The effects of encoding instruction and opportunity on the recollection of behaviourally relevant events

Abstract

How does attending to a brief, behaviourally relevant stimulus affect episodic memory encoding? In the attentional boost effect, increasing attention to a brief target in a detection task boosts memory for items that are presented at the same time (relative to distractor-paired items). Although the memory advantage for target-paired items is well established, the effects of attending to targets on other aspects of episodic memory encoding are unclear. This study examined the effects of target detection and goal-directed attention on memory for task-irrelevant information from a single event, focusing on the contributions of recollection and familiarity during recognition. In Experiment 1, participants viewed a series of briefly presented faces as they performed a detection task on unrelated squares, pressing the space bar only when the square was a target colour (e.g., blue) rather than a distractor colour (e.g., orange). Half of the participants were told to memorise the faces, and half were told to ignore them. Results indicated that both recollection and familiarity were greater for target-paired faces than for distractor-paired faces, regardless of whether the faces were intentionally encoded. Experiment 2 examined whether these effects are present for single events, replicating the recollection benefit when encoding time is sufficient. Attending to behaviourally relevant targets appears to facilitate both intentional and incidental memory for the background item and the context in which it occurred, boosting subsequent recollection as well as familiarity.

Keywords

Dual-task interference recollection attention episodic memory attentional boost effect

Attention can significantly influence the ability to successfully remember events. Intentionally memorising an image or word, for example, results in better memory for that item than if the item had been incidentally encoded (Craik & Lockhart, 1972; Tulving & Thompson, 1973). However, the effects of attention on encoding are not always as expected. Although dividing attention across tasks or stimuli often impairs memory (dual-task interference; for example, Kinchla, 1992; Mulligan, 1998), increasing attention to a target in a detection task facilitates memory for concurrently presented pictures and words (referred to here as items). This finding, known as the attentional boost effect, suggests that attending to a behaviourally relevant stimuli, such as a target in a detection task, boosts memory for other, unrelated information presented at that time (Swallow & Jiang, 2010). The attentional boost effect has been widely replicated and occurs for a variety of tasks and stimuli (see Swallow & Jiang, 2013, for review). However, the nature of the attentional boost effect remains unclear. Some accounts suggest it reflects the prioritisation of behaviourally relevant moments (temporal selection), independent of goal-directed or spatial selection (Swallow & Jiang, 2013). Other accounts suggest that it reflects an increase in item-related processes (Spataro et al., 2017). The present study focuses on whether the benefits of target detection extend to other information encountered during the moment when a target appears, thereby boosting memory for the target’s context regardless of its relevance for the encoding task. It therefore addresses whether target detection during encoding increases quantitative estimates of how well a person remembers an item’s context during retrieval, whether the effect of target detection on subsequent memory relies on the intention to encode the background item, and whether the effect of target detection on context occurs within a single event. Because context encoding takes time (Fanselow, 1990; Hirshman et al., 2002; Malmberg & Shiffrin, 2005), we also address whether these effects are limited by the opportunity (or the total amount of time available over item presentations) to encode the item’s context. To address these questions, this study used a well-characterised quantitative model to examine which aspects of episodic memory are boosted by target detection under both intentional and incidental encoding conditions.

Context memory in the attentional boost effect

Event memory encompasses a studied item as well memory for the context in which that item was encountered (Davachi, 2006). Context can include information about when the item was encoded, the item’s location, the features and spatial configuration of the environment in which the item was encoded, inter-item temporal and semantic associations, and perceptual features of the item that are unrelated to its semantic representation (see Yonelinas et al., 2010, for review). One way to distinguish between memory for an item and memory for its context has been captured by the constructs of familiarity and recollection in recognition memory (Tulving & Murray, 1985). Familiarity reflects the feeling of having encountered an item previously. Recollection, in contrast, reflects retrieved associations between the study item and aspects of the original event during which the item was encountered.

Studies on the attentional boost effect have demonstrated that it enhances memory for items that are paired with an unrelated target during encoding, relative to those presented on their own or with a distractor (Swallow & Jiang, 2014). What is less clear, however, is whether it also facilitates memory for the context in which those items occurred. In studies examining the attentional boost effect on word memory, Mulligan et al. (2016) found that participants showed no memory advantage for the sensory modality, font, colour, or presentation order of words that appeared with a target rather than a distractor during encoding. Spataro et al. (2017) found no target-related advantage for a task sensitive to memory for the semantic relationships between encoded words. However, target detection could facilitate other aspects of context memory. Two studies indicated that target detection enhances relational memory for the target cue and a concurrently presented scene (Swallow & Atir, 2018; Turker & Swallow, 2018). Other data suggest that scene orientation memory, memory for the colour of a square in a particular location, and scene location are enhanced in the attentional boost effect (Leclercq et al., 2014; Makovski et al., 2011; Swallow & Jiang, 2010). Data from experiments using the Remember-Know procedure are also consistent with the suggestion that context memory is enhanced, though the effects were small (Leclercq et al., 2014; Meng et al., 2019, but see Mulligan et al., 2016). These data suggest that target detection enhances at least some aspects of context memory as well as memory for the item itself.

However, these studies also offer an incomplete picture of context memory. The Remember-Know procedure requires participants to accurately report their subjective experience of whether they remember contextual details, and depends on terminology and instruction (Geraci & McCabe, 2006; McCabe & Geraci, 2009). In contrast, tests of relational and feature memory risk missing memory for contextual information that was not directly tested (sometimes referred to as “noncriterial recollection”; Parks, 2007). These direct tests could also reflect familiarity if the items are unitised (see Yonelinas et al., 2010). In fact, unitisation is more likely to occur when items are presented multiple times, as it becomes more difficult to differentiate context information for each individual presentation (Diana et al., 2008; Giovanello et al., 2006). Therefore, it is important to characterise the effects of target detection on context memory using additional, convergent methods. Here, we used quantitative models to estimate recollection and familiarity.

Clarifying the effects of target detection on context memory will provide critical insight into how attending to behaviourally relevant stimuli influences encoding. Current accounts of the attentional boost effect agree that it reflects processes that are engaged during the initial encoding of an image or word into memory (Mulligan et al., 2016; Swallow & Jiang, 2013). However, they differ in their claims about whether this benefit is confined to the item being encoded. Based on their recent data, Spataro, Mulligan, and colleagues suggest that target detection facilitates the encoding of items and has little effect on memory for context (Mulligan et al., 2016; Spataro et al., 2017). This suggests that the attentional boost effect should be observed mainly in measures of familiarity, which reflect item memory strength, rather than recollection, which reflects memory for contextual details of the episode in which the item was encountered (Yonelinas et al., 2010).¹

However, other accounts suggest that the detection of a target triggers a temporal selection mechanism that broadly enhances encoding for the moment the target appeared, while selection based on features, locations, or goals is proposed to modulate encoding through other, already well-characterised means (e.g., Desimone & Duncan, 1995). Thus, target detection should have relatively broad effects on encoding that could include task-irrelevant aspects of the target’s context (Swallow & Jiang, 2013, 2014) boosting both recollection and familiarity for concurrently presented items, regardless of their goal relevance.

This account of the attentional boost effect also emphasises the importance of examining whether the effects of target detection on encoding include task-irrelevant information about a single event. In the standard version of this paradigm, the background item is part of the context in which a particular target or distractor appears and it is often presented multiple times. It may be remembered because it is part of the event or because participants are directing attention to it. Under intentional encoding instructions, participants should intentionally process the item in ways that are likely to facilitate later memory (e.g., processing it in a way that distinguishes the item from other items; Hunt & McDaniel, 1993). However, when participants are told to ignore the background item as they perform the detection task, the item becomes a task-irrelevant feature of the context in which a target or distractor appeared. If target detection facilitates context memory, then we might also expect to find a benefit for incidentally encoded background items. However, findings from previous studies have been mixed (e.g., Dewald et al., 2011, 2013; Swallow & Jiang, 2011, 2014; Walker et al., 2017).

