Biopsychology memory paper

Authors:

Address:

Source:

NLM Title Abbreviation:

Publisher:

Other Publishers:

ISSN:

Language:

Keywords:

Abstract (English):

Hunting down the source: How amnesic patients avoid fluency-based memory errors.

Geurten, Marie. Cyclotron Research Center, University of Liege, Liege, Belgium, [email protected] Bastin, Christine. Cyclotron Research Center, Psychology and Neuroscience of Cognition Unit, University of Liege, Liege, Belgium Salmon, Eric. Cyclotron Research Center, University of Liege, Liege, Belgium Willems, Sylvie. Psychology and Neuroscience of Cognition Unit, University of Liege, Liege, Belgium

Geurten, Marie, Cyclotron Research Center, University of Liege, B33 Trifacultaire— Quartier Agora, Place des Orateurs 1, 4000, Liege, Belgium, [email protected]

Neuropsychology, Jun 20, 2019.

Neuropsychology

US : American Psychological Association

US : Educational Publishing Foundation US : Philadelphia Clinical Neuropsychology Group United Kingdom : Taylor & Francis

0894-4105 (Print) 1931-1559 (Electronic)

English

amnesia, fluency, metacognition, recognition memory

Objective: The primary aim of this study was to test whether differences in the ability of amnesic and healthy participants to detect alternative sources of fluency can account for differences observed in the use of fluency as a cue for memory. Method: Patients with severe memory deficits and matched controls were presented with 3 forced-choice recognition tests. In each test, an external source of fluency was provided by manipulating the perceptual quality of the studied items during the test phase. The detectability of the perceptual manipulation varied in each test (i.e., a 10%, 20%, or 30% contrast reductions were given). Results: The results indicated that all participants were able to rely on fluency when making recognition decisions as long as the perceptual manipulation remained unnoticed. It is interesting that our data also revealed that the level of contrast reduction at

Impact Statement:

Document Type:

Subjects:

PsycINFO Classification:

Population:

Location:

Age Group:

Tests & Measures:

Grant Sponsorship:

Methodology:

which the alternative source is detected differs between healthy controls and amnesic patients. Specifically, patients with amnesia appeared to disqualify fluency as a cue for memory even when the contrast reduction was moderate, whereas healthy participants disqualified fluency only when the contrast reduction was clearly visible. Conclusion: Overall, our results seem to suggest that the ability to use fluency is probably not impaired in amnesia but undergoes metacognitive changes resulting in the implementation of explicit or implicit strategies aiming at tracking alternative sources in order to reduce memory errors. (PsycINFO Database Record (c) 2019 APA, all rights reserved)

General Scientific Summary: Despite amnesic patients’ severe deficits, some of their memory processes are preserved. Unfortunately, they do not appear to take full advantage of these spared memory abilities. This study is an attempt to determine whether adaptive metacognitive changes could account for the apparent inability of amnesic patients to rely on their preserved memory skills. (PsycINFO Database Record (c) 2019 APA, all rights reserved)

Journal Article

*Amnesia; *False Memory; *Memory; *Metacognition; *Verbal Fluency; Errors; Implicit Memory; Memory Disorders; Patients; Test Items

Neurological Disorders & Brain Damage (3297)

Human Male Female

Belgium

Adulthood (18 yrs & older) Young Adulthood (18-29 yrs) Thirties (30-39 yrs) Middle Age (40-64 yrs)

Contrast Reduction Test Wechsler Abbreviated Scale of Intelligence--Second Edition DOI: 10.1037/t15171- 000 Wechsler Memory Scale III

Sponsor: Fund Maria-Elisa and Guillaume de Beys (FRB) Recipients: No recipient indicated

Sponsor: National Fund for Scientific Research Recipients: No recipient indicated

Empirical Study; Quantitative Study

Format Covered:

Publication Type:

Publication Status:

Publication History:

Release Date:

Copyright:

Digital Object Identifier:

PsycARTICLES Identifier:

Accession Number:

Electronic

Journal; Peer Reviewed Journal

Online First Posting

Accepted: Apr 20, 2019; Revised: Apr 2, 2019; First Submitted: Jan 7, 2019

20190620

American Psychological Association. 2019

http://0-dx.doi.org.wizard.umd.umich.edu/10.1037/neu0000566

neu-neu0000566

2019-33413-001

Hunting Down the Source: How Amnesic Patients Avoid Fluency-Based Memory Errors

By: Marie Geurten Cyclotron Research Center and Psychology and Neuroscience of Cognition Unit, University of Liège; Christine Bastin Cyclotron Research Center, Psychology and Neuroscience of Cognition Unit, and National Fund for Scientific Research, University of Liège Eric Salmon Cyclotron Research Center and Psychology and Neuroscience of Cognition Unit, University of Liège Sylvie Willems Psychology and Neuroscience of Cognition Unit, University of Liège Acknowledgement: This research was supported by a grant from the Fund Maria-Elisa and Guillaume de Beys (FRB) and by the National Fund for Scientific Research. The authors have no conflict of interest to declare.

Over the past 50 years, research focusing on memory impairments associated with amnesia has

Listen American Accent

generated a large array of findings, which in turn has led to many theoretical advances. Specifically, much attention has been directed toward the identification of increasingly refined and sophisticated dissociations. Thereby, researchers have learned that amnesic patients usually show spared short-term memory versus impaired long-term memory (Baddeley & Warrington, 1970) and demonstrate intact nonconscious long-term memory versus altered conscious long-term memory (Squire & Zola, 1996). Within the conscious long-term memory deficits, patients with amnesia have also been found to exhibit pronounced deficits in recollection, defined as the ability to mentally relive past events in vivid details, while showing no or less impairment in familiarity, defined as a vague feeling of “oldness” associated with past experiences (e.g., Addante, Ranganath, Olichney, & Yonelinas, 2012; Bastin et al., 2004; Yonelinas, Kroll, Dobbins, Lazzara, & Knight, 1998). Indeed, in a comprehensive review of studies focusing on recollection and familiarity in amnesia, Yonelinas et al. (1998) concluded that both processes are compromised in amnesic patients but that the impairment in familiarity is typically less severe than that in recollection. Since then, the question of whether and when familiarity is impaired in amnesia has been hotly debated (e.g., Keane, Orlando, & Verfaellie, 2006; Ozubko & Yonelinas, 2014; Squire, 2004).

