A simulation of older adults' associative memory deficit using structural process interference in young adults.

Associative memory deficit underlies a part of older adults' deficient episodic memory due to the reduced ability to bind units of information. In this article we further assess the mechanism underlying this deficit, by assessing the degree to which we can model it in young adults under conditions of divided attention. We shall describe two experiments in this paper; these experiments investigate item and associative recognition in young adults under full- or divided-attention conditions. The secondary tasks employed were N-back like (NBL), which serves as a working memory updating task, and parity judgement and visuospatial (VS) tasks, which serve as non-working memory tasks. The results of both experiments show that only the NBL specifically affected associative recognition, while the other tasks affected item and associative memory to the same degree, indicating a general resource competition. These results presented a convergence of evidence for the associative deficit in older adults by modelling it in young adults.

Introduction

There is a body of research on lifespan cognitive processes that specifically investigated memory performance in older adults to identify the causes of their memory decline. The research examined age-related differences in memory performance and types of information. Various studies suggested several explanations for older adults’ memory decline: a reduction in attentional resources, processing speed and failure of inhibitory processes [1-4]. While all these explanations described a deficit in memory processing, they did not fully explain the significant episodic memory deficit in older adults [5, 6]. Some studies explained that the episodic memory age decline was because they struggle to bind together a contextual feature of the event [7, 8]. Based on this explanation, the Associative-Deficit Hypothesis (ADH) framework was proposed [6]. This framework indicates that deficiency in creating and retrieving associations between different components of an episode is the primary cause for the decline in older adults’ episodic memory.

To examine this framework, researchers used a task that differentiates between item and associative information [6, 9] and thus allowed its independent estimation. Participants were required to memorize a list of pairs of items (the basic units can be two items, an item and its context, two contextual elements, or more generally the representation of two mental codes) and then to perform either an item or an associative recognition task. In the item task, participants were required to decide whether a presented item stimulus appeared in the learning phase compared to non-presented items. In the associative task, participants were required to decide whether a presented pair of two items appeared as an identical pair in the learning phase or was a recombined pair. Many studies using this paradigm supported the ADH, indicating that older adults exhibit a specific associative memory deficit, compared to young adults, while the ability to retrieve the item itself remains intact [5, 10–13]. A simple index was created to measure the level of associative memory deficit (associate deficit index, ADI). This ADI is a calculation of the difference between item recognition and associative recognition performance. In this index, item memory was used as the baseline and no association can be created without its components. A higher ADI value represented a greater associative deficit [14, 15].

One question regarding the ADH is whether it presents a new direction in aging memory research or whether it can, as previously, be described as a case of episodic memory failure due to insufficient attentional resources to process the associative task. Indeed, there are studies [1, 16] that support the idea that a reduction in attentional resources underlies older adults’ decline in episodic memory. Thus, it is reasonable to hypothesize that young adults under divided attention (DA) will be a good simulation of older adults. Yet, when testing item and associative recognition in young adults under DA, simple resource deficiency is insufficient to explain age memory decline. These studies found an additive effect in which both item and associative recognition were affected to the same degree [10-13]. These results support the ADH framework and show that it is an independent mechanism beyond general attentional resources.

While the importance of modulating older adults’ cognition in young adults is clear, to expand our understanding regarding the locus of the processes involved, a simple DA paradigm is insufficient. We suggest that using secondary tasks that are known to employ a specific mechanism is the way to simulate older adults’ memory decline in young adults.

Thus, the failure to simulate older adults’ associative deficit in young adults might stem from the secondary tasks used in the above research [11, 17, 18]. It appears that in much of the research, the secondary tasks were continuous reaction time (CRT) tasks, involving monitoring, and shallow perceptual processes or requiring simple information that already exists in the long-term memory (LTM). Therefore, these tasks served more as a mirror for the attentional cost of the primary memory task (for example [19-21]) and were developed to minimize structural interference. Here we suggest a complementary direction using a task with a specific feature that will interfere and compete with the shared process of the primary memory task. This configuration might teach us about the processes involved in the primary task.

There is a claim in the literature that a working memory (WM) deficit is the cause of an age-related decline in cognitive abilities [22, 23]. Previous studies showed that older adults have reduced WM span compared to young adults [24, 25]. Other studies found that a deficit in the efficiency or processing component in WM results in its age-related decline [26-30]. More specifically, a few studies suggested that there might be a relation between a reduction in WM resources and associative deficit in the LTM of older adults [31, 32].

Following the above, it appears that the secondary tasks used in previous research did not involve WM updating or any other demanding process that requires WM, and therefore did not interfere with the associative binding. Indeed, one recent study [33] examined whether the mediation of a reduction in WM resources explains the associative binding deficit in older adults. In those experiments, the researchers used secondary tasks that manipulated the amount of WM storage or processing resources by using math operations. Their results indicate that under DA conditions, participants obtained a differential decline in associative rather than item memory. They concluded that the kind of secondary task that is being used is important, and only a task that involves WM resources is good enough to simulate older adults’ associative deficit in young adults. Yet, their research has some limitations. First, they used a secondary task that employed simple math calculations (addition and subtraction), which yielded the same additive effect to the full-attention condition. Only division math calculations yielded significant interaction, presenting associative deficits compared to full-attention condition. Furthermore, manipulating letter strings as a WM storage (e.g., Experiment 1) also presented partial support because the most demanding condition (4-letter strings) failed to support the prediction that gradually reducing WM storage resources elicits an associative deficit. Moreover, the secondary tasks used in that research did involve WM, but did not utilize a continuous task and left an indeterminate amount of time and attention to carry out other memory-related processes [13].

The aim of the following research

The purpose of our study is to examine whether continuous WM secondary tasks can simulate older adults’ associative deficit in young adults by creating a clear distinction between the cognitive mechanisms that the tasks required. The importance of using a continuous WM secondary task is that it requires the participants to divide their attention while performing the task. Therefore, such a task does not leave the participant enough time to switch between the tasks. The innovation of our experiments is that only a secondary task that involves WM updating will create an associative deficit in young adults. Each secondary task that will involve the relevant mechanism of binding in WM will lead to an associative deficit in young adults. For this purpose, we manipulated the secondary task in the DA condition. We used three different secondary tasks: N-back like (NBL), parity judgement and visuospatial (VS), and these were performed simultaneously with the memory encoding.