If the effects of target detection on context memory reflect the selection of behaviourally relevant moments, then they may be present after a single presentation. However, most studies on the attentional boost effect that demonstrate context memory benefits involve multiple presentations of the background images. There are two main consequences of these repeated pairings: (1) The item is presented for more time in total, giving participants greater opportunity to form context representations (Hirshman et al., 2002; Malmberg & Shiffrin, 2005), and (2) the item is paired with a target or distractor cue multiple times, allowing participants to encode multiple instances of the item. In addition, because participants are sensitive to multiple presentations (Hasher & Zacks, 1988; Thunell & Thorpe, 2019), they could become aware of the consistent pairing of the item with the cue and use this to aid performance in the target detection task. Alternatively, it is possible that target detection has effects that are similar to increasing the study duration of the items, which increases both recollection and familiarity (Hockley, 1992; Kelley & Wixted, 2001). Interestingly, increasing study duration diminishes the magnitude of the attentional boost effect, suggesting that there may be some overlap in the benefits resulting from target detection and those resulting from having a longer period of time to encode items into memory (Mulligan & Spataro, 2015; Spataro et al., 2017). The nature of those benefits, however, has yet to be defined.

Estimating recollection and familiarity with the receiver operating characteristic

In this study, we estimated the contributions of recollection and familiarity to the recognition of target-paired and distractor-paired items by computing receiver operating characteristics (ROCs) and applying the dual-process signal detection (DPSD) model (Yonelinas, 1994). According to the DPSD model, recognition memory reflects the combination of a continuous memory strength signal (item familiarity) and a separate, threshold recollection signal that reflects memory for context. Recollection and familiarity are derived by fitting the equations described in the DPSD model to ROCs, which capture how hit rates (i.e., the proportion of old items correctly identified as “old”) and false alarm rates (i.e., the proportion of new items incorrectly identified as “old”) change as participants relax their criterion for indicating that an item is old. Although more complex models exist (e.g., Onyper et al., 2010), this study used the simpler DPSD model as it has been validated and used in a variety of studies on the effects of attention on episodic memory (Chan & McDermott, 2007; Diana et al., 2006; Ghetti & Angelini, 2008). Behavioural and neurophysiological evidence also indicate that recollection and familiarity can capture differences in the activation of separate neural systems during recognition (Kahn et al., 2004; Rugg & Curran, 2007). Finally, the DPSD model does not rely on participants’ interpretation of Remember-Know instructions or their ability to accurately judge whether they recollected an item. It can also estimate recollection even when participants fail to retrieve a particular feature of the encoding context (as in noncriterial recollection; Yonelinas et al., 2010).

To summarise, two accounts of the effects of attending to targets on episodic memory make competing predictions about whether the attentional boost effect will emerge in recollection. According to an item-only account of the attentional boost effect, attending to a target in this task enhances only item memory, increasing the feeling of having encountered the item previously (i.e., familiarity). In contrast, a context account of the attentional boost effect predicts that attending to a target in this task also enhances memory for other, contextual aspects of the moment in which the target and item occurred (i.e., recollection and familiarity). It further suggests that any effects of target detection on recollection estimates should be present under incidental encoding conditions (Kishiyama & Yonelinas, 2003) and when items are presented just one time.

Experiment 1

Estimating the contributions of recollection and familiarity to the attentional boost effect under incidental and intentional encoding conditions could provide additional insight into the effects of target detection on task-irrelevant information. Three outcomes are consistent with different accounts of the attentional boost effect. First, relative to intentional encoding, incidental encoding reduces familiarity more than recollection (Chalfonte, 1996; Hirst & Volpe, 1984; Kishiyama & Yonelinas, 2003). If the effects of target detection on encoding are broad and include context information, then it should influence processing under both incidental and intentional encoding instructions, and increase recollection estimates in both cases. However, a target-related boost to familiarity may be reduced under incidental encoding (Kishiyama & Yonelinas, 2003). An alternative possibility is that target detection simply allows participants to process the items in a manner that is similar to encoding under full attention. Because single-task encoding is associated with greater recollection and familiarity than dual-task encoding (Gardiner & Parkin, 1990; Mangels et al., 2001; Yonelinas, 2001), this possibility predicts that any effects of target detection on recollection or familiarity should be reduced when encoding is incidental rather than intentional. To test whether target detection influences recollection and familiarity when the item is task irrelevant, we asked participants to perform a target detection task on coloured squares while images of faces were presented in the background. Half of the participants were instructed to ignore the background faces, whereas the other half were instructed to remember them. We then fit the DPSD model to old/new confidence responses obtained from a recognition memory test, with hit rates computed separately for target- and distractor-paired items.

Method

Participants

Participants were recruited from the Cornell University community and received course credit. They had normal or corrected-to-normal vision. Normal colour vision was verified with the Hardy, Rand & Ritler pseudoisochromatic colour blindness test (Richmond Products, Albuquerque, NM, USA). Experimental procedures were reviewed and approved by Cornell’s Institutional Review Board. All participants provided informed consent and were debriefed about the purpose of the study upon completion.

A target sample size of 64 was selected a priori using G*Power 3.1 (Faul et al., 2007) and estimates of effect sizes based on previous studies of temporal selection in episodic memory encoding (Turker & Swallow, 2018). With α < .05, this sample size resulted in a power (1 − β) of .8 to detect an effect of d ⩾ 0.25 in a paired-samples t test. Data from three participants whose performance on the detection task did not reach a priori inclusion criteria (target hit rate >.8 and false alarm rate <.1) were excluded and replaced, resulting in a final sample of 64 participants (21 male, 43 female, 18–23 years old, age M = 19.27, SD = 1.2).

Study materials

Participants sat unconstrained in normally lit interior room, approximately 57 cm away from a ViewSonic E70fB 17″ CRT monitor (1024 × 768 pixels, 75 Hz refresh rate) controlled by a Dell PC and responded on a keyboard. All experiments were programmed in MATLAB (MathWorks, LLC, Apple Hill Dr., Natick, MA) using Psychtoolbox-3 (Brainard, 1997; Pelli, 1997).

A total of 120 face images (256 pixels height) were randomly selected from the Bainbridge 10K U.S. Adult Faces Database (Bainbridge et al., 2013). Faces were divided into 20 faces for use during the practice task, 50 old faces that were presented during the target detection and recognition tasks, and 50 new faces that were presented only during the recognition task. For each participant, old faces were randomly assigned to appear with targets or distractors. Because previous work suggests that item-level variations among the background images such as attractiveness, gender, desirability, and emotional valence do not influence the attentional boost effect (Rossi-Arnaud et al., 2018; Swallow & Atir, 2018), face attributes were not controlled. Randomisation of the faces across target and distractor conditions resulted in each face appearing with each square type at least 22 times (M = 32, SD = 3.8, across 64 participants). Face masks were created by dividing the image into 256 equally sized squares and shuffling their locations. Squares for the detection task (coloured squares) were drawn in Psychtoolbox-3 and were either blue (RGB [0 0 255]) or orange (RGB [255 127 0]) in colour.

Procedure

The experiment was performed in two parts. Participants first completed the encoding and detection task (Figure 1). Each trial (1,000 ms long, 0 ms inter-trial interval) consisted of a face located at the centre of the screen (7.5° × 5.5° in size; 500 ms duration), a blue or orange square that appeared at the centre of the face (1.4° × 1.4° in size; 100 ms duration; face-square onset asynchrony 0 ms), and a mask that immediately replaced the face at the centre of the screen (7.5° × 5.5° in size; face-mask inter-stimulus interval 0 ms; 500 ms mask duration). All participants pressed the space bar as quickly as possible whenever the square was in a predefined target colour (counterbalanced across participants) but not when it was in the other, distractor colour. For each participant, half of the faces were randomly assigned to always appear with a target coloured square, and the other half to appear with a distractor coloured square (encoding condition). Each face-square type pair was presented 8 times, resulting in 400 trials evenly divided among target and distractor squares. All button presses and response times (RTs) were recorded.

Figure 1.

Items presented during the encoding period. Participants were instructed to press the space bar on a keyboard when they saw a square in the target colour (either blue or orange), and to withhold responses for distractor squares (either blue or orange depending on target colour). Half of the participants were instructed to remember the faces presented in the background, and the other half were instructed to ignore the background faces and focus on the target search task.

To examine the interaction of goal-directed attention to the faces and the selection of behaviourally relevant events, encoding instructions were manipulated across participants. Half of the participants (N = 32) were instructed to memorise all of the faces for a later memory test (intentional encoding). The other half of the participants (N = 32) were told that their only task was to respond to the target coloured squares as quickly as possible, and to ignore the faces because they could interfere with the detection task (incidental encoding).

