Planting misinformation in episodic memory through reconsolidation

Bachelorthese Klinische Psychologie

Anne-Sophie Venhuizen (10380922)

Begeleider: Vanessa van Ast

Inleverdatum: 27-05-2016

Planting misinformation in episodic

memory through reconsolidation

Anne-Sophie Venhuizen

Abstract

The incorporation of false memories into human recollection after exposure to misinformation is a well established phenomenon. A possible mechanism through which this misinformation effect can be achieved is reconsolidation, the process of retrieving and restoring existing memories. The present study investigated if misinformation could be implemented in declarative memory by making use of the malleable state of a memory directly after retrieval. Twenty eight psychology students from the UvA were shown pictures of which some were replaced with misinformation directly after retrieval of the original memory. Recognition of these original pictures was impaired upon testing, but regardless of retrieval being achieved or not. This study therefore did not find that reconsolidation is the underlying mechanism for the misinformation effect.

I.

Introduction

A

trauma-debriefing is a widely applied practice aimed at helping people process the traumatic event and preventing any further mental damage, like PTSS. Unfortunately, re-search has shown that the opposite can result. It appears that trauma-debriefing can alter a person’s memory in such a way that a person can be made to remember things that they didn’t experience during the event, for ex-ample by the debriefer telling the traumatized person about experiences of other victims. This in turn makes the memory of a traumatic event worse and can even increase the likelihood of negative consequences for mental health [1,2]. How is it possible for people to be convinced of having certain memories of events that never actually happened?

This effect, also known as the

”misinformation-effect”, has been demon-strated in many studies over the past years. People can be led to believe having seen things that were never there and even incorporate an entire episode of events into the recollection of their childhood [3]. However, debate still very much exists concerning the mechanisms underlying the misinformation-effect.

A possible explanation for this can be found in reconsolidation, the process of retrieving and re-storing existing memories. Reconsoli-dation is seen as a mechanism for updating existing memories with newly acquired infor-mation. After having retrieved a memory into consciousness, it becomes unstable and suscep-tible to modification. This implies a window of opportunity for not only updating memories but also intervening in the re-storage of a mem-ory, making alteration of the original memory possible (figure 1).

The field of reconsolidation research lay dormant for a while after its initial discov-ery in the mid 1900’s, until rediscovered by Nader and colleagues in 2000 [4]. Since then a lot of support for the existence of reconsolida-tion and it’s possibilities for altering memories has come from research on the effects of beta-blockers administered during reconsolidation. When propranolol was administered during the retrieval of a (fear)memory, this could erase the emotionality of that memory, making the experience of that memory less emotional from then on [5,6]. However, episodic memory re-mained unaltered. Could it be possible to use retrieval as a window of opportunity to plant misinformation in existing memories? Recent research has found this to be the case.

Figure 1: Visual representation of memory processes during reconsolidation

Chan et al. [7] showed participants a video and later, during retrieval, verbally replaced certain elements of the story with misinforma-tion. Participants later believed this misinfor-mation to be part of the original story, claiming to have seen it in the video. Other research has also shown that participants did not rec-ognize the original information, implying that the original memory was overwritten by the misinformation [8]. These studies planted mis-information verbally, but it is also possible to replace one image with another, as was shown in a study using photographs [9]. This shows that misinformation can be incorporated into memory in multiple ways.