To understand better which secondary task will lead to an associative deficit, a closer observation of the cognitive aspects of the assignments is required. Our modified NBL task required the participants to hold in mind three digits at a time and compare them to each other (i.e., verify if they are all odd numbers in a sequence) and update the three-digit string whenever a new digit is presented. In the standard N-back task, a participant is asked to recall only one digit at a time, depending on the interval defined (N = 1, N = 2, etc.). Thus, our modified task can be presented as a 1+2+3-back version. This task is usually referred to as a WM updating measure and requires active maintenance of WM [34]. Since a participant needs to hold three digits at a time in WM and also to compare the presented stimulus to the previous stimuli, the steps of the experiment are dependent on each other. As a result, an association is created and therefore, this task requires a binding mechanism. In the parity task, participants were shown a random number from 1 to 9 and were asked to decide whether the number was odd or even. The decision whether a number is odd or even relies on information that already exists in LTM. In the VS task, participants were shown a digit on a quadrant of a screen and were required to decide which quadrant the digit appeard in and then press the matching response key. This kind of task requires only shallow perceptual resources. The parity and VS tasks are considered to be CRT tasks. As mentioned above, the CRT task does not involve the binding mechanism in WM but rather is a shallow perceptual process and relies on retrieval from LTM.

We claim that the NBL task engaged and competed with the mechanism needed for long-term episodic binding that leads to a recollection experience, which is essential for successful associative recognition. While recollection involves executive functioning and is associated with a clear sense of remembering, item recognition relies more on familiarity based processing [5, 35, 36]. Thus, we predict that the NBL, but not parity or VS tasks, will cause a differential deficit in young adults’ memory for associative information compared to item information, although they are all attention and resource demanding.

To this end, two experiments were conducted: The first experiment compared full-attention performance for item and associative recognition to two DA conditions. The first secondary task in the DA condition was a parity judgement task, which is considered to be a CRT task, and which relies on retrieval from LTM. The second secondary task was an NBL task that is considered to be a WM task, which requires updating and therefore involves the binding mechanism. The second experiment had the same design but compared the same NBL task to a VS task. The VS task is considered to be a CRT task that relies on retrieval from LTM (see Fig 1).

Fig 1

Procedure of the secondary task presentation employed in Experiments 1 and 2.

CWMU = continuous working memory updating.

Experiment 1

Method

Participants

In the current study there were 34 psychology undergraduate students (29 women) from Achva Academic College who were rewarded for their participation with course credit, an acceptable procedure in a first-year introductory psychology academic course. The mean age and education were 22.64 (1.98) years and 13.0 (0.29) years, respectively. All the participants had normal vision and hearing abilities, as was indicated in self-reports and in their ability to report standard stimuli presented to them visually and auditorily. In addition, the participants all reported being in good health. Exclusion criteria included past/current psychiatric or neurological disorders (confirmed by the participants in a self-report) and/or a formal diagnosis of learning disabilities. The study was approved by the local institutional review board of Achva Academic College. All participants gave their written informed consent for study participation.

Design

Three independent variables were used: test (item/association), attention (full attention (FA)/divided attention (DA)) and task-group (NBL/parity). All factors except task-group were within subject variables. Memory accuracy was computed as the percentage of hits minus the percentage of false alarms. This measure equates the memory performance scale of the item and the associative recognition tests.

Materials

The stimuli were four lists of 26 pairs of words that were not related visually, auditorily or semantically. For the four lists, two versions were created, each version included all four of the lists but in one version two lists were presented in the FA condition and the other two lists were presented in the DA condition, while in the second version the list presentations were reversed. The order of the attention conditions was counterbalanced.

Procedure

Tasks were designed to create minimal structural interference in modality and stimuli. The memory tasks were presented auditory, and the secondary tasks were presented visually.

Before the experimental session, participants were given a practice session of the memory and the relevant secondary task to familiarize themselves with the procedure and with the secondary task.

Memory learning phase

The participants were tested individually in a neutral and quiet room. They heard four lists, each included 26 words pairs, at a rate of four seconds per pair, through headphones. Participants were instructed to pay attention not only to each item but also to the pairs, and were told their memory for both items and pairs would be tested.

Each list was followed by an interpolated activity of 60 sec in which the participants counted backwards in sevens from a given number, to prevent rehearsal between the learning and the test phase (recency effect). Then, after the distracting task, two memory tests, as described below (item test and association test), were presented. The order of the tests (item or associative recognition) was counterbalanced across all participants in each group, and each word appeared in only one of the tests.

Memory test phase

Yes–no item recognition test. In this test, the participants heard 16 words, one at a time, at the rate of four seconds per word. Of these, eight were target words that were selected from the studied pairs (one from each pair) and eight words were distractors. The distractors had the same characteristics as the target words, except that they had not appeared in the study phase. The participants were instructed to respond to target words with a designated “yes” response key and to press the "no" response key for a distractor word.

Associative recognition test. In this test, 16 pairs of words were presented, one at a time, at the rate of four seconds per pair, by a female voice. Eight of the words were the intact pairs from the study phase, and the other eight pairs were recombined (rearranged) pairs, that is, they consisted of words taken from different study pairs (distractors). Participants were instructed to respond to target words by pressing the “yes” response key, and the "no" response key for distractor words.

Scoring

To assess differences between the item and associative recognition tests, we computed a measure for the percentage of hits minus the percentage of false alarms (a false alarm occurs where a non-target stimulus is identified as a target [37]) for each participant and experiment. This measure of memory accuracy equated the scales for the item and the associative recognition tests, with chance level performance at 0.00 and perfect performance at 1.00. In accordance with this, the associative deficit is referred to as the difference between item recognition and associative recognition performance.

DA secondary tasks

The participants performed a secondary task during the study phase of two of the lists. This task involved a visual presentation on a computer monitor of a single-digit at the center of the screen, once every 1.5 sec, in a random order (see Fig 1).