According to many authors, processing fluency, defined as the speed and ease with which a stimulus is processed, is a key factor to understand familiarity-based memory decisions (e.g., Jacoby & Dallas, 1981; Whittlesea, 1993; Willems, Germain, Salmon, & Van der Linden, 2009). Specifically, because people intuitively know that an earlier encounter with a stimulus generally enhances processing fluency, it is usually assumed that a feeling a familiarity can result from attributional processes whereby people ascribe fluency to the past. This view, however, has been challenged by studies showing that some patients with amnesia are not able—or at least less able —to use fluency as a cue for recognition memory, despite successfully completing a priming task conducted on the same set of stimuli (e.g., Levy, Stark, & Squire, 2004). These data suggest that processing fluency can occur without giving rise to better explicit memory judgments, leading several authors to conclude that fluency has no or only small influences on amnesic patients’ memory decisions (Conroy, Hopkins, & Squire, 2005; Squire & Dede, 2015).

In contrast with this radical vision, research documenting metacognition as a key factor to better understand the circumstances under which processing fluency can generate a subjective experience of familiarity has revealed that several steps have to be completed for people to make familiarity-based memory decisions on the basis of fluency: (a) Participants have to understand at some general level that fluency is a cue that can be used to inform memory judgments, (b) they have to experience a feeling of fluency when processing a stimulus, and (c) they have to attribute this feeling of fluency to their memory (Jacoby, Kelley, & Dywan, 1989). In other words, fluency experiencers have to rely on metacognitive skills to decide whether fluency can be used as a

source of evidence when making a memory decision (Whittlesea & Williams, 1998). This inferential process may not necessarily come in the form of a conscious strategy. Rather, people may simply subconsciously note the occurrence of a feeling of fluency and with modest amounts of cognitive effort decide whether it is relevant to use this feeling to inform memory judgments. As evidence for this heuristic processing, studies have revealed that although people can sometimes verbalize that fluency is a cue to memory (Schwarz, 1998), these verbal reports did not appear to be related with the actual use of the fluency rule when making a decision (Geurten, Willems, & Meulemans, 2015). At any rate, according to this theory, familiarity results from the interaction between metacognition and fluency experiences that both have to be preserved for familiarity- based decisions to occur.

By manipulating fluency at the time of test through masked visual priming, numerous studies have shown that the ability to experience fluency is spared in amnesia (Conroy et al., 2005; Squire, 2004; Verfaellie & Keane, 2002). What remains a subject of debate is the extent to which attributional (metacognitive) processes are also preserved in amnesic patients. Indeed, although it is generally assumed that a decrease in the ability to engage in attributional processes accounts for the impairment in familiarity observed in amnesia (Keane et al., 2006; Verfaellie, Giovanello, & Keane, 2001), two recent studies have shown that it is possibly not so much an impairment than a change in these processes that explains amnesic patients’ pattern of results in fluency-driven recognition tests (Geurten & Willems, 2017; Ozubko & Yonelinas, 2014).

More specifically, Geurten and Willems (2017; Experiment 1) examined the influence of the introduction of an alternative source of fluency on patients’ recognition decisions by manipulating the perceptual quality of stimuli during a forced-choice recognition test. Their results revealed that healthy participants relied on the absolute level of fluency when making recognition decisions, whereas amnesic patients appeared to disqualify fluency as a cue to memory when an external source of fluency was detected. The authors suggested that patients’ underuse of fluency could result from a learned reinterpretation of fluency as a poor cue for memory rather than from a real inability to rely on it. Because of the high number of situations where fluency leads to memory errors in patients’ daily lives, the ecological validity of the correlation between fluency and past occurrence gradually decrease. Consequently, to reduce fluency-based memory errors, participants progressively learn to implement—possibly unconscious—strategies to track biasing fluency sources. Behaviorally, this leads them to rely on fluency only when they can attribute it to preexposure with a high level of confidence.

In another experiment, Geurten and Willems (2017; Experiment 2) tested the first part of their hypothesis, showing that healthy participants repeatedly exposed to evidence that perceptual-

quality-driven fluency led to memory errors started to disqualify fluency as a cue for memory, mimicking the pattern of responses demonstrated by amnesic patients. To date, however, the second part of their hypothesis—according to which patients with amnesia should be able to track alternative sources of fluency more effectively than do healthy participants—has still to be investigated.

In this context, the primary aim of the present study was to test whether differences in the ability of amnesic and healthy participants to detect alternative sources of fluency can account for differences observed in the use of fluency. To this end, patients with severe memory deficits and matched controls were recruited. The same procedure as the one used in the study by Geurten and Willems (2017; Experiment 1) was employed except that participants were presented with three forced-choice recognition tests instead of one. In each test, in addition to exposure-related fluency, an external source of fluency was provided by manipulating the perceptual quality of either the studied or the unstudied items during the test phase. To do so, we prepared three types of target−distractor pairs by combining stimuli with high and low visual quality. It has been shown that pictures with a high figure-ground contrast are perceived as clearer and easier to process than are low-contrast ones (Checkosky & Whitlock, 1973; Whittlesea, Jacoby, & Girard, 1990). Critically here, the detectability of the contrast reduction varied in each of the three recognition tests (i.e., the pictures included in the three tests were, respectively, given a 10%, 20%, or 30% contrast reduction). It is important to note that the representation of the stimuli was not manipulated in our study. Indeed, the representation of each item—created during the encoding phase—was the same in our three experimental conditions. However, we did manipulate factors that should influence the results of the attributional processes (for a recent integrative memory model presenting the distinctions and the interactions between the level of representation and the level of attribution in memory, see Bastin et al., 2019).