Although it may now appear to be quite easy to plant misinformation wherever you want whenever you want, this is not the case. Not only has research on planting misinforma-tion during reconsolidamisinforma-tion so far been scarce, there are also some boundary-conditions for reconsolidation to take place and for misin-formation to be incorporated into the original memory [10]. For example, misinformation needs to be similar enough to the original

infor-mation, making it possible to replace an apple with an orange, but not an apple with an ele-phant. This would explain why we don’t incor-porate misinformation all the time in everyday life, which would create highly dysfunctional memories. Though perhaps the most impor-tant boundary-condition is for retrieval to take place, otherwise the memory won’t become un-stable and hence susceptible to alterations. Of possible influence on retrieval is the strength of the memory, a stronger memory causing stronger retrieval and thus a more unstable memory, or at least a higher certainty that a memory will be retrieved and therefore unsta-ble. Although a recent study has found that older and therefore stronger memories are sus-ceptible to modification [11], there has long been the idea that the weaker a memory is, the easier it will be to implement misinformation [12]. What the exact relationship is between the strength of a memory and the susceptibility to misinformation is still unclear.

It has been shown that emotionally arous-ing memories are stronger and better remem-bered [13]. It can therefore be expected that

if stronger memories are more susceptible to alterations, emotional memories will be better replaceable with misinformation. In the case of the aforementioned trauma-debriefing and it’s possible negative effects, it is of particular importance to know if emotionally arousing memories are indeed so easily altered and if the underlying mechanism is reconsolidation, or if something else is going on.

The present study focuses on contributing to the current scientific base for reconsolida-tion as the underlying mechanism for misinfor-mation, as well as exploring the relationship between memory-strength and susceptibility to misinformation. Neutrally and negatively charged pictures were used as a means to cre-ate stronger and weaker memories, the nega-tive pictures being more emotionally arousing and therefore better remembered. We expect to find that reconsolidation is indeed the under-lying mechanism for the misinformation-effect by finding that when retrieval was achieved, pictures were replaceable with misinformation. We also expect the strength of a memory to be of influence on the susceptibility to misin-formation, making the negatively charged pic-tures better replaceable with misinformation and by finding a positive relationship between memory strength and susceptibility to misin-formation.

The experiment was spread out over 3 days. On the first day participants were exposed to neutral and negative pictures set against neu-tral backgrounds. On the second day partici-pants were shown background pictures from day 1, representing the reactivation condition, and scrambled backgrounds, representing the non-reactivation condition, and instructed to report their sense of reliving brought on by this background picture. Then they were shown either the original picture, a misinformation picture, or a completely different picture. On the third day participants were shown pictures they saw on day 1 that had been replaced by either a misinformation or different picture, pictures of day 1 that had not been replaced, misinformation and different pictures that re-placed the original day 1 pictures, and

com-pletely new pictures they hadn’t seen before. They were asked per picture if they recognized them.

Participants are expected to mistake the misinformation planted on day 2 for the origi-nal pictures seen on day 1. Seeing as the misin-formation should have overwritten the original memory, the day 1 pictures that were replaced with misinformation should not be recognized on day 3. Also reliving is hypothesized to indi-cate a higher degree of retrieval and therefore better opportunity for planting misinformation. Reliving is therefore expected to correlate pos-itively with susceptibility to misinformation, and since negative pictures are expected to be better remembered, susceptibility to misinfor-mation should be highest for pictures of this category.

II.

Methods

I.

Participants

For the experiment, 32 psychology and commu-nication students from the University of Ams-terdam were recruited. This was done through a recruitment-add on the LAB-site and flyers distributed throughout the university campus. After recruitment, 1 participant didn’t show up and 3 dropped out on the second day, 1 of which because of technical difficulties and 2 due to the unpleasant nature of some of the negative pictures. The remaining 28 partici-pants with a mean age of 20.96 (SD = 1.74, 18-26) completed the experiment and were re-warded with 1 participation-point per hour, accumulating to 3 in total after completion. Exclusion-criteria were: having had treatment in the past year for a psychological disorder, having epilepsy, and having had a severe trau-matic experience.

II.

Materials

Four questionnaires were used for exploratory means: the BDI-II-NL [14], trait and STAI-state [15], and the PANAS [16]. The BDI and

STAI-trait were used to determine participant’s general pattern of emotional states, especially depressive tendencies. The STAI-state and PANAS were used to map out participant’s moods before and after the experiment.