Participants were randomly assigned to either to the NBL task-group or the parity task-group. Half of the participants performed a parity judgement secondary task, in which they were asked to judge whether the represented digit was an odd or an even number as fast as they could (by pressing a response key for “odd” or a different response key for “even”). The other half of participants received an NBL secondary task in which they were instructed to judge if there was a continuous sequence of three odd digits. In this task, participants were instructed to press on the response key "yes" whenever there was a sequence of three odd digits, otherwise the participants were instructed to press on the response key "no". While the parity task did not require the participant to maintain and update WM after each response, the NBL task required participants to keep and update their WM after each response (see Fig 1). These tasks were identical in their visual-numerical presentation and the defined pressing keys, and differed only in their instructions.

Prior to the study phase of each of the DA trials, participants were told to pay equal attention to memorizing the words and to performing the secondary digit task. Participants were told to perform the secondary task as quickly and as accurately as possible.

Results

The mean proportion of measures for hits and false alarms for each task-group and test can be seen in Table 1. To specifically address the hypothesis tested in this experiment, we computed a three-way analysis of variance (ANOVA), with 2 (task-group) X 2 (test) X 2 (attention) for the equated measure of memory accuracy (proportion of hits minus proportion of false alarms). Results (Fig 2) indicated three significant main effects. The first main effect for attention was significant (F(1, 32) = 20.92, p < 0.01, η2 = 0.39), meaning that memory performance in the FA condition (M = .57 (0.23)) was better than in the DA condition (M = .40 (0.19)). The second main effect for test was also significant (F(1, 32) = 43.52, p < 0.01, η2 = 0.57), with performance on the item test (M = .58 (0.22)) surpassing performance on the associative test (M = .40 (0.16)). The third main effect for task-group was also significant (F(1, 32) = 4.46, p < 0.01, η2 = 0.12), with performance on the parity task (M = 0.55 (0.18)) surpassing performance on the NBL task (M = .42 (0.18)).

Table 1

Memory performance; proportion of hits and false alarms (FAL) in full- and divided-attention conditions for the two experimental task-groups.

		Parity secondary task				N-Back like secondary task
		Full attention		Divided attention		Full attention		Divided attention
		M	SD	M	SD	M	SD	M	SD
Item recognition
	Hits	0.77	0.19	0.76	0.16	0.75	0.14	0.63	0.19
	FAL	0.12	0.11	0.17	0.12	0.14	0.11	0.17	0.12
	%Hit-%FAL	0.65	0.25	0.58	0.20	0.61	0.19	0.47	0.17
Associative recognition
	Hits	0.73	0.19	0.64	0.18	0.65	0.14	0.53	0.17
	FAL	0.15	0.14	0.22	0.20	0.18	0.14	0.38	0.15
	%Hit-%FAL	0.57	0.20	0.42	0.29	0.47	0.22	0.14	0.21

Fig 2

Memory performance in Experiment 1; proportion of hits minus false alarms (FAL) in full- (FA) and divided-attention (DA) conditions for the two experimental groups.

The error bars represent standard error of the mean.

The interaction between attention and test was significant (F (1, 32) = 12.90, p < 0.01, η2 = 0.29). Post-hoc comparisons on the different effect of the attention condition on item and associative recognition yielded greater decrease in the DA condition between tests compared to the FA condition (F(1, 32) = 40.49, p = 0.000, η2 = 0.55; F(1, 32) = 14.13, p = 0.000, η2 = 0.30; respectively); see Fig 2B.The interaction between test and task-group was also significant (F(1, 32) = 5.09, p = 0.030, η2 = 0).14. Post-hoc comparisons showed a difference between groups only in the associative test but not in the item test (F(1, 32) = 5.99, p = 0.020, η2 = 0.16; F(1, 32) = 1.40, p = 0.244, η2 = 0.04; respectively).

The three-way interaction used to test our hypotheses did not reach significance (F(1, 32) = 2.16, p = 0.15, η2 = 0.06). Further analysis of the interaction between attention and test were significant only for the NBL task (F(1, 32) = 12.80, p < 0.01, η2 = 0.29) but not for the parity task (F(1, 32) = 2.25, p = 0.14, ns). Planned comparison on the above significant interaction for the NBL task showed greater associative deficit for the DA condition (F(1, 32) = 37.55, p < 0.01, η2 = 0.54) than for the FA condition (F(1, 32) = 11.82, p < 0.01, η2 = 0.27); see Fig 2 and Table 1. Despite the fact that the three-way interaction was not significant, we performed the planned comparisons in order to test the simple interaction between attention and test in the NBL task, as a significant ANOVA is not a pre-condition for performing focused contrasts (see Rosnow & Rosenthal [38]).

To ensure there were no differences between the task-groups in the FA condition, we performed a 2X2 ANOVA between task-group and test in the FA condition. The results indicated no significant effect for task-group (F(1, 32) = 0.72, p = 0.40, η2 = 0.01) nor for the interaction between test and task-group (F(1, 32) = 1.31, p = 0.26, η2 = 0.04). Thus, the differences between the memory performance in the DA condition stemmed from the secondary task we used.

Discussion

The above results consistently replicate previous results. DA causes a significant decrease in memory performance [11, 12, 21, 39] and item recognition performance is superior to associative recognition performance [10–13, 39]. The advantage of these results is that they demonstrate an interaction effect between tests and attention when an NBL task was used, but not when a parity task was used. The results might suggest that a simple CRT task such as parity judgement affects item and associative recognition in an additive manner as was found in previous research, and thus presented a general resource interference. However, the NBL task, which is considered to be a WM continuous task, presents a structural interference that affects and interacts specifically with the associative mechanism. Nevertheless, this interpretation cannot exclude other interpretations of our observations. The parity task requires less effort than the NBL task, thus it is reasonable to assume that it would increase the difficulty of the task rather than create a specific structural interference. Therefore, in the next experiment, we decided to use four choice response keys rather than two and we concentrated on this issue. Experiment 2 compares the NBL task to the VS task using the same numerical stimuli but with different instructions, which results in a separate process.

Experiment 2

The purpose of the second experiment was to extend and replicate the findings of the first experiment by comparing the NBL task to a visuospatial task. Two groups performed memory tasks under FA and DA conditions. The tasks employed were the same as in Experiment 1, except for the VS task, where participants were instructed to indicate the spatial location of a number (see Fig 1). We hypothesize that while the VS secondary task would cause an additive effect on item and associative recognition, the NBL task would replicate Experiment 1’s results by presenting interaction between tests and the attention condition.