During the second part of the experiment, participants performed an old/new recognition task on the faces. On each trial, a single face was presented in the centre of the screen, and participants pressed the “b” key to indicate they believed the face was old or the “n” key to indicate they believed it was new. Participants then numerically rated their confidence in their response on a scale from 1 to 7 (1 = unsure, 7 = very sure). Participants were given 10 s to make a response. All 50 old faces (25 target-paired and 25 distractor-paired) and all 50 new faces were presented during the recognition test, for a total of 100 trials. All button presses and RTs were recorded.

Model-based ROC analysis

For each participant, we computed separate ROC curves for target-paired and distractor-paired faces. Criterion levels were defined by combining an old/new response with each of the seven confidence ratings (resulting in 14 criterion levels). Each ROC curve was fit to the DPSD model (Yonelinas, 1994), which estimates the hit rate for the kth criterion using the following equation

P (″ yes ″ \leq k | old) = R_{o} + 1 - R_{o} \times Φ (C_{k}, - d^{'} / 2, 1)

where Ck is the response criterion, Ro is the probability that an item is recollected, and $d^{'}$ is the distance between the old and new item distributions in standard deviation units. The DPSD model estimates false alarm rates using the following equation

P (″ yes ″ \leq k | new) = Φ (C_{k}, d^{'} / 2, 1)

To avoid overfitting ROCs, $d^{'}$ was constrained with a lower boundary of 0 and an upper boundary of 3, and Ro was constrained between 0 and 1 (Koen et al., 2017). MATLAB’s GlobalSearch function was used in conjunction with custom scripts to identify model parameters that minimised the root-mean-square deviation (RMSD) between the observed and modelled data.

Statistical analysis

To test the effects of square type and instruction condition on recognition strength and accuracy, we constructed linear mixed-effects models in R using the lmer function within the lme4 package (Bates et al., 2015). In cases where the dependent variable was binomial, we constructed a generalised linear mixed-effects model using the glmer function. All models included random intercepts for subjects and items and were further characterised using the emmeans package in R (Russell, 2018). Including random intercepts for each subject and image allowed us to account for uncontrolled differences among the faces and address the item as a fixed-effect fallacy (Hayes, 1973). Each linear mixed-effects model was also created in R using the stan_glm function within the rstanarm package (Goodrich et al., 2018). These models provided Bayesian parameter estimates and credible intervals, which we report in the “Supplemental Materials” section. All Bayesian model estimates are consistent with the frequentist statistical results reported here. Because a minimum of 60 responses are necessary to obtain adequately powered DPSD fits to ROCs (Yonelinas, 2002), there were not enough trials to perform an item-level analysis on recollection and familiarity estimates. To test the effects of square type and instruction condition on these estimates, we conducted a repeated-measures analysis of variance (ANOVA) with subject as a random effect and nested within-subject factors using R’s aov function.

Results

Detection task performance

Hit (button presses within 1 s of a target square) rates, hit RTs, and false alarm (button press within 1 s of a distractor square) rates in the detection task are reported in Table 1. Hit rates, hit RTs, and false alarm rates were all similar across encoding conditions, largest t(62) = 0.31, p = .76, d = 0.24, for hit rates. The lack of an effect of instruction on detection task performance likely reflects ceiling effects.

Table 1.

Mean and standard deviation of the hit rate (target HR), hit RT, in milliseconds (target RT), and false alarm rate (distractor FAR) during the detection task for each condition in Exps 1, 2a, and 2b.

Measure	Experiment 1		Experiment 2a			Experiment 2b
	Intentional	Incidental	1 pres.	2 pres.	8 pres.	500/500 ms	1,000/1,000 ms
Target HR	0.95 (0.04)	0.96 (0.04)	0.95 (0.05)	0.95 (0.05)	0.94 (0.04)	0.96 (0.05)	1.00 (0.01)
Target RT (ms)	334 (27)	332 (25)	348 (26)	340 (28)	344 (20)	345 (28)	383 (48)
Distractor FAR	0.03 (0.03)	0.04 (0.04)	0.03 (0.03)	0.04 (0.03)	0.04 (0.02)	0.03 (0.04)	0.02 (0.04)

HR: hit rate; RT: response time; FAR: false alarm rate pres.: presentation(s); ms: milliseconds.

Recognition task performance

To examine the effects of the detection task and instructions on subsequent face memory accuracy, recognition hit rates and false alarm rates were calculated (Table 2). Because false alarm rates were marginally greater for the incidental group than the intentional group, t(62) = 1.8, p = .07, d = 0.45, they were subtracted from the recognition hit rates to account for numerically different response biases across instruction conditions. The same false alarm rate was used for the square type conditions within each group. This result is reported as Adjusted Hit Rate in Table 2 to facilitate comparisons to prior work (e.g., Mulligan et al., 2014; Swallow & Jiang, 2014). To examine recognition accuracy at the level of both item and subject, we created a generalised linear mixed-effects model with recognition accuracy as a binomial dependent variable, and with encoding instruction, item type (old/new), and their interaction as fixed effects. The model revealed a significant main effect of encoding instruction on recognition accuracy, F = 5.8, p = .02, β = 0.50, 95% confidence interval (CI) = [0.18, 0.72]. An interaction of item type and instruction condition indicated that the reduction in recognition accuracy under incidental encoding was driven by an increased false alarm rate, F = 7.8, p = .005, β = 0.34, 95% CI = [0.10, 0.58].

Table 2.

Experiment 1 mean and standard deviation of the HR, FAR, adjusted HR, and item strength derived from confidence ratings for each square type and encoding instruction condition.

Measure	Intentional encoding		Incidental encoding
Measure	Target	Distractor	Target	Distractor
HR	0.72 (0.13)	0.52 (0.15)	0.68 (0.19)	0.52 (0.16)
FAR	0.15 (0.09)	0.15 (0.09)	0.21 (0.16)	0.21 (0.16)
Adjusted HR	0.57 (0.15)	0.37 (0.17)	0.47 (0.20)	0.30 (0.16)
Strength	10.4 (1.3)	8.3 (1.4)	9.7 (2.0)	8.1 (1.6)

HR: hit rate; FAR: false alarm rate.

To determine the influence of square type on recognition accuracy of old items, we ran another generalised mixed-effects model with square type and encoding instruction as fixed effects. The model revealed a main effect of square type, F = 156, p < .001, β = 1.1, 95% CI = [0.76, 1.4], but no interaction of square type and instruction condition, F = 1.6, p = .21, β = 0.34, 95% CI = [0.10, 0.59]. Old items paired with targets rather than distractors were therefore better remembered in both conditions.

Item- and subject-level analyses were conducted on strength estimates derived from confidence ratings obtained during recognition. A linear mixed-effects model included strength ratings from individual recognition trials as the dependent variable, and square type and encoding instruction as fixed effects. Joint tests revealed a significant main effect of square type, F = 310, p < .001, β = 5.0, 95% CI = [4.5, 5.5]. However, no main effect of instruction condition was observed, F = 0.35, p = .29, β = 0.68, 95% CI = [0.19, 1.2]. Finally, an interaction between encoding instruction and square type was observed, F = 20.8, p < .001, β = 0.87, 95% CI = [0.50, 1.3], indicating that the target-related strength advantage was larger under intentional encoding than incidental encoding.

DPSD estimates

The DPSD model fits to the observed ROCs are plotted in Figure 2. The DPSD equation produced excellent fits to both target and distractor-paired ROC curves (Table 3).² RMSD did not significantly differ across square type, F(1, 62) = 2.08, p = .15, or encoding instructions, F(1, 62) = 0.52, p = .47, η² = .01, Square Type × Instruction interaction, F(1, 62) = 0.87, p = .35, η² = .02.

Figure 2.

Observed (solid) and DPSD-predicted (dashed) ROCs for target-paired old faces (blue) and distractor-paired old faces (orange), under intentional (left) and incidental (right) encoding instructions. Error bars and ellipses are 1 standard error of the mean.

Table 3.

Experiment 2a mean and standard deviation of the HR, FAR, adjusted HR, and item strength derived from confidence ratings for each square type with 1, 2, and 8 encoding presentations.