A medical screening was done before start-ing the experiment to make sure participants were fit to partake and didn’t meet any of the exclusion criteria. After completion of the ex-periment a short self-made questionnaire was administered to measure motivation among other things.

Setting

The task and questionnaires were administered through a desktop computer in a secluded, noise-proof testing-cubicle in the LAB-building on the University of Amsterdam campus. Computer task

Day 1: One picture-trial on day 1 consisted of the following: First a background picture was shown for 2 seconds. After these 2 seconds the target-picture, either neutral or negative, was shown in the middle of this background picture for 4 seconds. The target-picture then disappeared, leaving just the background pic-ture for another 0.5 second. Participants then had 2.5 seconds to rate how vividly they expe-rienced the picture-trial by adjusting a slider that ranged from 0 tot 10, having 0 indicate not vivid at all and 10 as vivid as can be. The screen then turned black for 1, 2 or 3 seconds before the start of the next trial. Every back-ground - picture combination was unique.

Amount of trials: every participant com-pleted 100 trials of backgrounds with negative pictures and 100 trials of backgrounds with neutral pictures.

Day 2: On day 2, a trial consisted of the following: Participants completed an equal amount of reactivation and non-reactivation tri-als. In a reactivation trial, participants needed to be reminded of the target picture for re-consolidation to take place and for a sense of reliving to be established. A conditioned stimulus that was previously linked to the tar-get memory can remind a person of the tartar-get

memory and therefore trigger reconsolidation [17]. To do this, the background picture that was paired with a specific picture the day be-fore was presented for 2 seconds. Research has shown that presenting this background picture is a sufficient context-cue for reconsolidation and therefore retrieval to be established [18,19]. During these 2 seconds participants had to im-mediately indicate their sense of reliving by adjusting a slider that ranged from 0 tot 10, having 0 indicate no reliving at all and 10 indi-cate total reliving. Then the target-picture was shown for 4 seconds, being the original picture, a misinformation picture or a new picture. Dur-ing the display of this picture, participants had to indicate whether they had seen the picture on day 1 (OLD) or not (NEW) by selecting one of the following answers:

1. Very sure that this picture is NEW. 2. Somewhat sure that this picture is NEW. 3. Guess that this picture is NEW.

4. Guess that this picture is OLD.

5. Somewhat sure that this picture is OLD. 6. Very sure that this picture is OLD. The screen then turned black for 1, 2 or 3 sec-onds before starting the next trial. A trial in the non-reactivation condition resembled this pro-cedure precisely, apart from the background picture. This was a scrambled and therefore unrecognizable one in stead of the original one from day 1, making retrieval impossible.

Amount of trials: in total 200 trials were completed per participant. Of these trials, 40 background - picture combinations were correct, as shown on day 1. There were 80 misinformation-trials, the target-picture that was shown being a picture similar to the mis-information trial. The remaining 80 trials were with the target-picture being a new picture, dis-similar to the original one shown on day 1. The division between neutral and negative pictures was equal per every type of trial. An example of these 3 types of trials can be found in figure 2.

Figure 2: Examples of consistent, misinformation and new picture-trials

Day 3: On day three a trial consisted of a target-picture with no background being shown to participants, having them select one of 6 options as described above to have them indicate if they saw the picture on day 1 (OLD) or not (NEW).

Amount of trials: 280 target-pictures were shown, of which 120 pictures that were shown on day 1. Of these day 1 pictures, 40 were shown on day 2 in correct combination, 40 were replaced by a new image on day 2, and 40 were replaced by a similar image on day 2. Another 40 pictures were the misinformation-pictures shown on day 2, and another 40 pictures were the new pictures shown on day 2. The remain-ing 80 pictures were new pictures that partici-pants hadn’t seen before, 40 of which neutral and 40 negative.

III.

Procedure

Day 1: Participants entered the lab and read the information brochure about the experiment. They then completed a short medical screening.