In the current study there were 32 psychology undergraduate students (23 women) from Achva Academic College who were rewarded for their participation with course credit. The mean age and education were 22.58 (1.64) years and 13.33 (0.59) years, respectively. All the participants had normal vision and hearing abilities, as was indicated in self-reports and in their ability to report standard stimuli presented to them visually and auditorily. In addition, the participants all reported being in good health. Exclusion criteria included past/current psychiatric or neurological disorders (confirmed by the participants in a self-report) and/or a formal diagnosis of learning disabilities. The study was approved by the local institutional review board of Achva Academic College. All participants gave their written informed consent for study participation.

Design & materials

The design and materials were the same as in Experiment 1 except that spatial judgement instructions replaced the parity judgement task instructions.

Procedure. The procedure was the same as in Experiment 1. Participants were randomly assigned to either the NBL task-group or the VS task-group.

Visuospatial task (VS)

Participants who were allocated to this task-group performed DA blocks in addition to the FA memory blocks. In the DA condition, which was performed simultaneously to the memory encoding, participants had to point to the spatial location of digits presented in a random order in four possible locations, using four matching response keys. This task involved a visual presentation on a computer monitor of single digits, one every 1.5 sec. the digit was presented on the screen in a random order in such a manner that each digit was presented in a quadrant of the screen. Participants had to decide in which quadrant the digit appeared and press on the matching response key (see Fig 1).

NBL task

The NBL task was the same as in Experiment 1.

Memory tasks

The memory task was the same as Experiment 1.

Note that the screen presentation stayed the same, but the instructions were changed for the spatial task, and the response keys were extended to four.

The mean proportion measures for hits and false alarms for each task-group and test can be seen in Table 2. In order to specifically address the hypothesis tested in this experiment, we computed a three-way ANOVA, with 2 (task-group) X 2 (test) X 2 (attention) for the equated measure of memory accuracy (proportion of hits minus proportion of false alarms). Results (Fig 3) indicated two significant main effects. The first main effect was for attention (F(1, 30) = 23.83, p < .01, η2 = .44) where memory performance in the full-attention condition (M = .53 (0.22)) was better than in the divided attention condition (M = .33 (0.14)). The second main effect was for test (F(1, 30) = 31.83, p < 0.01, η2 = 0.51) with performance on the item test (M = .52 (16)) surpassing performance on the associative test (M = .34 (0.17)).

Table 2

Memory performance; proportion of hits and false alarms (FAL) in full- and divided-attention conditions for the two experimental task-groups.

		Visuospatial secondary task				N-Back like secondary task
		Full attention		Divided attention		Full attention		Divided attention
		M	SD	M	SD	M	SD	M	SD
Item recognition
	Hits	0.72	0.18	0.64	0.18	0.70	0.18	0.71	0.16
	FAL	0.10	0.11	0.19	0.16	0.15	0.13	0.24	0.13
	%Hit-%FAL	0.62	0.21	0.44	0.19	0.54	0.25	0.47	0.17
Associative recognition
	Hits	0.64	0.15	0.57	0.13	0.65	0.14	0.50	0.17
	FAL	0.13	0.17	0.32	0.13	0.22	0.20	0.36	0.15
	%Hit-%FAL	0.51	0.25	0.27	0.12	0.43	0.27	0.14	0.22

Fig 3

Memory performance in Experiment 2; proportion of hits minus false alarms (FAL) in full- (FA) and divided-attention (DA) conditions for the two experimental groups.

The error bars represent standard error of the mean.

The interaction between attention and test was also significant (F(1, 30) = 7.01, p < 01, η2 = 0.19). Post-hoc comparisons on the different effect of the attention condition on item and associative recognition showed greater decrease in the DA condition between tests compared to the FA condition (F(1, 30) = 36.89, p = 0.000, η2 = 0.55; F(1, 30) = 6.84, p = 0.013, η2 = 0.18; respectively).

As in Experiment 1, the three-way interaction was not significant (F(1, 30) = 2.43, p = 0.12, ns). Further analysis yielded that the interaction between attention and test were significant only for the NBL task-group (F(1, 30) = 8.85, p < 0.01, η2 = 0.22) but not for the VS task-group (F < 1). Planned comparison on the above significant interaction for the NBL task-group showed greater associative deficit for the DA condition (F(1, 30) = 33.21, p < 0.01, η2 = 0.52) than for the FA condition (F(1, 30) = 3.58, p = 0.07, ns); see Fig 3.

As in Experiment 1, in order to ensure there were no differences between the task-groups in the FA condition, we performed a 2X2 ANOVA between task-group and test in the FA condition. The results indicated no significant effect for task-group (F(1, 30) = 1.10, p = 0.30, η2 = 0.03) nor for the interaction between test and task-group (F(1, 30) = 0.00, p = 0.99, η2 = 0.00). Thus, the differences between the memory performance in the DA condition stemmed from the secondary task we used.

Experiment 2 replicated and extended the results of Experiment 1. DA caused episodic memory deficiency, but only the NBL task made a differential effect in item and associative recognition performance compared to the FA condition.

In Experiment 1 we used a parity judgement task, but in Experiment 2 we used a VS task. The VS task compared to the parity and NBL tasks required participants to use four response keys rather than two. As such, it was more difficult, at least in terms of response production, and yet an additive effect was observed. Furthermore, taking the item recognition test as a baseline, it seems that the VS task did affect memory performance compared to the full-attention condition even more than the NBL task did (F(1, 30) = 7.23, p < 0.05, η2 = 0.19; F(1, 30) = 1.03, p = 0.32, η2 = 0.03; respectively); see Fig 3. Thus, the above result supports the specific effect of the NBL task on associative recognition.

Again, and in line with our hypothesis, only a task that involves the relevant mechanism for the creation of an association between two units of information, such as our NBL task, will lead to associative deficit.