In a similar experiment conducted with three different samples of healthy participants, Willems and Van der Linden (2006; Experiments 1–3) found that fluency due to preexposure influenced recognition responses less when the perceptual manipulation associated with the target was obvious compared to when it was only detectable or barely noticeable. In this context, as in the studies by Geurten and Willems (2017) and Willems and Van der Linden (2006), we expected participants to produce a greater correct recognition rate for targets with higher picture quality when the picture quality manipulation remained undetected (Jacoby & Whitehouse, 1989). However, when the perceptual manipulation was detected and judged to be the principal source of the feeling of fluency, we expected participants to attribute fluency to this external source (Whittlesea & Williams, 2000). In the latter case, fluency was not expected to be used as a guide for recognition decisions. In addition, if amnestic patients truly implemented strategies to more

effectively detect alternative sources of fluency, we hypothesized that they would demonstrate reluctance to use fluency at a low level of contrast reduction (i.e., when the external source is relatively difficult to detect; i.e., 20% contrast reduction), whereas healthy patients would disqualify fluency only at a high level of contrast reduction (i.e., when the external source is easily detectable; i.e., 30% contrast reduction). Finally, if attributional processes were truly preserved in amnesia, all participants should be able to rely on fluency at a very low level of contrast reduction (i.e., when the alternative source is barely noticeable; i.e., 10% contrast reduction).

Method

Participants Eight French-speaking patients (three female) with amnesia participated in this study. They were recruited from various neuropsychological rehabilitation units in Belgium. Major attentional and executive function deficits constituted an exclusion criterion. The time since diagnosis ranged from 1 to 11 years (M = 3.88, SD = 3.48). The mean age was 37.4 (SD = 12.09) years, and the mean education level was 13.4 (SD = 2.4) years. General intellectual efficiency was estimated using the Wechsler Abbreviated Scale of Intelligence (2nd ed.; WASI-II; Wechsler & Hsiao-pin, 2011). The Wechsler Memory Scale (3rd ed.; WMS-III; Wechsler, 1997) was used to appraise patients’ working memory and episodic memory abilities. All patients showed normal intellectual functioning (IQ = 98.4, SD = 7.5) and working memory performance (working memory index = 93.25, SD = 8.5). However, they had severe episodic memory deficits (general memory index = 58, SD = 5.8; visual delay index = 58.9, SD = 7.6; and auditory delay index = 64, SD = 7.1). Patients’ characteristics are presented in Table 1.

Amnesic Patients’ Demographic and Neuropsychological Characteristics

Moreover, two healthy participants who had no history of psychiatric or neurological illness were matched with each amnesic patient for age, gender (n = 16; six female), and education level. Their ages ranged from 21 to 55 years (M = 43.2 years, SD = 12.6); they had a mean IQ of 96 (SD = 10.15) and a mean education level of 13.7 (SD = 23.6) years. The control and amnesic groups did not differ significantly in age, education, or IQ (all ps > .50).

Required sample size was determined a priori on the basis of the medium to large effects that were observed in similar studies focusing on fluency use in amnesia (e.g., Geurten & Willems, 2017). Specifically, sample size was thus set to reach a predicted power of .80 for a within– between interaction (medium effect size).

Materials As in the study by Geurten and Willems (2017), unfamiliar drawings created from abstract paintings were used as stimuli in order to limit preexperimental familiarity. Specifically, three series of 60 drawings were created and randomly assigned to one recognition test. Each of the 60 figures of the three tests was randomly assigned to Sets A and B. Half of the participants were presented with Set A as targets and Set B as distractors; the other half of the participants were presented with the reverse design.

A high-fluency and low-fluency version of each drawing was created by manipulating the figure- ground contrast quality of the figures. To do so, we used the same method as the one employed by Reber, Winkielman, and Schwarz (1998), who degraded both the picture foreground and the picture background. This manipulation has repeatedly been shown to influence processing fluency through its impact on various types of judgments inside and outside the memory domain (e.g., Reber, Schwarz, & Winkielman, 2004; Willems & Van der Linden, 2006). Specifically here, in each of the three recognition tests, the high-contrast version of the figures was always the same (i.e., white on black). However, the quality of the low-contrast version of each abstract picture varied as a function of the test. In the first test, figures were given a 10% contrast reduction so the external source of fluency was barely noticeable. In the second test, figures were given a 20% contrast- reduction so the fluency manipulation was detectable but without attracting participants’ attention (Willems & Van der Linden, 2006). In the third test, figures were given a 30% contrast reduction so the external source of fluency was clearly visible. The level of contrast manipulation used in the second test was the same as the one used by Geurten and Willems (2017).

For each of the three test phases, 30 pairs of target−distractor figures were prepared based on the

60 figures: 10 Target+/Distractor− (i.e., targets had high alternative fluency), 10 Target=/Distractor= (i.e., no alternative fluency), and 10 Target−/Distractor+ (i.e., distractors had high alternative fluency) pairs. The “+” symbol indicates that the stimulus had a high contrast (i.e., high perceptual fluency), whereas the “−” indicates that the stimulus had a low contrast (i.e., low perceptual fluency). Stimuli that were assigned to these three contrast conditions were randomly counterbalanced between subjects. Figure 1 displays some examples of stimuli used in each contrast-reduction test.

Figure 1. Examples of pairs of abstract pictures used in each contrast-reduction test (10%, 20%, and 30% contrast-reduction). The items with the reduced contrast are on the left.