After signing the informed-consent slip they entered the testing-cubicle. After taking place behind the computer, they first completed the BDI and STAI-t. Participants filled in BDI and STAI-s. Before being shown the pictures, par-ticipants were instructed to ”Vividly imagine the central image in it’s surrounding context”. They then saw 200 trials, rating vividness per every trial. After completing the PANAS and STAI-s, the experimenter had a small chat with the participant about how they experienced the task, after which they went home.

Day 2: Participants entered the lab and were sat in the testing cubicle. They then completed the PANAS and STAI-s again. Before start-ing on the picture trials, participants were in-structed to ”Try to relive your memories from day 1. Indicate whether and how you recog-nize the central image”. 200 trials were shown with participants indicating reliving every trial, for which they had 2.5 seconds. Afterwards they completed the PANAS and STAI-s again. The experimenter had a small chat with the participant about how they experienced the

task, after which they went home.

Day 3: Participants were sat in the testing cubicle. They then completed the PANAS and STAI-s again. Before starting the picture-trials, they were instructed to ”Indicate whether you recognize the depicted image as seen during the first day of this experiment”. Participants then completed 280 trials with X seconds to indicate recognition. After completing the PANAS, STAI-s and a small exit-questionnaire, the experimenter chatted with the participants about how they experienced the study and what they thought the objectives were, after which they were done with the entire experi-ment.

IV.

Analysis

To confirm that reconsolidation is the process behind the misinformation effect, two things are expected to happen: in the reactivation

condition, the original pictures that were re-placed by misinformation pictures must not be recognized as such on day 3, as they were over-written and therefore forgotten. Consequently, these misinformation pictures that overwrote them should be mistaken for the original tures and therefore recognized as original pic-tures on day 3.

Also, if better remembrance will have better retrieval and thus better implementation of mis-information as effect, reliving should correlate positively with recognition of misinformation pictures in the reactivation condition. If neg-ative pictures are indeed better remembered, they should generally have a higher score on recognition than neutral pictures. For nega-tive misinformation pictures in the reactivation condition that should result in a higher score on recognition than neutral pictures, and for negative day 1 pictures that are overwritten with misinformation in the reactivation con-dition that should result in a lower score on recognition than neutral ones.

Day 1 pictures

To find this out, we will conduct a 3x2x2 re-peated measures ANOVA on the recognition of the pictures shown on day 1. The three inde-pendent variables are type of picture that day 1 pictures are replaced with on day 2 (same, mis-information, new), the reactivation condition (reactivation, non-reactivation) and emotion-ality of the picture (neutral, negative). The outcome variable is recognition of a picture on a scale from 0 to 1, having 0 indicate not recognized as a day 1 picture, and 1 indicate recognized as a day 1 picture. For a visual representation of the expected data, see figure 3.

Day 2 pictures

A 2x2x2 repeated measures ANOVA will be conducted on the recognition of day 2

pic-tures. The three independent variables here are the type of picture (misinformation, new), the reactivation condition (reactivation, non-reactivation) and emotionality of the picture (neutral, negative). The outcome variable is recognition of a picture on a scale from 0 to 1, having 0 indicate not recognized as a day 1 picture, and 1 indicate recognized as a day 1 picture. For a visual representation of the expected data, see figure 4.

Memory strength

Also a Pearson regression will be performed with reliving as the independent variable and recognition of day 2 misinformation pictures in the reactivation condition as the dependent variable. The aforementioned ANOVA’s will show if there is any effect of emotionality on recognition.

III.

Results

I.

Analysis of the day 1 pictures

To establish if the original pictures that were re-placed with misinformation were less well rec-ognized on day 3, a 3x2x2 repeated measures ANOVA of the Day 1 pictures was executed.