Further analysis across experiments

To directly compare the associative deficit across experiments and under different conditions, we used the ADI. Four attention conditions were compared in a one-way ANOVA: FA–participant under full attention; DA-parity; DA-VS; DA-N-back. Results are presented in Fig 4. The main effect of the attention condition was significant (F(3, 130) = 7.24, p = 0.00, η2 = 0.14). Planned comparisons showed that the two CRT DA tasks (parity judgement and VS) caused the same associative deficit (F(1, 130) = 0.00; e.g., about 16% decrease from item to associative recognition), and both were not different from the associative deficit in the FA condition (F(1, 130) = 1.03, p = 0.31, η2 = 0.01; e.g., about 11% in the FA condition). All three were significantly lower than the associative deficit caused under the WM updating task (N-back) (F(1, 130) = 15.47, p = 0.00, η2 = 0.11; e.g., about 33% in the N-back DA condition). These results emphasize our hypothesis that only the N-back task simulates associative deficit in young adults. We can draw from this observation that only tasks that employed WM updating, but not a dual task that competes with attentional resources, can affect associative binding. This simulation suggests that older adults suffer from damage to the same cognitive process involved in WM updating.

Fig 4

Associative deficit index (ADI) in the different attention conditions across the two experiments.

FA = full attention, VS = visuopatial, **p ≤ 0.01.

General discussion

The purpose of this research was to create process interference in young adults’ episodic memory, to simulate older adults’ associative deficit. This was achieved by using a continuous WM updating secondary task that involved the relevant mechanism (NBL). We suggest that only the involvement of a mechanism that is related to episodic binding (of time, objects, location, etc.) will lead to a recollection experience, which is essential for successful associative recognition. Therefore, we predicted in our task analysis that only an NBL secondary task, which is a task that requires WM updating and involves the relevant mechanism of comparison association between stimuli, would compete with associative memory binding. This, in turn, would cause a differential deficit in young adults’ memory for association compared to item information, as obtained in older adults’ performance. In the planned comparisons that we performed in both experiments, we found that the secondary task, whether parity judgement or VS, obtained additive deficits in the item and association recognition under the DA condition compared to the FA condition. However, in the NBL task, we found greater associative deficit for the DA condition compared to the FA condition, a pattern that characterizes the associative deficit of older adults.

Our findings replicate and expand the findings in the literature. In Experiments 1 and 2 we used the DA condition in three different secondary tasks. Two tasks out of three were CRT tasks—parity judgement and VS tasks. These kinds of tasks rely on semantic retrieval from LTM. These findings were supported by other research in this area that found an additive deficit in item and associative information [10-13]. In other words, these kinds of tasks were not good enough to simulate an associative deficit in young adults as could be observed in older adults.

Additionally, our results support and expand another study that successfully simulated older adults’ associative deficit in young adults, by manipulating the amount of WM storage or processing resources [33]. Their results indicate that under DA conditions, participants obtained differential decline in associative rather than item memory. However, their findings showed only partial support for associative deficit in young adults. In our experiment, we successfully simulated older adults’ associative deficit in young adults by using an NBL task, which is usually referred to as a WM updating measure, as well as active maintenance of WM. The innovation of our experiments is that only a continuous WM secondary task that requires updating, involves the relevant mechanism of binding and will lead to associative deficit in young adults.

In the above study [33], the researchers reported that associative deficit in young adults resulted from the amount of WM storage or processing. However, in another study [26] that criticized those results, it was claimed that their results showed an ordinal interaction, since binding memory in young adults was worse in comparison to their item memory at FA, and in the DA condition it increased. In contrast, the test X DA interaction in our Experiment 2 presents a disordinal interaction, which therefore reduces such a possible claim.

One possible criticism of the above results is that the NBL task is much more demanding and thus it causes a specific associative deficit. This task-demanding hypothesis was tested in the literature and consistently presented an additive result on the test (item/association) X task-group (easy/hard) interaction [10]. To target this possible criticism directly, we moved to four response keys in a VS task in Experiment 2, compared to only two response keys in the NBL task. Furthermore, considering item performance as a baseline condition, it can be seen that comparing the deficit from full attention to divided attention, the VS task affects performance much more than the NBL task (F(1, 30) = 7.23, p < 0.05, η2 = .19; F(1, 30) = 1.03, p = 0.31, η2 = 0.03; respectively). This pattern was not found when comparing a parity task to an NBL task, in which the former did not affect item recognition and the latter was close to being significant (F(1, 32) = 1.11, p = 0.29, η2 = 0.03; F(1, 32) = 4.14, p = 0.05, η2 = 0.11; respectively). These patterns found in both experiments suggest that task demand cannot explain our results when testing its effect on a regular item recognition task.

Another alternative explanation that cannot be excluded is that both the NBL task and the creation of associations employed inhibitory processes. In the process of creating an association, one needed to bind and retrieve an A-B pair and to inhibit an A-C pair, in which all A-B-C stimuli were presented in the learning phase (unlike item recognition in which the distractors are new). In one research that targeted this hypothesis, the results showed that inhibition processes cannot simulate associative deficit and the authors concluded that associative deficit seems to be independent of other cognitive processes, including inhibitory and other resource-demanding processes [40].

Future directions

The present research succeeded in obtaining an associative deficit in young adults by performing a secondary task that requires updating. Other than the current research, there is only one piece of evidence in the literature that showed an associative deficit in young adults by using WM storage or arithmetical calculations as a secondary task, tasks that require encoding or processing [33]. The common denominator for all the tasks is the involvement of the frontal-hippocampal circuit, and therefore more replications are needed by using different methods that involve this circuit.

Another task that can be used to simulate older adults’ associative deficit in young adults is the operation span (OSPAN) task. This task is a complex WM task that requires encoding, maintenance, storage and the processing of information. When tackling this sort of task, participants were required to decide whether the presented equation is correct or not. Between the presentation of the equations, they see letters and are required to remember them in the correct order and for this, updating is required. Additionally, there is evidence in the literature that an OSPAN task involves the same brain region as a traditional neuroimaging working memory task (TNWM) like NBL. Furthermore, the OSPAN task produced robust activity in the common brain region used when carrying out the above tasks but in particular, there is greater activation of the hippocampus [41].‏ Thus it can be assumed that if we use the OSPAN task, greater associative deficit will be obtained than in a similar task to the NBL task.

In line with the same logic, and to examine the linear decline of associative deficit in young adults, it would be beneficial to manipulate the N-back task. EEG research in the domain of the NBL task has shown that the MTL/H (medial temporal lobe/hippocampal complex) is more activated during a 3-back task rather than a 1-back task [42]. We claim that only the involvement of the prefrontal-hippocampal circuit during DA led to an associative deficit in young adults. Therefore, it will be interesting to examine the influence of the 3-back versus 1-back tasks on the associative deficit. We hypothesize that during a 1-back task, the associative deficit will be very small.

