of the present study was to examinethe link between checking compulsions and habitual prospectivememory

The task requiredparticipants to remind the experimenter to send an email to hersupervisor

Fig. 1.

Episodic

Habitual

Method1.1

we also manipulated therelevance of the time- and event-cued prospective memory tasksin the present study

checkers indeed reportexperiencing more problems with various aspects of prospectivememory

Instead, our focus was on episodicand habitual prospective memory.

f individuals frequently experience prospective memoryfailures (e.g., they frequently forget to attend appointments, to paybills and/or to turn off the iron), they may begin to worry and haveintrusive doubts concerning prospective memory failures (e.g., didI turn off the iron?).

Until recently, the memory deficit theory was tested exclusivelyin the domain of retrospective memory

memory deficit theory

The most common subtype, characterized bychecking compulsions, occurs in over 50% of OCD patients(Henderson & Pollard, 1988; Rasmussen & Eisen, 2002; Steinet al., 1997), with an additional 15% of the general populationdemonstrating sub-clinical checking compulsions

Obviously it isimpossible to reproduce precisely those condi-tions that prevail for a phenomenon such as FBMwhen importing an event into the laboratory fromthe natural setting in which it is manifested.

although the naturalistic methodtakes advantage of capturing a naturally occur-ring phenomenon, it has the disadvantage oflacking sufficient control of both the assessmentof FBMs and the factors intervening in theirformation

Average occurence of Spatial Memory

Now Print!

As a general hypothesis, being in anemotional state of alarm was expected toinfluence directly and indirectly*through recon-structive processes*the individual’s memoryaccuracy

Figure 1.

read immediately after the emo-tional event (negative vs neutral words).

the experimental approach is the major strengthof the present work.

Indeed, research work on FBMs has beenprimarily conducted using a naturalistic question-naire method.

From the aforementioned it follows that twopossible pathways might be hypothesised toaccount for the formation of FBMs

For many years, researchers have being debat-ing whether FBMs can be considered either aspecial class of episodic emotional memories(Bohannon, 1988; Conway et al., 1994; Pillemer,1984), or are affected by reconstructive factors, asordinary autobiographical formations.

Traditional FBM literature suggests that FBMevents are public in nature

Typically, a restudy condition involves presenting the materialagain for participants to read, without any overt response required.In such a condition, it can be difficult to know the degree to whichparticipants are processing the material or even paying attention toit at all. The current study helps rule out this potential problem byincluding a restudy condition that required participants to judge therelatedness between two words in a pair. The consistency of theratings across the encoding and restudy trials indicates that partic-ipants were processing the items well enough to make relevant andaccurate judgments that were in line with what would be expected.

Both experiments revealed an interaction whereby strong cueswere advantageous for initial recall, but items recalled from weakcues were retained better over time.

According to the elaborative retrieval hypothesis, the activationof such information is less likely to occur during restudy and sofinal test recall of restudied items would not be expected to benefitas a function of manipulations that encourage elaborative process-ing during retrieval

elaborative information

4.26 (0.53

Two experiments

If this information is beneficial for retention,items recalled from weak cues should be retained better over time,such that the initial advantage for strong cues is eliminated orreversed at the time of the final test.

Testing could therefore have a mediating benefit onretention by influencing how well learners process the materialduring their next opportunity to study it

Carpenter andDeLosh (2006

transfer-appropriate processing

known answers

This study investigated whether ideomotor actions can express the contents of implicit semantic memory. We comparedresponses to yes/no questions using both volitional reports and Ouija responses. When participants believed they knew theanswer, responses for both modalities were similar. When participants believed they did not know the answer, accuracy fellto chance for volitional reports (50%) but remained significantly higher for Ouija responses (65%). These results indicate thatinformation inaccessible through volitional report can nevertheless be expressed in an ideomotor action in the complete ab-sence of conscious awareness.

we calculated for every participanta similarity indexSdescribing the correlation of responses

   -1 (opposite)    0 (no similarity)    + (same)

guessed answers

Results3.1

To examine this possibility

Material and methods

In this study we examine

Ideomotor actions