There was no significant main effect for re-activation condition, F(1, 26) = 0.988, p> 0.05, indicating that it did not matter if the pictures on day 2 were shown after a scrambled or cor-rect background for the day 1 picture to be recognized on day 3. It had no effect on recog-nition of original pictures that were replaced by misinformation if reconsolidation took place or not.

Also there was no significant effect for emo-tionality of the picture, F(1, 26) = 0.379, p>0.05, meaning that there was no difference in

recog-nizing neutral or negative pictures. It appears that it is not the case here that negative memo-ries are stronger than neutral ones.

There was however a significant main ef-fect for type of picture, F(2, 52) = 208.423, p<0.05, which means that the type of picture shown on day 2 had effect on recognition of day 1 pictures on day 3. Contrasts revealed that recognition was significantly higher for consistent pictures than for pictures replaced with new ones, F(1, 26) = 383.253, p<0.05, and that recognition was significantly higher for pictures replaced with misinformation than for pictures replaced with new ones, F(1, 26) = 63.723, p<0.05. Though pictures replaced with misinformation were less well remem-bered than consistent pictures, the difference was not as big as expected. Also for pictures replaced with new pictures recognition was lowest, which was not predicted.

No significant interaction effect was found between the picture type, emotionality and re-activation, F(2, 52) = 1.631, p> 0.05. This indi-cates that reactivation did not have different effects for different types of pictures of differ-ing emotionality. It appears that day 1 pictures that were replaced with misinformation in the reactivation condition were not clearly over-written as a result of reconsolidation, with no difference in memory strength for negative and neutral pictures. For an overview, see figure 5.

II.

Analysis of the day 2 pictures

To see if the misinformation pictures were rec-ognized as original pictures on day 3, a 2x2x2 repeated measures ANOVA of the day 2 pic-tures was performed.

There was a significant main effect for emo-tionality, F(1, 26) = 8.796, p>0.05, which

indi-cates that emotionality had effect on recogni-tion of day 2 pictures on day 3, where recog-nition was significantly higher for neutral pic-tures than for negative picpic-tures in all cases but the non-reactivation misinformation pic-tures. Emotional memories were not consis-tently stronger than non-emotional ones as ex-pected, possibly the other way around.

There was also a significant main effect for type of picture, F(1, 26) = 25.752, p>0.05, indicating that the type of picture had effect on recognition of day 2 pictures on day 3, recognition being significantly higher for mis-information pictures than for new pictures. Misinformation was indeed often mistaken for original pictures, though recognition was not as high as expected and recognition of new pictures was not as low.

There was a significant main effect for re-activation condition, F(1, 26) = 19.783, p>0.05, which means that reactivation condition had effect on recognition of day 2 pictures on day 3, with pictures in the reactivation condition having significantly higher recognition than pictures in the non-reactivation condition. For misinformation this was as expected, though for the new pictures this was not predicted.

There was no significant interaction effect between emotionality, type of picture and re-activation condition, F(1, 26) = 1.292, p>0.05. Reactivation did not have different effects for different types of pictures of differing emotion-ality. There was however a significant interac-tion effect between emointerac-tionality and type of picture, F(1, 26) = 4.412, p<0.05, with higher recognition for neutral new pictures than for negative new pictures. The interaction effect be-tween emotionality and reactivation condition was also significant, F(1, 26) = 5.793, p<0.05, with higher recognition for neutral picture than for negative pictures in the reactivation condi-tion. Finally, the interaction effect between type of picture and reactivation condition was also significant, F(1, 26) = , p<0.05, with higher recognition for misinformation than new pic-tures and higher recognition in the reactivation condition than in the non-reactivation condi-tion. Misinformation was indeed more often mistaken for original pictures in the reactiva-tion condireactiva-tion, though not as much as expected and this was also the case for new pictures. For an overview, see figure 6.

III.

Analysis of the relationship between

reliving and recognition

To find out if the better a picture is remem-bered, the better retrieval will take place and therefore the better misinformation will be in-corporated into memory, a Pearson regression was performed with reliving as the indepen-dent variable and recognition of day 2 misin-formation pictures as dependent variable.