Previous studies that examined divided attention in young adults reported on minimum memory deficit during retrieval compared to encoding [43, 44]. Our study simulated older adults’ associative deficit by using a dual task that required WM updating, only in the encoding phase. Therefore, it is important to investigate the associative deficit in young adults under DA conditions using a dual task at retrieval. Hence, future research needs to simulate older adults’ associative deficit in young adults by using a dual task in the test phase in parallel to, and separated from, the encoding phase. It is possible that associative deficit will be obtained not only in the encoding but also in the retrieval phase and therefore will generate a larger associative deficit. Such results, if obtained, might provide a more accurate understanding of older adults’ associative deficit.

Conclusion

The present study simulated a pattern of associative deficit in young adults. This was achieved by using secondary tasks that involved continuous working memory updating and without CRT tasks. Our interpretation is that only the involvement of the relevant mechanisms of bindings will lead to an associative deficit in young adults, similar to that in older adults. Further studies in the neuroimaging field are required to determine if the frontal-hippocampal circuit is involved when performing an NBL task and creating the associative binding. The importance of this study is that it presents a convergence of evidence for the associative deficit in older adults by modelling it in young adults.

YafitOscar-NBack like vs Parity18-9-2020 (1).

(XLS)

Click here for additional data file.

YafitOscar-NBack like vsVS-exp2-18-9-2020 (2).

(XLS)

Click here for additional data file.

25 Nov 2020

PONE-D-20-30627

A simulation of older adults’ associative memory deficit

using structural process interference in young adults

PLOS ONE

Dear Dr. Oscar-Strom,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not meet PLOS ONE’s publication criteria as it currently stands. In particular, the research question needs to be clearly motivated and discussed in the context of the relevant literature, and there are open questions regarding the methods and results. Therefore, we invite you to submit a substantially revised version of the manuscript that addresses the points raised during the review process.

Editor Comments:

Two expert reviewers assessed this work, and they largely agree that substantial revisions are required to make this manuscript acceptable for publication. Both reviewers agree that the introduction and discussion cover aspects (e.g. neural processes) that are speculative and irrelevant to the research question at hand, while omitting more relevant literature. These sections should thus be rewritten to reflect the relevant background, and appropriately motivate the current experiments. It is also important to clarify the choice of task, and whether the intention was to use a task (NBL) with a working memory (WM) component compared to the other tasks not involving WM.

The reviewers also comment that insufficient information is provided regarding participants and tasks.

The other comments mainly concern the statistical analyses, where additional information is required to aid interpretation of the results.

Please note that while novelty is not a publication criterion for PLOS ONE, any published work need to be methodologically sound, and the conclusions need to be supported by the data.

Please submit your revised manuscript by Jan 09 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Maria Wimber

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Our internal editors have looked over your manuscript and determined that it is within the scope of our Cognitive Developmental Psychology Call for Papers. The Collection will encompass a diverse range of research articles in developmental psychology, including early cognitive development, language development, atypical development, cognitive processing across the lifespan, among others, with an emphasis on transparent and reproducible reporting practices.  Additional information can be found on our announcement page: https://collections.plos.org/s/cognitive-psychology.  If you would like your manuscript to be considered for this collection, please let us know in your cover letter and we will ensure that your paper is treated as if you were responding to this call. Please note that being considered for the Collection does not require an additional peer review beyond the journal’s standard process and will not delay the publication of your manuscript if it is accepted by PLOS ONE. If you would prefer to remove your manuscript from collection consideration, please specify this in the cover letter.

3. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: No

Reviewer #2: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: No

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The paper proposed here explores the impact of divided attention at encoding on subsequent item and associative memory in young adults, across two experiments that vary on the task used to create divided attention (Experiment 1 using N-back like (NBL) task vs. parity decision and Experiment 2 using NBL task vs. a visuo-spatial decision task), with the aim to simulate the age-related associative memory deficit in young subjects. I have several major concerns about the paper in its current form, that make me lean towards recommending its rejection.

I found the introduction to be incomplete. The use of references is not specific enough, and references in general are not sufficiently provided given the wide literature existing on the topic of age-related associative memory decline. Some details that are irrelevant for the current study are provided (e.g., the associative index, that is then not used in the analyses, or brain regions associated with the cognitive mechanisms while the study does not include neuroimaging data), while I found that the details necessary to understand the rationale of the current study were lacking (i.e (p.2), why the previous studies manipulating divided attention to simulate the associative decline weren’t good enough to interfere with the specific associative binding process; this needs to be much more developed in my opinion).

The methods should give more information concerning the participants: was the sample size determined by a-priori power analysis? Were the participants screened for any neurological or psychiatric history? As participants were divided into two groups, were the groups matched in terms of age and education? Was the study approved by an Ethics Committee?

Concerning the methods still, I am not convinced by the choice of the secondary tasks and what they allow to show evidence for, and I did not find the information put clearly in the introduction or hypotheses. From my understanding, only the NBL task requires working memory, while the two other secondary tasks to which it is compared do not; they are parity and visuo-spatial judgment tasks. So, if I understood correctly, the task comparison implies working memory task vs. visuo-spatial or parity decision, non-working memory tasks. I think this should be put very clearly to make it clear to the reader what are the stakes of such comparison, and what they are not.

I have one problem with the results where the attention (full or divided; within subject factor) and task (NBL or alternative secondary task; across subjects factor) variables are analyzed in a factorial design. Caution should be given to indicate whether there is, or not, a difference between groups in the full attention condition (e.g., according to the attention x task interaction). Indeed, one would expect to observe no “group-task” difference in the full attention condition, as there is no secondary task to perform in that condition. Showing this absence of difference would help ensuring that the effect of task is indeed caused by the actual secondary task, when there is one, and not group difference in “baseline” (full attention) memory. Yet, these results are not reported, but there is a main effect of task, and thus, of group.

In addition, I have a great concern about the discussion part, which I found highly speculative in several aspects. Among my concerns are that many of the conclusions drawn are not supported by the data as they imply brain regions, but also are extended to other type of populations than older adults whose associative decline was simulated, such as patients with post-traumatic stress disorder.

Finally, I found that the paper would highly benefit from some English editing.