Procedure The study was conducted in accordance with the ethics committee of the participating institutions. Written consent was obtained before the study began. Participants were tested individually in a quiet room. They underwent an approximatively 60-min session during which they completed three forced-choice recognition tests. These three tasks were conducted in the following order: (a) the test in which the contrast manipulation was barely noticeable (contrast reduction of 10%), (b) the test in which the contrast manipulation was detectable (contrast reduction of 20%), and (c) the test in which the contrast manipulation was visible (contrast reduction of 30%). These three tasks were completed in that specific order so that the inevitable detection of the contrast manipulation in the 30% contrast reduction test would not induce participants to look for contrast differences in the other tests. The three recognition tests were composed of two experimental phases (i.e., a study phase and a test phase) and separated by approximatively 10-min delays filled with cognitive tasks

(i.e., the subtests of the WASI-II).

Study phase

As in the study of Geurten and Willems (2017), participants were shown and told to study 30 white- on-black figures, four times each, in random order. Each study stimulus was presented in the center of the screen for 50 ms, followed by a 17-ms interval. A rapid serial visual presentation (Potter & Levy, 1969) was used to promote fluency-based recognition and eliminate the influence of declarative memory (Whittlesea, Masson, & Hughes, 2005).

Test phase

A forced-choice recognition test immediately followed the study phase. Participants were randomly presented with the 30 target−distractor pairs (10 Target+/Distractor−, 10 Target−/Distractor+, and 10 Target=/Distractor=). Both figures of each pair were presented simultaneously to each participant for 2,000 ms followed by a self-spaced interstimulus interval. The side of the screen in which the target stimulus was displayed was randomized over the trials. Participants were asked to point to the drawing they had previously seen.

Contrast detection

At the end of the experiment, participants were randomly presented with 45 target−distractor pairs of abstract pictures (i.e., 15 pairs retrieved from each recognition test) and were asked to judge which of the two pictures was of better perceptual quality. This procedure was used to examine whether patients with amnesia and healthy participants truly differed in their ability to detect alternative sources of fluency when their attention is clearly focused on the picture’s perceptual quality.

Manipulation Check To ensure that the levels of detection of the three contrast manipulations (10%, 20%, or 30%) truly differed from one another but were still sufficient for participants to develop fluency expectations, we carried out a pretest. A group of 12 participants (between 21 and 55 years of age) was randomly presented with the 90 pairs of pictures (Target+/Distractor–, Target=/Distractor=, and Target–/Distractor+) and asked to judge which of the two pictures of the pairs (if any) was of better perceptual quality. Statistical analyses revealed that high-contrast stimuli were selected in a proportion greater than chance when targets were given a 10% contrast reduction (M = .57), t(29) = 2.8, p = .015, d = 1.03; a 20% contrast reduction (M = .70), t(29) = 3.2, p < .001, d = 2.09; and a

30% contrast reduction (M = .95), t(29) = 11.28, p < .001, d = 4.70. They also revealed that the level of detection was significantly lower with a 10% contrast reduction than with a 20% contrast reduction (p = .004), which was significantly lower than with a 30% contrast reduction (p < .001). These results indicated that, when the participants’ attention was focused on the detection of perceptual differences, the level of detection of the contrast manipulation differed across the three conditions while remaining noticeable in each of them.

Results

Contrast Detection Rate A 2 (group: control or amnesic) × 3 (contrast reduction: 10%, 20%, 30%) mixed-variables analysis of variance (ANOVA) was carried out to determine whether the ability of participants to detect the perceptual manipulation differed across groups. The results revealed that the effect of contrast reduction was significant, F(2, 34) = 184.27, p < .001, ηp = .92. Specifically, the high-contrast stimuli were selected more often after a 30% contrast reduction (M = .98) than after a 20% contrast reduction (M = .71) and after a 10% contrast reduction (M = .61). No other result reached significance (Fs < 1.01).

Recognition Rate A 2 (group: control or amnesic) × 3 (contrast reduction: 10%, 20%, 30%) × 3 (target fluency: Target+/Distractor–, Target=/Distractor=, Target–/Distractor+) mixed-variables ANOVA was carried out to examine the influence of the perceptual fluency manipulation on participants’ correct recognition decisions. The group was the only between-subjects variable. The results revealed a Contrast Reduction × Target Fluency interaction, F(4, 88) = 6.74, p < .001, ηp = .23, and a Group × Contrast Reduction × Target Fluency triple interaction, F(4, 88) = 4.17, p = .004, ηp = .16. The triple interaction resulted from the fact that, in the 10% contrast reduction test (i.e., barely noticeable manipulation), both healthy participants (M = .57 vs. .43), F(1, 22) = 5.21, p = .03, ηp = .30, and patients with amnesia (M = .63 vs. .41), F(1, 22) = 6.83, p = .016, ηp = .42, produced more correct old responses when the visual manipulation induced a strong feeling of fluency (Target+/Distractor–) compared to when it induced a weak feeling of fluency (Target–/Distractor+). Conversely, in the 30% contrast reduction test (i.e., obvious manipulation), both groups gave fewer correct old responses when the competing source induced a strong feeling of fluency (Target+/Distractor–) than when it induced a weak feeling of fluency (Target–/Distractor+), M = .46 versus .65, F(1, 22) = 3.96, p = .05, ηp = .17, and M = .38 versus .65, F(1, 22) = 4.25, p = .04, ηp = .63, for controls and amnesic patients, respectively. Finally, an opposite profile was observed between our two groups after a 20% contrast reduction (i.e., detectable manipulation). Indeed, our data showed that the controls produced more correct old responses when the visual manipulation induced a strong feeling of fluency than when it induced a weak feeling of fluency (M = .60 vs. .41),

2

2

2

2

2

2 2

2

F(1, 22) = 3.79, p = .05, ηp = .19, whereas patients with amnesia seemed to give fewer correct old responses when the competing source induced a strong feeling of fluency than when it induced a weak feeling of fluency (M = .31 vs. .65), F(1, 22) = 7.04, p = .015, ηp = .63. No other result reached significance (F < 2; see Figure 2).