There was no significant correlation be-tween the two variables, F(1, 25) = 1.292, p>0.05, with an R squared of 0.011,

mean-ing that relivmean-ing did not significantly predict recognition of day 2 misinformation pictures. A stronger memory was not predictive of stronger retrieval and therefore a stronger mis-information effect.

IV.

Discussion

I.

Main findings

This study investigated if reconsolidation is the underlying mechanism for the misinformation effect. This was not unequivocally found: the data showed that day 1 pictures that were re-placed by misinformation on day 2 hovered somewhere halfway between recognized and not recognized, regardless of whether or not reconsolidation was achieved beforehand. It is therefore unlikely that reconsolidation had effect or maybe even took place. The misin-formation pictures from day 2 were also in this same range, with a higher recognition in the reactivation condition, which is more con-form our hypothesis that these pictures would be mistaken for original pictures. However, to conclude that there was a misinformation effect that was established through reconsoli-dation, there should have been a difference in recognition for original pictures that were re-placed by misinformation after reconsolidation allegedly took place. This study therefore did not clearly contribute to the scientific base for reconsolidation as an underlying mechanism for the misinformation effect.

The relationship between reliving and recognition was also explored. Though there was an expected positive correlation between memory strength and susceptibility to misin-formation, no such relationship was found. Also the expected higher remembrance of neg-ative pictures did not consistently come forth in the data. Recognition was sometimes higher for negative pictures within a condition, often higher for neutral, and sometimes there was no difference. At the present no specific explana-tion for this was found, leaving this for future research to explore.

day 1 pictures had a remarkably different dis-tribution than expected. Recognition scores of the consistent day 1 pictures were somewhat as expected, but the scores for both reactiva-tion an non-reactivareactiva-tion misinformareactiva-tion day 1 pictures were significantly lower than these. Also the recognition scores of the new day 1 pictures were significantly lower than those of the misinformation day 1 pictures. A pos-sible explanation for the lower scores of the misinformation and new pictures could be that consistent pictures were seen by participants twice, on day 1 as well as day 2, before being tested on day 3. Having only seen the day 1 pictures that were replaced by misinformation or new pictures on day 2 once, these were less well learned and remembered than the consis-tent pictures. Possibly for this same reason misinformation day 1 pictures could be better remembered than new day 1 pictures, having seen the misinformation day 2 pictures be simi-lar to having seen the consistent pictures twice. It is quite peculiar that scores for the new day 2 pictures are so relatively high, though still lower than misinformation day 2 pictures. A possible explanation for the higher scores for misinformation pictures in the reactivation condition could be that after reactivation, par-ticipants were more confused about whether or not these pictures were the same or slightly different than the pictures seen on day 1, since they went through the reliving of the back-ground in combination with the picture. This picture being so similar in their memory of the background-picture combination, for example a snake interacting with a living room back-ground, that the exact details of the picture mattered less. Not having to go through the reliving of the background-picture combina-tions and therefore not having this clouded judgment, a misinformation picture in the non-reactivation condition can possibly be better detected as being discrepant. However, this confusion is probably not caused by a misin-formation effect through reconsolidation but rather by confusion caused by the strong simi-larity of the pictures.

There also appeared to be a more

gen-eral confusion concerning which pictures be-longed to which day, also referred to as source-confusion [8], based on the relatively high scores of all day 2 pictures, and the relatively low scores of day 1 pictures that were replaced by misinformation or new pictures. This confu-sion was also reported by participants, who were often unsure about which picture be-longed to which day. Possible reasons could be fatigue and lapse of concentration, which was also often reported by participants.

II.