Reviewer #2: With two behavioral experiments, the present study compared item and associative memory performance for word pairs encoded with full vs. divided attention. In each experiment, participants encoded half of the stimuli with full attention, and the other half of the stimuli while also performing one of two secondary tasks (N-back like task and parity task in experiment 1, N-back like task and visuospatial task in experiment 2). Findings indicated that the N-back like task affected associative memory to a larger extent than item memory. By contrast, the other two tasks showed comparable influences on item and associative memory performance. I find these findings are interesting, although a few questions and comments need to be addressed.

1. From an overall point of view, the current study provides behavioral evidence for the shared cognitive processes recruited by NBL and associative encoding. There is a decently sized discussion of the potential neural mechanisms contributing to these processes in both the intro and discussion, which I’m not sure are clearly tied to the predictions/results, given that this is a behavioral study, and no direct comparison was made between the neural correlates of NBL and associative encoding. I would suggest the authors to focus more on the implications of the findings from a cognitive perspective.

Methods:

2. Were the participants randomly assigned to the groups with different secondary tasks?

3. Please clarify the instruction for the FA condition.

Results:

4. Across the two experiments, the group x test x attention interaction term from the initial 2 (task) x 2 (test) x 2 (attention) ANOVA was not significant. Therefore, it is not justified to examine the attention x test interaction for each group.

5. In p9, paragraph 1: the author mentioned larger associative deficit in the DA condition than the FA condition. Was the deficit measured as ADI? Please clarify the operational definition of associative deficit in the method section. Also, it would be informative to show whether the associative deficit (relative to the FA condition) was larger in the NBL task compared to the other two tasks.

6. Table 1&2: It would be helpful to add one row for each memory type to present the hits-FAs. It seems that in the FA condition, the group with NBL task in both experiments had lower associative memory performance compared to the other groups. Have the authors compared memory performance between the task groups? Is it possible that poor associative memory in the NBL group drove the significant attention x test interaction in this group?

Discussion:

7. p13, paragraph 2: Because this is a behavioral study on young adults only, it is difficult to draw a conclusion that circuit brain areas that are involved in the NBL tasks are impaired in older adults. Is there any behavioral evidence linking performance on N-back task to the performance of associative memory?

8. p14, paragraph 2: I’m not sure if the findings from PTSD and ASD patients are relevant, given that the main focus of the current study was to simulate age-related associative deficit in young adults.

9. p14, paragraph 3: Could the authors further discuss the study of Hara et al. (2015)? I think this study provide consistent findings with the current study that WM processing is a key component contributing to associative encoding.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Emma Delhaye

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

20 Feb 2021

The manuscript has been revised. we have addressed the reviewers comments in the response to reviewers document.

Submitted filename: response to reviewers 9.2.docx

Click here for additional data file.

10 Mar 2021

PONE-D-20-30627R1

A simulation of older adults’ associative memory deficit

using structural process interference in young adults

PLOS ONE

Dear Dr. Oscar-Strom,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not yet meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

The two original reviewers of this manuscript have now assessed the revision. As you can see from their comments, Reviewer #1 still has substantial concerns that require major revisions. Comments of Reviewer #2 are more minor in nature, but still require additional revisions. To summarize,

- Both experts agree that there are results reported in the Discussion section, which need to be moved to the Results section.

- Some theoretical arguments regarding the relationship between WM, LTM and the hypothesized binding mechanism need to be clarified.

- There are speculations about the role of MTL and mPFC in supporting memory retrieval that do not relate to the central aims and results of this study. The relationship should either be made explicity and very clear, or these sections should be removed.

- There are also further requests where methods are unclear, which need to be addressed.

- The use of the term "associative deficits" and other terms needs to be clarified and kept consistent across the manuscript.

- Finally, it is strongly recommended that the authors have their manuscript language-edited by a native English speaker to enhance intelligibility.

Please submit your revised manuscript by Apr 24 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

We look forward to receiving your revised manuscript.

Kind regards,

Maria Wimber

Academic Editor

PLOS ONE

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: (No Response)

Reviewer #2: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: (No Response)

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: (No Response)

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: (No Response)

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: No

Reviewer #2: (No Response)

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: I reviewed the manuscript of the paper « A simulation of older adults’ associative memory deficit using structural process interference in young adults”. This paper intends to simulate the age-related associative memory deficit in young adults through divided attention during encoding, with the hypothesis that such deficit can be simulated specifically by a working memory task requiring memory updating, but not using other non-working memory tasks. The results support this hypothesis as the associative memory decline in young adults is specifically shown following the working memory N-back like while it is not the case following two other non-working memory tasks varying in terms of resource demands.

While I think that the study is interesting and could make a nice contribution to the literature, I have several concerns about the manuscript.

First, I really think that the manuscript would highly benefit from English editing as many sentences seem grammatically incorrect, which makes the reading laborious.

Introduction

p.6, lines 139-141: “As mentioned above, the CRT task does not involve the binding mechanism in WM but rather relies on retrieval from the LTM” (same idea also appear on p. 8, lines 168-170). Does this imply that the NBL task involves the binding mechanism in WM? And if yes, in what way exactly? Also, it is not clear to me in what way both CRT tasks rely on the retrieval from LTM. Could the authors be more precise on that point?

Methods

p. 9: it is not clear to me if there was a memory test after each list, or if the memory test took place after the 4th list. If it took place after the 4th list, it is not clear to me why each list was followed by a distracting phase. Also, did the memory test mix items from the FA and from the DA conditions?

p. 10, lines 228-231: “Of these, eight were target words which were selected from the studied pairs (1 from each pair) and eight words were distractors. The distractors were eighteen words with the same characteristics as the target words, except that they had not appeared in the study phase.” Was it eight or eighteen distractors?

Results

Please homogenize the variable names, as they are sometimes called “group” and sometimes called “task”, sometimes “task-group”.

Discussion

p. 20-21: I think results reports would fit better in the result section than in the discussion part

Reviewer #2: The authors have addressed most of my prior questions. Now I only have some minor comments for the revised manuscript.

First, please incorporate the results about ADI (in discussion) into the results section.

Second, the authors mentioned the interaction of MTL/H and mPFC in supporting memory retrieval in the discussion. Is there any evidence suggesting mPFC is involved in NBL-like working memory? Again, I think the discussion of the specific brain regions is not highly relevant to the present study.