Figure 2. Mean proportion of old responses for targets in the three contrast reduction tests (10%, 20%, and 30%) and the quality of the three pictures for each group (control vs. amnesic

participants). Error bars display the standard deviations. T+D– = Target+/Distractor– (high-contrast target, low-contrast distractor); T=D= = Target=/Distractor= (high-contrast target, high-contrast

distractor); T–D+ = Target–/Distractor+ (low-contrast target, high-contrast distractor).

Finally, to further ensure that the contrast reduction manipulation was truly successful to enhance participants feeling of fluency, we compared whether participants truly showed a higher rate of correct recognitions for the pairs where the fluency of the target was high (Target+/Distractor–) than for the pairs where fluency was not manipulated (Target=/Distractor=), at least when the level

2

2

of contrast reduction was discreet enough not to induce a disqualification of the fluency cue. In control participants, results revealed a trend toward a higher hit rate for pairs with a high-fluency target than for pairs where the fluency was not manipulated in the 10% contrast reduction test (M = .57 vs. .52), F(1, 22) = 2.96, p = .08, ηp = .15. A higher hit rate was also found for pairs with a high-fluency target than for pairs where fluency was not manipulated in the 20% contrast reduction (M = .60 vs .50), F(1, 22) = 4.16, p = .04, ηp = .18. Similarly, in the 10% contrast reduction, amnesic patients gave more correct responses when the fluency of the target was high than when the perceptual fluency of the pairs were not manipulated (M = .63 vs. .49), F(1, 22) = 6.18, p = .02, ηp = .52. Overall, these findings confirm the validity of the fluency manipulation.

Discussion

The main goal of this experiment was to determine whether differences in how patients with amnesia and healthy controls rely on fluency can be explained by the fact that amnesic patients detect alternative sources of fluency more effectively than do healthy participants, leading them to more often disqualify fluency as a cue for memory. Our findings seem to confirm this hypothesis. Indeed, our results indicate that all participants relied on the absolute level of fluency when making recognition decisions (i.e., the higher the fluency, the higher their correct recognition rate) as long as the perceptual manipulation (i.e., contrast reduction) that served as an alternative source of fluency remained unnoticed. The main finding of the present study is that the level of contrast reduction at which the alternative source was detected differed between our groups.

Specifically, in the 10% contrast reduction test, our results revealed that both healthy participants and amnesic patients gave more correct responses on pairs where recognition of the target was facilitated by a high-contrast picture than on pairs where the processing of the distractor was facilitated. This pattern suggests that when the perceptual manipulation is sufficient to induce a feeling of fluency but inconspicuous enough not to be explicitly detected, patients with amnesia are able to rely on fluency to guide their memory decisions in the same way as do healthy participants. Many studies in which participants remain unconscious of the artificial manipulation of their processing experience have demonstrated that type of pattern in healthy participants (e.g., Jacoby & Whitehouse, 1989; Willems & Van der Linden, 2006).

In the 30% contrast reduction test, our data showed that both healthy and amnesic participants performed better on pairs where the distractor was made easier to process than on pairs where the target was made easier to process. This pattern indicates that all participants disqualified fluency as a relevant cue for memory when an external source was clearly visible. Consistent with this view, our analyses revealed that when participants were explicitly asked to compare the

2

2

2

perceptual quality of these pairs, their detection rate was nearly perfect (M = .97), suggesting that the experimental manipulation is easily detectable. It is interesting that these results can be interpreted within the discrepancy-attribution framework (Whittlesea & Williams, 2000, 2001a, 2001b; Willems & Van der Linden, 2006). According to this model, high processing fluency is interpreted as a sign of memory when the degree of fluency that is experienced is surprisingly greater than expected given the context. However, if an external source is detected that produces more fluency expectations than in past experience, even healthy participants are likely to attribute their feeling of fluency to this source rather than to the past. In recognition tests, this usually leads them to give more “yes” responses to items with a lower level of fluency.

Taken together, the results obtained in the 10% and 30% contrast reduction tests are interesting because, to our knowledge, it is the first time that, in the same experiment, a sample of patients with amnesia showed either a strong reliance or a disqualification of fluency depending on the characteristic of the test items. These findings are crucial because they could help to explain why, in previous studies, the influence of processing fluency on patients’ recognition decisions varied from large (Keane et al., 2006) to small (Verfaellie & Cermak, 1999) or even inconsistent (Levy et al., 2004) as a function of the experimental manipulation. For instance, using a subtle manipulation of fluency including one condition in which the constituent letters for studied and unstudied words were distinct (nonoverlap) and another condition in which the constituent letters for studied and unstudied words were the same (overlap), Keane et al. (2006) found a large influence of fluency on patients’ recognition judgments. Conversely, using a procedure manipulating fluency through (probably) detectable perceptual priming (83 ms), Verfaellie and Cermak (1999) found only a small effect of their manipulation on patients’ memory performance.

Finally, the results observed in the 20% contrast reduction test are particularly important because they replicated those of Geurten and Willems (2017) by showing different patterns of responses between healthy controls and patients with amnesia. Specifically, control participants performed better on pairs where the processing fluency of the target was high than on pairs where the processing fluency of the distractor was high. Conversely, in amnesic patients, poorer recognition performance was observed for pairs where the processing of the target was facilitated by higher picture quality, whereas better recognition performance was observed for pairs where the processing of the distractor was facilitated by higher picture quality. According to the discrepancy- attribution hypothesis, these findings suggest that patients with amnesia, but not controls, have detected the perceptual manipulation and judged it as the source of their feeling of fluency, leading them to disqualify fluency as a relevant memory cue. All this occurred although our analyses revealed that both patients and controls showed similar detection rates when they were explicitly asked to focus on the differences in perceptual quality between stimuli (Ms = .69 and .74 for

control and amnesic participants, respectively). These findings indicate that differences observed in the correct recognition rate between our two groups are not due to a better ability of the patients to detect the contrast manipulation per se. Indeed, all our participants were shown to be able to detect the manipulation when their attention was focused on the pictures’ perceptual quality. In this context, we hypothesize that differences in fluency use between our groups resulted from the fact that patients with amnesia could allocate resources to the detection of perceptual differences during the recognition test, leading them to more readily detect the alternative source, which remained unnoticed by control participants.