Our research compared to other

re-search

Our study did not clearly contribute to the scientific base for reconsolidation as the under-lying mechanism of the misinformation effect. Also the link between memory strength and susceptibility to misinformation is therefore still unclear. There are multiple possible rea-sons for this, the first of which is that there is no misinformation effect through reconsol-idation. But before the conclusions of many other previous studies on this subject are dis-carded as incorrect, let’s explore the possible weaknesses of the current study.

The lack of difference between the reactiva-tion condireactiva-tion and non-reactivareactiva-tion condireactiva-tion for the day 1 pictures can be due to the way in which reconsolidation was operationalized. Background-pictures were used as a means to bring about retrieval of a center-picture on day 2 and establish reconsolidation thusly. This was based on previous research stating that this was a sufficient cue for activating a memory. It is possible that this is either not the case, or this was not good enough for our specific research set-up or question. No other previous study that we know of tried to bring about a misin-formation effect through reconsolidation in the way that we did, so future research should ex-plore different aspects and effects of different operationalizations of reconsolidation.

A previous study using photographs as a means to implement misinformation did find the effect [9]. This was done by having peo-ple partake in a museum tour whilst having

a camera take pictures of things they passed during this tour. Later these pictures were replaced by pictures of things they did not pass in the museum, which participants later believed to have seen. A possible difference with the current study could be that this mu-seum tour set-up is much closer to day to day episodic memory, where our set-up is much more unnatural and fabricated. Also our set-up is more dependent on how much of an active effort participants make to imagine and relive background-picture combinations, in stead of having it be dependent on actual experience.

As mentioned before, participants often re-ported confusion about what pictures belonged to which day and also indicated being quite fatigued after the trials on day 1 and 2, saying that it was long and even boring. This could have an effect on concentration and the active effort people put into remembering and recall-ing pictures. Compared to other studies [7, 8, 9], our amount of items to be remembered and recalled was around 10 times more, these other studies often using about 10 to 20 items. This could have a big effect on memory and future research could take this into account by comparing the effects of different quantities of pictures.

Finally, the number of participants in this study could be called limited. Though even with a small number of participants a differ-ence between reactivation end non-reactivation should have been visible, it is still a point of improvement for future research.

Despite the possible weaknesses of our study, there could also be some possible strengths. After closer comparison with an arti-cle mentioned before by Chan and LaPaglia [7], it appeared that their non-reactivation condi-tion differed from the one in the present study. Theirs consisted of playing Tetris in stead of being presented with misinformation without the reactivation component, as was done in this study. Participants therefore weren’t con-fronted with misinformation until the memory test, making it less confusing for the partici-pants in the non-reactivation condition which

picture was seen during the initial learning phase. Also this study, consisting of 6 sepa-rate experiments, couldn’t completely rule out source-confusion even in the reactivation condi-tion, which came forth in the experiment were they asked if the picture was from the original learning phase or not, as we asked too. When participants solely had to indicate whether or not they recognized the picture, a misinforma-tion effect was found, but when the quesmisinforma-tion was posed as was done in our article, this effect became smaller and less reliable. The conclu-sions of this study are therefore not as strong as initially thought.

Finally, very recent research that tried to alter different types of memory by disrupting reconsolidation [20] did not find that it reli-ably altered existing declarative memories. It should be noted, however, that these memories consisted of number sequences, which is quite different from more visual memories of events as used in other research as well as our own.

An explanation for the phenomenon men-tioned at the beginning of this article, namely that of people incorporating false memories into their recollection of traumatic events, is yet to be found. It is important to obtain more clar-ity concerning this problem in order to prevent psychological damage being done to people during future trauma-debriefings. This may still lie in reconsolidation, though for this to be established, further research is necessary.

References

[1] Paterson, H. M., Whittle, K. & Kemp, R. I. (2014). Detrimental Effects of Post-Incident Debriefing on Memory and Psy-chological Responses. J Police Crim Psych 30, 27-37

[2] Devilly, G. J., Varker, T., Hansen, K., & Gist, R. (2007). An analogue study of the effects of psychological debriefing on eye-witness memory. Behaviour research and therapy, 45(6), 1245-1254.