Third, as the authors described now, associative deficit is defined as the decreased performance from item memory to associative memory, which is different from the original definition by Naveh-Benjamin (2000). If the authors stick to the current definition (which means AD is present whenever an effect of test is identified), then please make sure to use this term properly throughout the manuscript. For example, please clarify the sentence in line 614: previous associative deficit studies reported on minimum memory deficit during retrieval compared to encoding.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

1 May 2021

Please see our responses for the reviewers in the responses file

Submitted filename: respones R2 24.4 finalSD.docx

Click here for additional data file.

25 May 2021

PONE-D-20-30627R2

A simulation of older adults’ associative memory deficit

using structural process interference in younger adults

PLOS ONE

Dear Dr. Oscar-Strom,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

The manuscript has now been re-reviewed by both original reviewers. While Reviewer #2 feels their concerns are now sufficiently addressed, Reviewer #1 still has substantial concerns regarding the readibility of the introduction, the statistics and figures reported in the results section, and parts of the discussion where some conclusions do not seem supported by the results. All of these points need to be thoroughly addressed in the next revision.

Please submit your revised manuscript by Jul 09 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Maria Wimber

Academic Editor

PLOS ONE

Additional Editor Comments (if provided):

The manuscript has now been re-reviewed by both original reviewers. While Reviewer #2 feels their concerns are now sufficiently addressed, Reviewer #1 still has substantial concerns regarding the readibility of the introduction, the statistics and figures reported in the results section, and parts of the discussion where some conclusions do not seem supported by the results. All of these points need to be thoroughly addressed in the next revision.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: (No Response)

Reviewer #2: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: (No Response)

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: No

Reviewer #2: (No Response)

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: (No Response)

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: No

Reviewer #2: (No Response)

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: I acknowledge that the authors did a good amount of work in revising the manuscript according to the concerns that I raised. While I see an improvement in the quality of the manuscript, I still have several major concerns preventing publication.

- The introduction still needs to be substantially revised for the sake of clarity and smoothness. I still had trouble reading it, and I found the reading highly laborious. Here are a few examples of the kind of things that lack clarity, but the revision needs to be extended to the entire introduction section:

o P.2: «Various studies suggested several explanations to older adults’ memory decline: a reduction in attention, processing resources and speed [1-3], a decline in memory performance…» : the authors state that older adults’ memory decline is explained by a decline in memory performance.

o P. 4: «Failures to simulate older adults’ associative-deficit in younger adults stems from the secondary tasks used in the research [10, 16-17]» : it is not clear if this is a result shown in the literature, or is the hypothesis that the authors wish to test, or is it a hypothesis suggested by other authors?

o P. 4: «Additionally, these results support the possible validity of an independent associative mechanism» : it is not clear what results this refers to

o P. 5: «Moreover, the secondary tasks used in the this research do involve WM, but not continuous WM. Namely, each step in the experiment is independent of the other [12].»: it is not clear why and how this is important, while it should be clear at this stage of the introduction

- Result section:

o P. 11: Could the authors detail the effects of the interactions that are significant, using post-hocs? (attention x test & attention x task-group)

o P. 11: «Further analysis of the interaction between attention and test were significant only for the NBL task (F (1, 32) = 12.80, p < .01, η2 = .29) but not for the parity task (F (1, 32) = 2.25, p = 0.14, ns.)» : in my understanding, this is actually a further analysis of the triple interaction, which is non-significant

o P.16: could the authors detail the direction of the significant interaction using post-hocs? (attention x test)

o Intermediate discussions p. 13 and p.18: caution in the interpretation of the results should be taken, as the triple interactions were not significant

- Overall, in the interpretation of the results in the discussion, I think caution should be taken in sentences like, e.g., «The present research succeeded in obtaining an associative-deficit in younger adults by performing a secondary task that requires updating» as the triple interactions were non-significant in both experiments

- Figure 2 and 3: precise in the legend to which experiment each figure refers to. Also, in Figure 2, please homogenize the graphs so that we can visually compare performance across tasks

Reviewer #2: (No Response)

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

7 Jul 2021

The responses are in the responses to reviewers file.

Submitted filename: responses R2 7.7.21.docx

Click here for additional data file.

20 Jul 2021

PONE-D-20-30627R3

A simulation of older adults’ associative memory deficit

using structural process interference in young adults

PLOS ONE

Dear Dr. Oscar-Strom,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

This second revision of the manuscript was now reviewed by the editor, due to unavailability of the original reviewer ( Reviewer #1 who still had substantial concerns). The authors addressed most of the concerns raised by Reviewer #1, including the statistical analyses, and a restructuring and clarification of many points in the introduction. Before this study can be accepted for publication, the following revisions need to be undertaken.

(1) There are still many grammatical errors throughout the manuscript, in particular in the introduction and abstract, which should be corrected to improve the quality and readability of the manuscript. These sections should also avoid vague language like “it seems” and “it appears”, and instead clearly state what can be inferred based on the existing literature.

(2) The figure legends should be expanded to specify exactly what is represented by the bar graphs and error bars (e.g., mean +/- SEM). They should also contain an explanation of the abbreviations (FA and DA) in the legends themselves. Finally, text in the figures should be corrected for typos (e.g., “visuospatial” instead of “visuospatial” in Fig. 3.

Please submit your revised manuscript by Sep 03 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Maria Wimber

Academic Editor

PLOS ONE

Journal Requirements:

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

Additional Editor Comments (if provided):

(see above)

[Note: HTML markup is below. Please do not edit.]

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

28 Sep 2021

attached a response to reviewers file

Submitted filename: response to reviewers 03.09.docx

Click here for additional data file.

1 Oct 2021

A simulation of older adults’ associative memory deficit

using structural process interference in young adults

PONE-D-20-30627R4

Dear Dr. Oscar-Strom,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Maria Wimber

Academic Editor

PLOS ONE

Additional Editor Comments (optional): The remaining editorial comments have been addressed.

Reviewers' comments: There were no further comments from the reviewers.

26 Oct 2021

PONE-D-20-30627R4

A simulation of older adults’ associative memory deficit using structural process interference in young adults

Dear Dr. Oscar-Strom:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Maria Wimber

Academic Editor

PLOS ONE