Overall, the findings of the present study seem to confirm the hypothesis of Geurten and Willems (2017) according to which patients with amnesia progressively start to track alternative sources of fluency to reduce the frequency of their fluency-based memory illusions. Specifically, given that recollection control processes are disturbed in amnesia (Bastin et al., 2004; Yonelinas et al., 1998), it is possible that amnesic patients frequently experience situations where fluency leads to memory errors in their daily life, creating the need to implement strategies to help them to decide with a high level of certainty whether their feeling of fluency results from prior exposure or from another source. This could explain why patients appear to use fluency only in a context where the external manipulation is hardly noticeable. On the other hand, healthy participants have no reason to closely track alternative sources of fluency in an attempt to compensate for impaired recollection control processes. Consequently, as in the study of Willems and Van der Linden (2006), the manipulation of the perceptual quality of the picture has to be glaringly obvious for them to disqualify fluency as a cue to memory.

Despite the relative clarity of these results, the question of whether the monitoring processes involved in the tracking of external sources of fluency are explicit−effortful or implicit−automatic still has to be investigated. Indeed, according to the cue-utilization approach of memory (Koriat, 1997, 2007), monitoring processes can sometimes occur without explicit goals and even without consciousness. To test this hypothesis, future experiments in which patients with amnesia would have to verbally report the strategies they used while completing some recognition tasks should be conducted. Another option could be to put patients in a divided-attention situation while performing our three recognition tests in order to determine whether a disqualification of fluency is still observed in the 20% contrast reduction.

Moreover, it is important to note that other metacognitive mechanisms may be suggested to account for the findings reported in the present study. Indeed, we postulate that patients with amnesia implement (implicitly or explicitly) strategies to track alternative sources of fluency to avoid memory errors. However, a pattern of responses similar to the one obtained in the present

experiment would have been observed if patients had simply set a more conservative response threshold on their global feeling of familiarity to effectively discount fluency as a diagnostic cue of information. Indeed, as the perceptual manipulation has presumably produced more fluency in the 20% contrast reduction test than in the 10% contrast reduction test, if patients changed their response criterion, the experienced fluency would logically be more likely to be disqualified in the former than in the latter test. Because patients with amnesia do not expect their impaired memory to produce a strong memory feeling, they would be more likely to reject strong as compared to weak feelings of familiarity. Within this framework, patients are not supposed to allocate more resources than do healthy participants to the tracking of alternative fluency sources but are assumed to react differently to the absolute level of fluency that is experienced. This could explain why in one study, Ozubko and Yonelinas (2014) found that amnesic patients’ recognition decisions were driven by fluency for new, but not old, items. However, because in their experiment the prime used to enhanced fluency was detectable, the hypothesis that patients had tracked the alternative source of fluency is still plausible. To truly disentangle these two hypotheses, an experimental manipulation designed to induce a strong feeling of familiarity while the external source of fluency remains undetectable should be carried out. If the changing criterion hypothesis is correct, such a manipulation would give rise to a disqualification of fluency in amnesic patients. On the reverse, if the tracking hypothesis is correct, patients with amnesia should rely on fluency to inform their recognition decisions in such a design.

There are several limitations in this study. First, the small number of patients with amnesia means that the results of our statistical analyses must be interpreted with caution. Nevertheless, the fact that, in the 20% contrast reduction test, we replicated the results of Geurten and Willems (2017) seems to speak in favor of the robustness and validity of our findings. Moreover, to determine whether our results could be generalized, it would be interesting to replicate these results in other clinical populations where severe memory problems are widespread and where, as in amnesia, fluency-based memory decisions are not shown to translate into better recognition performance (e.g., Simon, Bastin, Salmon, & Willems, 2018). In the same vein, the impact of the etiology of the amnesia could also be investigated. In this study, the recognition performance of all our patients was quite homogeneous. Of note, patients were selected to present only memory deficits. However, it could be interesting to explore whether all types of amnesic patients in more heterogeneous samples would have the same profile of results on our tests. Given the potential involvement of frontal lobes in attributional processes, it is possible that amnesic patients with head trauma or Korsakoff syndrome (i.e., who frequently show frontal damage) demonstrate more deficits in attributional processes than, for example, patients with anoxia.

A second limitation of this study is that the three recognitions tests (10%, 20%, and 30% contrast

reduction) were always presented in the same order. Although this specific procedure was selected because we wanted the fluency manipulation to remain undetected as long as possible, this confounding of test order may have influenced our results through, for example, an increase of proactive interference for the last tests. Even though the global performance of our participants was shown to remain stable across tests, which seems to rule out the possibility of an interference effect, our results should nevertheless be replicated using other types of designs. One possibility to overcome this problem could be, for example, to replace the block design used in this experiment with a between-subjects design where three groups of patients see pairs of stimuli with either a 10%, a 20%, or a 30% contrast reduction at test.

Another concern is the fact that, in the present study, participants performed mostly at chance in the control condition (Target=/Distractor=). This poses the question of whether the current results could generalize to tests in which the recognition performance is above chance. Although future experiments should be conducted to formally test this issue, some responses are already available in the literature. For instance, in studies where a counterfeit encoding was used (i.e., a procedure where participants are told that stimuli are presented in a subliminal manner at study when, in fact, there are not), participants’ performance was usually at chance on subsequent tests. Despite this, however, data have revealed similar variations in fluency effects after a counterfeit encoding than after a classic encoding condition that leads to above-chance recognition performance (e.g., Lloyd, Westerman, & Miller, 2003; Westerman, Miller, & Lloyd, 2003).