Measure	1 presentation		2 presentations		8 presentations
	Target	Distractor	Target	Distractor	Target	Distractor
HR	0.47 (15)	0.38 (0.13)	0.52 (0.18)	0.42 (0.16)	0.65 (0.21)	0.45 (0.20)
FAR	0.32 (0.13)	0.32 (0.13)	0.25 (0.13)	0.25 (0.13)	0.17 (0.09)	0.17 (0.09)
Adjusted HR	0.15 (0.12)	0.07 (0.11)	0.26 (0.13)	0.17 (0.12)	0.48 (0.23)	0.29 (0.20)
Strength	7.7 (1.1)	6.9 (1.0)	8.1 (1.8)	7.1 (1.7)	9.6 (2.3)	7.6 (2.0)

HR: hit rate; FAR: false alarm rate.

Repeated-measures ANOVAs were performed on the Ro and $d^{'}$ parameter estimates to test the effects of square type and instructions on recollection and familiarity. If attending to behaviourally relevant targets boosts memory for an item’s context, then recollection estimates should be greater for target-paired faces than for distractor-paired faces. This was indeed the case, main effect of square type, F(1, 62) = 15.3, p < .001, η² = .20. In contrast, encoding instructions did not affect recollection estimates, main effect of instruction, F(1, 62) = 0.001, p = .98, η² < .001, Square Type × Instruction interaction, F(1, 62) = 0.04, p = .85, η² < .001. This indicates that the effect of target detection on recollection estimates did not depend on the intention to encode the images.

As with recollection, familiarity estimates were greater for target-paired than distractor-paired faces, main effect of square type, F(1, 62) = 54.5, p < .001, η² = .47. However, unlike recollection, instructions to memorise rather than ignore the faces increased familiarity estimates during recognition, main effect of instructions, F(1, 62) = 7.0, p = .01, η² = .10. There was no interaction between square type and instructions, F(1, 62) < 0.001, p = .999, η² < .001.

These results are consistent with our strength analysis and are evident in the qualitative properties of the observed ROCs (Figure 2). The Y-intercepts of respective target and distractor-paired ROCs are roughly the same across encoding instruction conditions, while the intentional encoding ROC has a higher degree of curvilinearity across the negative diagonal axis than that of the incidental encoding condition, indicating greater contribution of familiarity to recognition (Onyper et al., 2010). An alternative model of recognition memory, the Unequal Variance Signal Detection (UVSD; Egan, 1958; Wixted, 2007) model, proposes that recognition memory is a function of a single, continuous strength-based decision process. According to the UVSD model, asymmetries in ROC curves reflect unequal variance in the strength distributions of old and new items, rather than a threshold-like recollection process (Egan, 1958; Wixted, 2007). Greater variance in the strength of old items (relative to new items) could reflect variability in successful encoding (Wixted, 2007), attention to the items (DeCarlo, 2002), or intrinsic differences in the memorability items (Pratte et al., 2010). If target detection adds memory strength to concurrently presented items, then the UVSD estimates of old item strength $(d^{'})$ and old item variability (σ_old) should be greater for target-paired items than for distractor-paired items. UVSD fits to the ROC curves in Figure 2 did produce significantly larger $d^{'}$ estimates for the target-paired items (M = 1.80, SD = .73) than the distractor-paired items, M = 1.13, SD = 0.67; t(62) = 5.38, p < .001, d = 1.06. However, although σ_old estimates were greater for target-paired items (M = 1.49, SD = 0.58) than distractor-paired items (M = 1.35, SD = 0.43), they did not significantly differ, t(62) = 1.59, p = .11. This is consistent with prior findings that UVSD σ_old estimates do not always capture differences in strength variability due to encoding manipulations (Koen & Yonelinas, 2010, 2013; but see Starns et al., 2012). Starns et al. (2012) suggest that differences in σ_old will only be observed when there are large differences in memory performance, so it is possible that our manipulation was too weak to produce a significant effect in this parameter.

In summary, Experiment 1 demonstrated that estimates of both recollection and familiarity are facilitated by target detection, regardless of whether the images were intentionally encoded. This finding implies that the goal relevance of the background items does not influence the ability to form a contextually rich representation of the scene. However, as with many previous studies of the attentional boost effect, images were presented multiple times. This leaves open the possibility that subjects are forming item–cue associations over multiple presentations to facilitate target detection. To determine whether target-related recollection advantages are a result of task-specific learning, additional studies varying the number and duration of presentations are needed.

Experiment 2a

Our finding that target detection increases estimates of subsequent recollection and familiarity under both intentional and incidental encoding conditions contrasts with some existing studies on the attentional boost effect while affirming the findings of others. An important parameter that varies among these studies is the number of times the items were presented—once in studies showing no context memory advantage (e.g., Mulligan et al., 2016; Spataro et al., 2017), and multiple times or for longer durations in studies showing a context memory advantage (e.g., Leclercq et al., 2014; Meng et al., 2019; Swallow & Atir, 2018; Turker & Swallow, 2018; Experiment 1). This raises two possibilities. One is that the memory benefit could reflect the predictive relationship between the item and the presence (or absence) of a target. However, breaking the predictive relationship between items and detection task stimuli does not eliminate the attentional boost effect (e.g., Lin et al., 2010; Makovski et al., 2011). In addition, that predict that a target will appear in 100 ms are remembered about as well as items that predict distractor onset (Swallow & Jiang, 2011). Another possibility is that the target-related recollection advantage depends on having adequate opportunity to form a contextual representation of the event through repeated or prolonged presentations. Additional studies that vary the number and duration of presentations will help address both of these concerns. In Experiments 2a and 2b, we manipulated the amount of exposure participants had to each item–cue pair by varying the number or duration of presentations. Because encoding instructions did not influence recollection estimates in Experiment 1, all participants were instructed to remember the images to counter the potential for floor effects in recognition memory when the items are presented once.

Method

The method for Experiment 2a was identical to that of the Experiment 1 except for several key changes. First, the number of presentations, rather than being fixed at 8, was varied across participants. Participants viewed the images either 1, 2, or 8 times. In addition, the number of Old Faces and New Faces increased from 50 to 80 to produce adequate power across conditions. As in the Intentional Encoding condition in Experiment 1, participants were instructed to remember all background images presented during the encoding and detection task. To attain our target sample size of 32 participants per group, we recruited a total of 104 participants (31 male, 73 female, 18–22 years old, age M = 19.2, SD = 1.1) for the study. Of these participants, 35 viewed the images 1 time, 37 viewed the images 2 times, and the remaining 32 viewed the images 8 times. Presentation duration of the faces and squares was identical to those in Experiment 1 (500 ms face duration, 100 ms square duration). The duration of the mask was shortened to 250 ms to mitigate the increase in the length of the experiment and meet time constraints. Previous work suggests this difference in encoding time does not produce appreciable differences in the attentional boost effect in old/new recognition hit rates (Mulligan & Spataro, 2015).

Results

Detection task performance

Hit rates, hit RTs, and false alarm rates are reported in Table 1 and were all similar across presentation conditions, largest F = 1.6, p = .21, d = 0.03, for false alarm rates.

Recognition task performance

Table 3 shows the adjusted hit rates in the recognition task for each experimental condition. As with Experiment 1, adjusted hit rates were larger for target-paired items than for distractor-paired items across each presentation condition (1, 2, and 8).

A generalised linear mixed-effects model with recognition accuracy as a binomial dependent variable and with number of presentations, item type (old/new), and their interaction as fixed effects revealed a main effect of number of presentations, F = 30, p < .001, linear β = 0.68, 95% CI = [0.53, 0.82]. An interaction of item type and number of presentations indicated that an increased false alarm rate was the primary reason for reduced accuracy with fewer presentations, F = 9.7, p < .001, linear β = 0.27, 95% CI = [0.15, 0.40]. With new items removed and square type added as a fixed effect, the model revealed a main effect of square type, F = 151, p < .001, β = 0.60, 95% CI = [0.51, 0.70], and an interaction of square type and number of presentations, F = 14, p < .001, linear β = 0.43, 95% CI = [0.26, 0.60]. These findings indicate that adjusted hit rates were greater for images paired with targets than distractors, and this difference increased with the number of presentations.