[3] Loftus, E. F. (2005). Planting misinforma-tion in the human mind: A 30-year in-vestigation of the malleability of memory. Learning & Memory, 12(4), 361-366. [4] Nader, K., Schafe, G. E., & Le Doux, J. E.

(2000). Fear memories require protein syn-thesis in the amygdala for reconsolidation after retrieval. Nature, 406(6797), 722-726. [5] Kindt, M., Soeter, M., & Vervliet, B. (2009).

Beyond extinction: erasing human fear re-sponses and preventing the return of fear. Nature neuroscience, 12(3), 256-258.

[6] Brunet, A., Orr, S. P., Tremblay, J., Robert-son, K., Nader, K., & Pitman, R. K. (2008). Effect of post-retrieval propranolol on psy-chophysiologic responding during subse-quent script-driven traumatic imagery in post-traumatic stress disorder. Journal of psychiatric research, 42(6), 503-506.

[7] Chan, J. C., & LaPaglia, J. A. (2013). Im-pairing existing declarative memory in hu-mans by disrupting reconsolidation. Pro-ceedings of the National Academy of Sciences, 110(23), 9309-9313.

[8] Hupbach, A., Gomez, R., & Nadel, L. (2009). Episodic memory reconsolidation: updating or source confusion?. Memory, 17(5), 502-510.

[9] Jacques, P. L. S., & Schacter, D. L. (2013). Modifying Memory Selectively Enhanc-ing and UpdatEnhanc-ing Personal Memories for a Museum Tour by Reactivating Them. Psychological science, 24(4), 537-543. [10] Finnie, P. S. B. & Nader, K. The role of

metaplasticity mechanisms in regulating memory destabilization and reconsolida-tion. Neurosci Biobehav Rev 36, 1667-1707 (2012).

[11] Wichert, S., Wolf, O. T., & Schwabe, L. (2013). Changing memories after reactiva-tion: A one-time opportunity?. Neurobiol-ogy of learning and memory, 99, 38-49.

[12] Tousignant, J. P., Hall, D., & Loftus, E. F. (1986). Discrepancy detection and vulnera-bility to misleading postevent information. Memory & Cognition, 14(4), 329-338. [13] McGaugh, J. L., McIntyre, C. K., & Power,

A. E. (2002). Amygdala modulation of memory consolidation: interaction with other brain systems. Neurobiology of learn-ing and memory, 78(3), 539-552.

[14] Beck, A. T., Steer, R. A., Brown, G. K.,& Van der Does, A. J. W. (2002). BDI-II-NL Handleiding [BDI-II-Dutch Manual]. Lisse, the Netherlands: Psychological Corporation. [15] Spielberger, C. D. (1970). STAI manual

for the state-trait anxiety inventory. Self-Evaluation Questionnaire, 1-24.

[16] Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: the PANAS scales. Journal of person-ality and social psychology, 54(6), 1063. [17] Mueller, D., & Cahill, S. P. (2010).

Nora-drenergic modulation of extinction learn-ing and exposure therapy. Behavioural brain research, 208(1), 1-11.

[18] van Ast, V. A., Cornelisse, S., Meeter, M., & Kindt, M. (2014). Cortisol me-diates the effects of stress on the

contextual dependency of

memo-ries.Psychoneuroendocrinology, 41, 97-110. [19] van Ast, V. A., Cornelisse, S., Meeter,

M., Joëls, M., & Kindt, M. (2013). Time-dependent effects of cortisol on the con-textualization of emotional memories. Bio-logical psychiatry, 74(11), 809-816.

[20] Hardwicke, Tom E., Mahdi Taqi, and David R. Shanks. "Postretrieval new learn-ing does not reliably induce human mem-ory updating via reconsolidation." Pro-ceedings of the National Academy of Sciences (2016): 201601440.