Finally, one last point to discuss is the low detectability of the contrast manipulation in the 10% contrast reduction test. This condition allowed us to confirm that, in some circumstances, patients with amnesia are able to rely on fluency to guide their memory decisions to an extent that was similar to that for healthy participants. However, because the contrast reduction of most pairs included in this condition was not detectable (i.e., correct detection rate of 57% in the pilot data), we could not determine whether patients relied on fluency in this condition because they failed to find an alternative source of fluency or because their experienced level of fluency was not high enough to prompt them to search for an alternative source.

Despite these limitations, our results could already have major implications. From a theoretical perspective, our findings could help to resolve the conceptual debate on the question of whether and when familiarity is impaired in amnesia. Specifically, our study adds to the small amount of literature showing that attributional processes—which have long been assumed to account for the emergence of familiarity (Jacoby & Dallas, 1981)—are probably not impaired in amnesia but undergo some metacognitive changes that are the product of both a decrease in the ecological validity of the fluency−memory correlations in daily life and the implementation of a more

conservative response criterion or of strategies aiming at tracking alternative sources to reduce memory errors (Geurten & Willems, 2017; Ozubko & Yonelinas, 2014). More generally, our findings emphasize the importance of looking beyond the mere behavioral pattern that is observed following a memory task in amnesia. Indeed, what could, at first sight, appear to be an impaired or abnormal test performance may actually result from subtle metacognitive changes that are very adaptive for patients’ day-to-day functioning.

References Addante, R. J., Ranganath, C., Olichney, J., & Yonelinas, A. P. (2012). Neurophysiological evidence for a recollection impairment in amnesia patients that leaves familiarity intact. Neuropsychologia, 50, 3004–3014. 10.1016/j.neuropsychologia.2012.07.038

Baddeley, A. D., & Warrington, E. K. (1970). Amnesia and the distinction between long- and short- term memory. Journal of Verbal Learning & Verbal Behavior, 9, 176–189. 10.1016/S0022- 5371(70)80048-2

Bastin, C., Besson, G., Simon, J., Delhaye, E., Geurten, M., Willems, S., & Salmon, E. (2019). An integrative memory model of recollection and familiarity to understand memory deficits. Behavioral and Brain Sciences. Advance online publication. 10.1017/S0140525X19000621

Bastin, C., Linden, M., Charnallet, A., Denby, C., Montaldi, D., Roberts, N., & Andrew, M. (2004). Dissociation between recall and recognition memory performance in an amnesic patient with hippocampal damage following carbon monoxide poisoning. Neurocase, 10, 330–344. 10.1080/13554790490507650

Checkosky, S. F., & Whitlock, D. (1973). Effects of pattern goodness on recognition time in a memory search task. Journal of Experimental Psychology, 100, 341–348. 10.1037/h0035692

Conroy, M. A., Hopkins, R. O., & Squire, L. R. (2005). On the contribution of perceptual fluency and priming to recognition memory. Cognitive, Affective & Behavioral Neuroscience, 5, 14–20. 10.3758/CABN.5.1.14

Geurten, M., & Willems, S. (2017). The learned reinterpretation of fluency in amnesia. Neuropsychologia, 101, 10–16. 10.1016/j.neuropsychologia.2017.05.012

Geurten, M., Willems, S., & Meulemans, T. (2015). Are children conservative, liberal, or

metacognitive? Preliminary evidence for the involvement of the distinctiveness heuristic in decision making. Journal of Experimental Child Psychology, 132, 230–239. 10.1016/j.jecp.2014.12.010

Jacoby, L. L., & Dallas, M. (1981). On the relationship between autobiographical memory and perceptual learning. Journal of Experimental Psychology: General, 110, 306–340. 10.1037/0096- 3445.110.3.306

Jacoby, L. L., Kelley, C. M., & Dywan, J. (1989). Memory attributions. In H. L. R. I. F. I. M.Craik (Ed.), Varieties of memory and consciousness: Essays in honour of Endel Tulving (pp. 391–422). Hillsdale, NJ: Erlbaum.

Jacoby, L. L., & Whitehouse, K. (1989). An illusion of memory: False recognition influenced by unconscious perception. Journal of Experimental Psychology: General, 118, 126–135. 10.1037/0096-3445.118.2.126

Keane, M. M., Orlando, F., & Verfaellie, M. (2006). Increasing the salience of fluency cues reduces the recognition memory impairment in amnesia. Neuropsychologia, 44, 834–839. 10.1016/j.neuropsychologia.2005.08.003

Koriat, A. (1997). Monitoring one’s own knowledge during study: A cue-utilization approach to judgments of learning. Journal of Experimental Psychology: General, 126, 349–370. 10.1037/0096-3445.126.4.349

Koriat, A. (2007). Metacognition and consciousness. In P. D.Zelazo, M.Moscovitch, & E.Thompson (Eds.), The Cambridge handbook of consciousness (pp. 289–326). 10.1017/CBO9780511816789.012

Levy, D. A., Stark, C. E. L., & Squire, L. R. (2004). Intact conceptual priming in the absence of declarative memory. Psychological Science, 15, 680–686. 10.1111/j.0956-7976.2004.00740.x

Lloyd, M. E., Westerman, D. L., & Miller, J. K. (2003). The fluency heuristic in recognition memory: The effect of repetition. Journal of Memory and Language, 48, 603–614. 10.1016/S0749- 596X(02)00535-1

Ozubko, J. D., & Yonelinas, A. P. (2014). The disruptive effects of processing fluency on familiarity- based recognition in amnesia. Neuropsychologia, 54, 59–67.