Joint tests on a linear mixed-effects model with strength ratings from individual recognition trials as the dependent variable and with square type, encoding instruction, and their interaction as fixed effects demonstrated a significant main effect of square type, F = 221, p < .001, β = 1.3, 95% CI = [1.1, 1.5], as well as a main effect of number of presentations, F = 5.8, p = .003, linear β = 0.47, 95% CI = [–0.11, 1.0], and an interaction between number of presentations and square type, F = 21, p < .001, linear β = 0.95, 95% CI = [0.65, 1.3].

DPSD estimates

To determine the effects of square type and number of presentations on recollection and familiarity, we fit the DPSD model to ROC curves derived from recognition responses and confidence ratings. Estimates of recollection and familiarity for Experiment 2 are provided in Figure 3. The DPSD model provided excellent fits to the data (M RMSD = 0.04). RMSDs for each experimental condition are provided in the “Supplemental Materials” section and did not differ across square type or number of presentations.

Figure 3.

DPSD recollection and familiarity estimates for Experiments 2a and 2b.

A repeated-measures ANOVA with recollection estimates as the dependent variable revealed main effects of square type, F(1, 101) = 25, p < .001, η² = .18, replicating our findings from Experiment 1. It also revealed a main effect of number of presentations, F(2, 101) = 21, p < .001, η² = .29. Importantly, an interaction between square type and number of presentations was observed, F(2, 101) = 7.1, p = .001, η² = .10, with the effect of target detection on recollection estimates increasing with number of presentations. Similarly, in our familiarity estimates, we observed a main effect of square type, F(1, 101) = 51, p < .001, η² = .31, main effect of number of presentations, F(2, 101) = 33, p < .001, η² = .39, and Square Type × Number of Presentations interaction, F(2, 101) = 6.4, p = .002, η² = .08. These findings suggest that both target detection and increasing the number of presentations increase estimates of both recollection and familiarity. As indicated by the DPSD estimates (Figure 3), there was no target-related recollection advantage for images that were only presented once, while the target-related familiarity advantage is observed in all presentation conditions.

These results indicate that increasing the number of presentations resulted in increased target-related advantages in recognition strength and accuracy measures, as well as in recollection and familiarity estimates. Importantly, although a familiarity boost was observed, we did not observe a target-related recollection advantage with one presentation, t(34) = .37, p = .72. However, it is still possible that one brief presentation does not provide an adequate opportunity to form a contextual representation of the scene, particularly because recollection estimates were at floor (Malmberg & Shiffrin, 2005). Addressing whether the target-related recollection advantage relies on learned item–cue associations across repeated presentations or simply providing adequate exposure to the background image requires additional experimentation with images presented once for varying durations.

Experiment 2b

To test whether the effects of target detection on recollection and familiarity increase with more encoding time, we conducted an experiment in which the presentation duration was varied within participants. This will address the question of whether the target-related context memory advantage occurs as a result of multiple pairings or depends on having an adequate opportunity to form a contextual representation of the scene. If the advantage depends on having adequate encoding opportunity, then we can expect to see a target-related recollection advantage with a single presentation at long durations, but not at short durations.

Method

The method for Experiment 2b was identical to that of the Experiment 2a, but the number of presentations was fixed at 1 and the presentation duration was varied within participants. Participants encoded 80 images in two blocks. For one of the blocks, the presentation duration was identical to that of Experiment 1 (500/500 condition; 100 ms square duration, 500 ms image duration, 500 ms mask duration). Because recollection estimates were at floor in Experiment 2a, we increased the mask duration to the value used in Experiment 1 (500 ms). In the other block, the face and mask presentation duration were both doubled to 1,000 ms (1,000/1,000 condition). The order of long- and short-duration blocks was counterbalanced between participants, and old faces were randomly assigned to square type and duration conditions for each participant. Because this experiment had four within-subject conditions (as opposed to two in Experiments 1 and 2a), the number of hit responses used to compute each subject’s ROC curve was half of those in Experiment 2a. To account for this power deficit, we increased the target sample size by 50%. We recruited a sample of 45 participants (10 male, 35 female, 18–24 years old, age M = 20.4, SD = 1.6) to take part in the study.

Model-based ROC analysis

As with Experiments 1 and 2a, we computed ROC curves from the recognition responses and confidence ratings provided. However, in addition to computing separate ROC curves from target- and distractor-paired images, for each participant we also computed separate ROC curves for each duration condition (500/500 and 1,000/1,000), resulting in four separate ROC curves calculated using the same false alarm rate.

Results

Detection task performance

Hit rates, hit RTs, and false alarm rates are reported in Table 1. Hit rates were high in both conditions but were significantly better in the 1,000/1,000 duration condition, t(44) = 4.9, p < .001, d = 0.73, reflecting near-perfect detection in that condition for almost all participants. This increase in accuracy came with a trade-off in speed, as RTs were also longer in the 1,000/1,000 condition than in the 500/500 condition, t(44) = 6.6, p < .001, d = 0.98. Distractor false alarm rates did not differ between duration conditions, t(44) = 0.86, p = .39, d = 0.13.

Recognition task performance

Table 4 shows the adjusted hit rates in the recognition task for each experimental condition. As with Experiments 1 and 2a, adjusted hit rates were larger for target-paired items than for distractor-paired items across each duration condition. Because the manipulation of presentation duration occurred within participants, false alarm rates could not differ between conditions and new items were therefore not included in the analysis. A linear mixed-effects model revealed a main effect of square type, F = 15, p < .001, β = 0.27, 95% CI = [0.13, 0.41], but no main effect of duration condition, F = 1.0, p = .31, β = 0.001, 95% CI = [−0.14, 0.14], or interaction between duration condition and square type, F = 0.96, p = .33, β = 0.10, 95% CI = [−0.10, 0.29]. The attentional boost effect was therefore similar in magnitude with short and long durations.

Table 4.

Experiment 2b mean and standard deviation of the HR, FAR, adjusted hit rate, and item strength derived from confidence ratings for each square type and duration.

Measure	500/500 ms		1,000/1,000 ms
Measure	Target	Distractor	Target	Distractor
HR	0.50 (0.14)	0.46 (0.16)	0.53 (0.18)	0.46 (0.16)
FAR	0.30 (0.11)	0.30 (0.11)	0.30 (0.11)	0.30 (0.11)
Adjusted HR	0.21 (0.15)	0.16 (0.14)	0.24 (0.18)	0.16 (0.17)
Strength	8.0 (1.3)	7.6 (1.6)	8.3 (1.6)	7.6 (1.6)

HR: hit rate; FAR: false alarm rate.

Joint tests on a linear mixed-effects model with strength ratings from individual recognition trials as the dependent variable and with square type, duration condition, and their interaction as fixed effects revealed a significant main effect of square type on strength ratings, F = 17, p < .001, β = 0.58, 95% CI = [0.30, 0.85], though no main effect of presentation duration was observed, F = 0.93, p = .34, β = 0.001, 95% CI = [−0.27, 0.27]. In addition, no interaction of presentation duration and square type was observed, F = 0.91, p = .34, β = 0.19, 95% CI = [−0.20, 0.58].

DPSD estimates

To determine the effects of square type and presentation duration on recollection and familiarity, we fit the DPSD model to ROC curves derived from recognition responses and confidence ratings. Estimates of recollection and familiarity for Experiment 2b are provided in Figure 3. The DPSD model provided excellent fits to the data (M RMSD = 0.04). RMSDs for each experimental condition are provided in the “Supplemental Materials” section and did not differ across square type or duration conditions.

A repeated-measures ANOVA with recollection estimates as the dependent variable revealed a significant main effect of square type on estimates of recollection, F(1, 44) = 5.9, p = .02, η² = .13. A small overall increase in recollection estimates with increases in duration was not significant, main effect of duration, F(2, 44) = 0.37, p = .54, η² = .01, nor was the interaction between square type and duration significant, F(2, 44) = 1.7, p = .20, η² = .037. The ANOVA therefore provides evidence that recollection estimates are greater for faces paired with a target. However, paired t tests indicated that this effect was driven by the 1,000/1,000 condition: Although there was no significant target-related recollection advantage in the 500/500 condition, t(44) = 0.55, p = .58, d = 0.08, the target-related recollection advantage was significant for faces in the 1,000/1,000 condition, t(44) = 3.8, p < .001, d = 0.57. As for familiarity estimates, we observed a marginal main effect of square type, F(1, 44) = 3.2, p = .08, η² = .068, but no main effect of presentation duration, F(1, 44) = 0.12, p = .73, η² = .052, or interaction effects, Square Type × Duration interaction, F(1, 44) = 0.044, p = .84, η² = .001. The results therefore suggest that target detection and presentation duration did not significantly influence familiarity estimates, while target detection only increased recollection estimates at longer durations.

In Experiment 2b, we found a target-related increase in recollection estimates when items were presented just one time. Although the interaction was not significant, simple effects tests indicated that this effect was present at longer presentation durations but not at shorter durations. This finding supports the idea that the attentional boost effect can reflect context memory for a single event.

General discussion

This study examined the effects of attending to behaviourally relevant stimuli on memory for concurrently presented items and their context. It used quantitative models of recollection and familiarity during recognition to estimate the effect of target detection during encoding on memory for an item and its context. The data replicated four important findings: (1) recognition accuracy was greater for target-paired items than distractor-paired items (the attentional boost effect); (2) the attentional boost effect occurred for incidentally encoded items; (3) familiarity, but not recollection, was reduced when the items were encoded incidentally rather than intentionally (Kishiyama & Yonelinas, 2003); and (4) the attentional boost effect is present even when background items are presented just one time. Most importantly, the memory advantage for target-paired items was reflected in estimates of both recollection and familiarity. These effects were observed under both incidental and intentional encoding instructions. The magnitude of these effects increased with the number of times each item was presented, and they were also observed with a single presentation of each item when the presentation duration was extended. The data supported the view that the attentional boost effect includes memory for an item as well as its context, regardless of the intention to remember it or the number of times the item is presented, as long as there is sufficient encoding time.

The data from our study help arbitrate between item-only and context accounts by providing estimates of the contributions of recollection, which reflects aspects of context memory, and familiarity, which reflects item memory strength (Diana et al., 2007; Yonelinas et al., 2010), to the attentional boost effect. The current study provides evidence that recollection, as well as familiarity, contributes more to the recognition of target-paired items than distractor-paired items. Because recollection does not require the retrieval of all aspects of the study event, it is possible for recollection to occur when source memory for a specific feature (or set of features) is poor (as in “noncriterial recollection”; Yonelinas et al., 2010). Our data also suggest that the effects of target detection on context memory are present even with encoding conditions that are likely to discourage the use of a unitisation strategy (asking participants to ignore the items, presenting items only once; Diana et al., 2008). The bulk of the evidence therefore supports the claim that attending to behaviourally relevant stimuli boosts memory for at least some elements of an item’s context as well as the item.

Context memory can include multiple aspects of the situation in which an item was encoded, suggesting that the conflicting data on context memory and the attentional boost effect could be reconciled in two ways. First, some elements of the context in which an item was encoded may provide a more sensitive measure of context memory if they are better remembered or easier to retrieve from memory than other contextual elements (e.g., location vs. colour; Chen & Wyble, 2015; Uncapher et al., 2006). Second, it may be possible that target detection has different effects on the encoding of different context elements, yet still increases recollection. For example, the generation effect (e.g., better memory for words that were generated in response to a cue word, such as when one produces the antonym to ‘good’; Slamecka & Graf, 1978) increases recollection estimates (Clark, 1995) but has opposite effects on memory for cue word’s colour (which is impaired) and location (which is sometimes enhanced) (Mulligan, 2004, 2011).

This experiment offers additional insight into the mechanism that produces the attentional boost effect by examining the interaction of attention to briefly presented targets and goal-directed attention to the background items. Incidental encoding resulted in a reduction of the estimates of familiarity, but not recollection, replicating Kishiyama and Yonelinas (2003). This finding also contradicts the suggestion that the attentional boost effect results from greater distinctiveness processing of target-paired items, as recollection of distinctive items is reduced when they are incidentally, rather than intentionally, encoded (Kishiyama & Yonelinas, 2003; see also Smith & Mulligan, 2018). The finding that the attentional boost effect is present even for items that are incidentally encoded replicates some earlier reports (Dewald et al., 2013; Swallow & Jiang, 2014), but not others (Dewald et al., 2011; Swallow & Jiang, 2011), and challenges accounts that suggest that task-irrelevant information should be inhibited at these moments (Choi et al., 2009; Leclercq et al., 2014; Leclercq & Seitz, 2012; Walker et al., 2017). It may be possible that detecting a target simply mitigates the effects of dividing attention during encoding, because items paired with targets are remembered as well as those that are encoded under full attention (Spataro et al., 2013; Swallow & Jiang, 2010). However, the recollection advantage during incidental encoding suggests that target detection does more than simply allow participants to more fully attend to items they are attempting to memorise. Therefore, the data support the claim that the critical factor for producing the memory advantage for target-paired items is detecting the presence of a stimulus that requires an overt or covert response, not the intention to remember the item (Swallow & Jiang, 2014).

Results from Experiments 2a and 2b replicate a rich literature showing a memory advantage with increasing number of item presentations, with comparable benefits to recollection and familiarity (see Yonelinas, 2002, for review). Our results extend these findings to show that target-related advantages for recollection and familiarity increase with the number of presentations. In addition, although the attentional boost effect has been previously observed with just one presentation (Makovski et al., 2011; Meng et al., 2019; Mulligan et al., 2016), this study systematically examined whether the context memory advantage for target-paired items relies on item–cue associations learned over multiple presentations. In Experiment 2a, we found that with one brief presentation (750 ms per trial), there is a target-related advantage for familiarity but not recollection. In Experiment 2b, extending the trial to 2,000 ms resulted in a recollection advantage. The results suggest that a target-related context memory advantage is observed when participants are given sufficient opportunity to form a contextually rich representation of the scene.

The observation that the attentional boost effect for recollection estimates increases with longer durations appears to contradict earlier reports that the advantage in hit rates decreases with longer study durations (Mulligan & Spataro, 2015). These and other data (Malmberg & Nelson, 2003; Spataro et al., 2017) formed the basis of an “early phase encoding” account of the attentional boost effect, which suggested that the benefits of target detection to encoding occur before more controlled, elaborative processes take place. The data presented in Experiments 2a and 2b suggest that the effects of target detection, however, may be more complex. In conjunction with Mulligan et al. (2014), Experiment 2b suggests one initial interpretation: that the enhancement to recollection of target-paired items reflects elaborative or late-stage processes, and that better recognition memory at shorter durations may primarily reflect an early phase enhancement to familiarity (observed in Experiments 1 and 2a). Although this may provide a straightforward account of the duration data, it is not easily reconciled with the effect of the number of presentations on both recollection and familiarity, or the incidental memory advantages, when the effects of effortful, controlled encoding should be reduced (Russo et al., 1998; Russo & Mammarella, 2002). Instead, the data are consistent with the suggestion that context representations can take longer than 2 s to form and may do so incidentally, but that a limited amount of information about context is stored with each item presentation (Malmberg & Shiffrin, 2005; see also Hirshman et al., 2002; Malmberg & Nelson, 2003, for related findings with the word frequency effect). When items are presented multiple times, however, context representations are accumulated and context memory should increase. If this is the case, then our findings suggest that target detection could enhance the formation of context memories by speeding up the formation of context representations or by increasing the amount of context information that is acquired during a single presentation (see also Kuhl & Wagner, 2010; Xue et al., 2010).

Future research

Because recollection refers to the experience of recalling extra information about the moment during which an item was encountered, it necessarily includes a variety of information, including which other items were present (as in relational memory), where the item was located, which items were encountered earlier, or later, other studied items that were semantically or perceptually related to the study item. It is possible that the attentional boost effect enhances memory for some aspects of an item’s context (e.g., relational memory; Swallow & Atir, 2018; Turker & Swallow, 2018), but not others (e.g., inter-item associations; Mulligan et al., 2016; Spataro et al., 2017). Focused studies exploring many of these features are necessary for clarifying which aspects of context memory are enhanced.

Additional studies are needed to examine other differences in the tasks that show a boost to context memory, and those that do not, such as the nature of the materials that were encoded. The use of different types of materials, words or images, across studies may have promoted different encoding strategies (Bernstein et al., 2002; Grady et al., 1998; Walker et al., 2017). Although our data show that target detection enhances context memory for faces when given sufficient time, it is not clear whether the effects of encoding instructions, item repetition, and duration are the same with other types of materials. Future research should characterise how recollection and familiarity for words change with longer, or spaced, presentations and when the words are incidentally encoded.