10.1016/j.neuropsychologia.2013.12.008

Potter, M. C., & Levy, E. I. (1969). Recognition memory for a rapid sequence of pictures. Journal of Experimental Psychology, 81, 10–15. 10.1037/h0027470

Reber, R., Schwarz, N., & Winkielman, P. (2004). Processing fluency and aesthetic pleasure: Is beauty in the perceiver’s processing experience?Personality and Social Psychology Review, 8, 364–382. 10.1207/s15327957pspr0804_3

Reber, R., Winkielman, P., & Schwarz, N. (1998). Effects of perceptual fluency on affective judgments. Psychological Science, 9, 45–48. 10.1111/1467-9280.00008

Schwarz, N. (1998). Accessible content and accessibility experiences: The interplay of declarative and experiential information in judgment. Personality and Social Psychology Review, 2, 87–99. 10.1207/s15327957pspr0202_2

Simon, J., Bastin, C., Salmon, E., & Willems, S. (2018). Increasing the salience of fluency cues does not reduce the recognition memory impairment in Alzheimer’s disease!Journal of Neuropsychology, 12, 216–230. 10.1111/jnp.12112

Squire, L. R. (2004). Memory systems of the brain: A brief history and current perspective. Neurobiology of Learning and Memory, 82, 171–177. 10.1016/j.nlm.2004.06.005

Squire, L. R., & Dede, A. J. O. (2015). Conscious and unconscious memory systems. Cold Spring Harbor Perspectives in Biology, 7: a021667. 10.1101/cshperspect.a021667

Squire, L. R., & Zola, S. M. (1996). Structure and function of declarative and nondeclarative memory systems. Proceedings of the National Academy of Sciences of the United States of America, 93, 13515–13522. 10.1073/pnas.93.24.13515

Verfaellie, M., & Cermak, L. S. (1999). Perceptual fluency as a cue for recognition judgments in amnesia. Neuropsychology, 13, 198–205. 10.1037/0894-4105.13.2.198

Verfaellie, M., Giovanello, K. S., & Keane, M. M. (2001). Recognition memory in amnesia: Effects of relaxing response criteria. Cognitive, Affective & Behavioral Neuroscience, 1, 3–9. 10.3758/CABN.1.1.3

Verfaellie, M., & Keane, M. M. (2002). Impaired and preserved memory processes in amnesia. In L.Squire & D. L.Schacter (Eds.), Neuropsychology of memory (pp. 35–46). New York, NY: Guilford Press.

Wechsler, D. (1997). WMS-III: Wechsler Memory Scale administration and scoring manual (3rd ed.). San Antonio, TX: Psychological Corporation.

Wechsler, D., & Hsiao-pin, C. (2011). WASI II: Wechsler Abbreviated Scale of Intelligence (2nd ed.). San Antonio, TX: Psychological Corporation.

Westerman, D. L., Miller, J. K., & Lloyd, M. E. (2003). Change in perceptual form attenuates the use of the fluency heuristic in recognition. Memory & Cognition, 31, 619–629. 10.3758/BF03196102

Whittlesea, B. W. A. (1993). Illusions of familiarity. Journal of Experimental Psychology: Learning, Memory, and Cognition, 19, 1235–1253. 10.1037/0278-7393.19.6.1235

Whittlesea, B. W. A., Jacoby, L. L., & Girard, K. (1990). Illusions of immediate memory: Evidence of an attributional basis for feelings of familiarity and perceptual quality. Journal of Memory and Language, 29, 716–732. 10.1016/0749-596X(90)90045-2

Whittlesea, B. W. A., Masson, M. E. J., & Hughes, A. D. (2005). False memory following rapidly presented lists: The element of surprise. Psychological Research, 69, 420–430. 10.1007/s00426- 005-0213-1

Whittlesea, B. W. A., & Williams, L. D. (1998). Why do strangers feel familiar, but friends don’t? A discrepancy-attribution account of feelings of familiarity. Acta Psychologica, 98, 141–165. 10.1016/S0001-6918(97)00040-1

Whittlesea, B. W. A., & Williams, L. D. (2000). The source of feelings of familiarity: The discrepancy-attribution hypothesis. Journal of Experimental Psychology: Learning, Memory, and Cognition, 26, 547–565. 10.1037/0278-7393.26.3.547

Whittlesea, B. W. A., & Williams, L. D. (2001a). The discrepancy-attribution hypothesis: I. The heuristic basis of feelings of familiarity. Journal of Experimental Psychology: Learning, Memory, and Cognition, 27, 3–13. 10.1037/0278-7393.27.1.3

Whittlesea, B. W. A., & Williams, L. D. (2001b). The discrepancy-attribution hypothesis: II. Expectation, uncertainty, surprise, and feelings of familiarity. Journal of Experimental Psychology: Learning, Memory, and Cognition, 27, 14–33. 10.1037/0278-7393.27.1.14

Willems, S., Germain, S., Salmon, E., & Van der Linden, M. (2009). Patients with Alzheimer’s disease use metamemory to attenuate the Jacoby-Whitehouse illusion. Neuropsychologia, 47, 2672–2676. 10.1016/j.neuropsychologia.2009.04.029

Willems, S., & Van der Linden, M. (2006). Mere exposure effect: A consequence of direct and indirect fluency-preference links. Consciousness and Cognition, 15, 323–341. 10.1016/j.concog.2005.06.008

Yonelinas, A. P., Kroll, N. E. A., Dobbins, I., Lazzara, M., & Knight, R. T. (1998). Recollection and familiarity deficits in amnesia: Convergence of remember-know, process dissociation, and receiver operating characteristic data. Neuropsychology, 12, 323–339. 10.1037/0894-4105.12.3.323

Submitted: January 7, 2019 Revised: April 2, 2019 Accepted: April 20, 2019

This publication is protected by US and international copyright laws and its content may not be copied without the copyright holders express written permission except for the print or download capabilities of the retrieval software used for access. This content is intended solely for the use of the individual user.

Source: Neuropsychology. Jun 20, 2019 Accession Number: 2019-33413-001 Digital Object Identifier: 10.1037/neu0000566

Mobile Site iPhone and Android apps EBSCO Support Site Privacy Policy Terms of Use Copyright

© 2019 EBSCO Industries, Inc. All rights reserved.