Conclusion

The attentional boost effect reflects enhanced memory for items and at least some aspects of their context. This memory advantage occurs even when items are incidentally encoded, suggesting that the effects do not depend on goal-directed attention to the items. In addition, these effects are present when items are only presented once, given sufficient encoding time. These results could reflect a selection mechanism that enhances the encoding of multiple aspects of events in which behaviourally relevant stimuli, such as targets in a detection task, appear. Such a mechanism could produce immediate benefits to planning and goal-directed activity in everyday contexts and has implications for understanding how multiple cognitive systems interact to produce adaptive behaviours in complex, dynamic environments.

Supplemental Material

ABE_Recollection_Supplemental_Materials – Supplemental material for The effects of encoding instruction and opportunity on the recollection of behaviourally relevant events

Supplemental material, ABE_Recollection_Supplemental_Materials for The effects of encoding instruction and opportunity on the recollection of behaviourally relevant events by Adam Wood Broitman and Khena Marie Swallow in Quarterly Journal of Experimental Psychology

Footnotes

Acknowledgements

The authors thank Mariel Emrich, Deanna Earle, and Ami Agrawal for assistance with data collection on this project.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Khena Marie Swallow

Supplemental material

The Supplemental Material is available at

Notes

References

Bainbridge

W. A.

Isola

Oliva

(2013). The intrinsic memorability of face photographs. Journal of Experimental Psychology: General, 142(4), 1323–1334. https://doi.org/10.1037/a0033872

Bates

Mächler

Bolker

Walker

(2015). Fitting linear mixed-effects models using lme4. Journal of Statistical Software, 67(1), 1–48. https://doi.org/10.18637/jss.v067.i01

Bernstein

L. J.

Beig

Siegenthaler

A. L.

Grady

C. L.

(2002). The effect of encoding strategy on the neural correlates of memory for faces. Neuropsychologia, 40(1), 86–98. https://doi.org/10.1016/S0028-3932(01)00070-7

Brainard

D. H.

(1997). The psychophysics toolbox. Spatial Vision, 10(4), 443–446.

Brainerd

Gomes

Nakamura

(2015). Dual recollection in episodic memory. Journal of Experimental Psychology: General, 144(4), 816–843.

Chalfonte

B. L.

(1996). Spatial location memory in amnesia: Binding item and location information under incidental and intentional encoding conditions. Memory, 4(6), 591–614.

Chan

J. C.

McDermott

K. B.

(2007). The testing effect in recognition memory: A dual process account. Journal of Experimental Psychology: Learning, Memory, and Cognition, 33(2), 431–437. https://doi.org/10.1037/0278-7393.33.2.431

Chen

Wyble

(2015). The location but not the attributes of visual cues are automatically encoded into working memory. Vision Research, 107, 76–85.

Choi

Seitz

A. R.

Watanabe

(2009). When attention interrupts learning: Inhibitory effects of attention on TIPL. Vision Research, 49(21), 2586–2590.

10.

Clark

S. E.

(1995). The generation effect and the modeling of associations in memory. Memory & Cognition, 23(4), 442–455.

11.

Craik

F. I. M.

Lockhart

R. S.

(1972). Levels of processing: A framework for memory research. Journal of Verbal Learning and Verbal Behavior, 11(6), 671–684.

12.

Davachi

(2006). Item, context and relational episodic encoding in humans. Current Opinion in Neurobiology, 16(6), 693–700. https://doi.org/10.1016/j.conb.2006.10.012

13.

DeCarlo

L. T.

(2002). Signal detection theory with finite mixture distributions: Theoretical developments with applications to recognition memory. Psychological Review, 109(4), 710–721.

14.

Desimone

Duncan

(1995). Neural mechanisms of selective visual attention. Annual Review of Neuroscience, 18(1), 193–222.

15.

Dewald

A. D.

Sinnett

Doumas

L. A.

(2011). Conditions of directed attention inhibit recognition performance for explicitly presented target-aligned irrelevant stimuli. Acta Psychologica, 138(1), 60–67.

16.

Dewald

A. D.

Sinnett

Doumas

L. A.

(2013). A window of perception when diverting attention? Enhancing recognition for explicitly presented, unattended, and irrelevant stimuli by target alignment. Journal of Experimental Psychology: Human Perception and Performance, 39(5), 1304–1312.

17.

Diana

R. A.

Reder

L. M.

Arndt

Park

(2006). Models of recognition: A review of arguments in favor of a dual-process account. Psychonomic Bulletin & Review, 13(1), 1–21. https://doi.org/10.3758/BF03193807

18.

Diana

R. A.

Yonelinas

A. P.

Ranganath

(2007). Imaging recollection and familiarity in the medial temporal lobe: A three-component model. Trends in Cognitive Sciences, 11(9), 379–386.

19.

Diana

R. A.

Yonelinas

A. P.

Ranganath

(2008). The effects of unitization on familiarity-based source memory: Testing a behavioral prediction derived from neuroimaging data. Journal of Experimental Psychology: Learning, Memory, and Cognition, 34(4), 730–740. https://dx-doi-org.web.bisu.edu.cn/10.1037/0278-7393.34.4.730

20.

Egan

J. P.

(1958). Recognition memory and the operating characteristic (Technical Note No. AFCRC-TN-58-51). Hearing and Communication Laboratory, Indiana University.

21.

Fanselow

M. S.

(1990). Factors governing one-trial contextual conditioning. Animal Learning & Behavior, 18(3), 264–270.

22.

Faul

Erdfelder

Lang

A.-G.

Buchner

(2007). G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behavior Research Methods, 39(2), 175–191. https://doi.org/10.3758/BF03193146

23.

Gardiner

J. M.

Parkin

A. J.

(1990). Attention and recollective experience in recognition memory. Memory & Cognition, 18(6), 579–583.

24.

Geraci

McCabe

D. P.

(2006). Examining the basis for illusory recollection: The role of remember/know instructions. Psychonomic Bulletin & Review, 13(3), 466–473.

25.

Ghetti

Angelini

(2008). The development of recollection and familiarity in childhood and adolescence: Evidence from the dual-process signal detection model. Child Development, 79(2), 339–358. https://doi.org/10.1111/j.1467-8624.2007.01129.x

26.

Giovanello

K. S.

Keane

M. M.

Verfaellie

(2006). The contribution of familiarity to associative memory in amnesia. Neuropsychologia, 44(10), 1859–1865.

27.

Goodrich

Gabry

Ali

Brilleman

(2018). rstanarm: Bayesian applied regression modeling via Stan (R Package Version 2.17.4). http://mc-stan.org/

28.

Grady

C. L.

McIntosh

A. R.

Rajah

M. N.

Craik

F. I.

(1998). Neural correlates of the episodic encoding of pictures and words. Proceedings of the National Academy of Sciences of the United States of America, 95(5), 2703–2708.

29.

Hasher

Zacks

R. T.

(1988). Working memory, comprehension, and aging: A review and a new view. In Bower

G. H.

(Ed.), Psychology of learning and motivation (Vol. 22, pp. 193–225). Elsevier.

30.

Hayes

W. L.

(1973). Statistics for social sciences. Holt, Rinehart and Winston.

31.

Hirshman

Fisher

Henthorn

Arndt

Passannante

(2002). Midazolam amnesia and dual-process models of the word-frequency mirror effect. Journal of Memory and Language, 47(4), 499–516.

32.

Hirst

Volpe

B. T.

(1984). Encoding of spatial relations with amnesia. Neuropsychologia, 22(5), 631–634.

33.

Hockley

W. E.

(1992). Item versus associative information: Further comparisons of forgetting rates. Journal of Experimental Psychology: Learning, Memory, and Cognition, 18(6), 1321–1330.

34.

Hunt

R. R.

McDaniel

M. A.

(1993). The enigma of organization and distinctiveness. Journal of Memory and Language, 32(4), 421–445.

35.

Kahn

Davachi

Wagner

A. D.

(2004). Functional-neuroanatomic correlates of recollection: Implications for models of recognition memory. Journal of Neuroscience, 24(17), 4172–4180. https://doi.org/10.1523/JNEUROSCI.0624-04.2004

36.

Kelley

Wixted

J. T.

(2001). On the nature of associative information in recognition memory. Journal of Experimental Psychology: Learning, Memory, and Cognition, 27(3), 701–722.

37.

Kinchla

R. A.

(1992). Attention. Annual Review of Psychology, 43(1), 711–742.

38.

Kishiyama

M. M.

Yonelinas

A. P.

(2003). Novelty effects on recollection and familiarity in recognition memory. Memory & Cognition, 31(7), 1045–1051. https://doi.org/10.3758/BF03196125

39.

Koen

J. D.

Barrett

F. S.

Harlow

I. M.

Yonelinas

A. P.

(2017). The ROC Toolbox: A toolbox for analyzing receiver-operating characteristics derived from confidence ratings. Behavior Research Methods, 49(4), 1399–1406. https://doi.org/10.3758/s13428-016-0796-z

40.

Koen

J. D.

Yonelinas

A. P.

(2010). Memory variability is due to the contribution of recollection and familiarity, not to encoding variability. Journal of Experimental Psychology: Learning, Memory, and Cognition, 36(6), 1536–1542.

41.

Koen

J. D.

Yonelinas

A. P.

(2013). Still no evidence for the encoding variability hypothesis: A reply to Jang, Mickes, and Wixted (2012) and Starns, Rotello, and Ratcliff (2012). Journal of Experimental Psychology: Learning, Memory, and Cognition, 39(1), 304–312.

42.

Kuhl

Wagner

(2010). Strategic control of memory. In Squire

L. R.

(Ed.), Encyclopedia of neuroscience (pp. 437–444). Elsevier.

43.

Leclercq

Le Dantec

C. C.

Seitz

A. R.

(2014). Encoding of episodic information through fast task-irrelevant perceptual learning. Vision Research, 99, 5–11. https://doi.org/10.1016/j.visres.2013.09.006

44.

Leclercq

Seitz

A. R.

(2012). The impact of orienting attention in fast task-irrelevant perceptual learning. Attention, Perception, & Psychophysics, 74(4), 648–660.

45.

Lin

J. Y.

Pype

A. D.

Murray

S. O.

Boynton

G. M.

(2010). Enhanced memory for scenes presented at behaviorally relevant points in time. PLOS Biology, 8(3), Article e1000337. https://doi.org/10.1371/journal.pbio.1000337

46.

Makovski

Swallow

K. M.

Jiang

Y. V.

(2011). Attending to unrelated targets boosts short-term memory for color arrays. Neuropsychologia, 49(6), 1498–1505.

47.

Malmberg

K. J.

Nelson

T. O.

(2003). The word frequency effect for recognition memory and the elevated-attention hypothesis. Memory & Cognition, 31(1), 35–43.

48.

Malmberg

K. J.

Shiffrin

R. M.

(2005). The “one-shot” hypothesis for context storage. Journal of Experimental Psychology: Learning, Memory, and Cognition, 31(2), 322–336.

49.

Mangels

J. A.

Picton

T. W.

Craik

F. I.

(2001). Attention and successful episodic encoding: An event-related potential study. Cognitive Brain Research, 11(1), 77–95.

50.

McCabe

D. P.

Geraci

L. D.

(2009). The influence of instructions and terminology on the accuracy of remember–know judgments. Consciousness and Cognition, 18(2), 401–413.

51.

Meng

Lin

(2019). The role of distractor inhibition in the attentional boost effect: Evidence from the R/K paradigm. Memory, 27(6), 750–757.

52.

Mulligan

N. W.

(1998). The role of attention during encoding in implicit and explicit memory. Journal of Experimental Psychology: Learning, Memory, and Cognition, 24(1), 27–47. https://doi.org/10.1037/0278-7393.24.1.27

53.

Mulligan

N. W.

(2004). Generation and memory for contextual detail. Journal of Experimental Psychology: Learning, Memory, and Cognition, 30(4), 838–855.

54.

Mulligan

N. W.

(2011). Generation disrupts memory for intrinsic context but not extrinsic context. Quarterly Journal of Experimental Psychology, 64(8), 1543–1562.

55.

Mulligan

N. W.

Smith

S. A.

Spataro

(2016). The attentional boost effect and context memory. Journal of Experimental Psychology: Learning, Memory, and Cognition, 42(4), 598–607. https://doi.org/10.1037/xlm0000183

56.

Mulligan

N. W.

Spataro

(2015). Divided attention can enhance early-phase memory encoding: The attentional boost effect and study trial duration. Journal of Experimental Psychology: Learning, Memory, and Cognition, 41(4), 1223–1228. https://doi.org/10.1037/xlm0000055

57.

Mulligan

N. W.

Spataro

Picklesimer

(2014). The attentional boost effect with verbal materials. Journal of Experimental Psychology: Learning, Memory, and Cognition, 40(4), 1049–1063. https://doi.org/10.1037/a0036163

58.

Onyper

S. V.

Zhang

Y. X.

Howard

M. W.

(2010). Some-or-none recollection: Evidence from item and source memory. Journal of Experimental Psychology: General, 139(2), 341–364. https://doi.org/10.1037/a0018926

59.

Parks

C. M.

(2007). The role of noncriterial recollection in estimating recollection and familiarity. Journal of Memory and Language, 57(1), 81–100.

60.

Pelli

D. G.

(1997). The VideoToolbox software for visual psychophysics: Transforming numbers into movies. Spatial Vision, 10(4), 437–442. https://doi.org/10.1163/156856897X00366

61.

Pratte

M. S.

Rouder

J. N.

Morey

R. D.

(2010). Separating mnemonic process from participant and item effects in the assessment of ROC asymmetries. Journal of Experimental Psychology: Learning, Memory, and Cognition, 36(1), 224–232.

62.

Rossi-Arnaud

Spataro

Costanzi

Saraulli

Cestari

(2018). Divided attention enhances the recognition of emotional stimuli: Evidence from the attentional boost effect. Memory, 26(1), 42–52.

63.

Rugg

M. D.

Curran

(2007). Event-related potentials and recognition memory. Trends in Cognitive Sciences, 11(6), 251–257. https://doi.org/10.1016/j.tics.2007.04.004

64.

Russell

(2018). emmeans: Estimated marginal means, aka least-squares means (R Package Version). https://cran.r-project.org/web/packages/emmeans/index.html

65.

Russo

Mammarella

(2002). Spacing effects in recognition memory: When meaning matters. European Journal of Cognitive Psychology, 14(1), 49–59.

66.

Russo

Parkin

A. J.

Taylor

S. R.

Wilks

(1998). Revising current two-process accounts of spacing effects in memory. Journal of Experimental Psychology: Learning, Memory, and Cognition, 24(1), 161–172.

67.

Slamecka

N. J.

Graf

(1978). The generation effect: Delineation of a phenomenon. Journal of Experimental Psychology: Human Learning and Memory, 4(6), 592–604.

68.

Smith

S. A.

Mulligan

N. W.

(2018). Distinctiveness and the attentional boost effect. Journal of Experimental Psychology: Learning, Memory, and Cognition, 44(9), 1464–1473.

69.

Spataro

Mulligan

N. W.

Bechi Gabrielli

Rossi-Arnaud

(2017). Divided attention enhances explicit but not implicit conceptual memory: An item-specific account of the attentional boost effect. Memory, 25(2), 170–175. https://doi.org/10.1080/09658211.2016.1144769

70.

Spataro

Mulligan

N. W.

Rossi-Arnaud

(2013). Divided attention can enhance memory encoding: The attentional boost effect in implicit memory. Journal of Experimental Psychology: Learning, Memory, and Cognition, 39(4), 1223–1231.

71.

Starns

J. J.

Rotello

C. M.

Ratcliff

(2012). Mixing strong and weak targets provides no evidence against the unequal-variance explanation of ʐROC slope: A comment on Koen and Yonelinas (2010). Journal of Experimental Psychology: Learning, Memory, and Cognition, 38(3), 793–801.

72.

Swallow

K. M.

Atir

(2018). The role of value in the attentional boost effect. Quarterly Journal of Experimental Psychology, 72(3), 523–542. https://doi.org/10.1177/1747021818760791

73.

Swallow

K. M.

Jiang

Y. V.

(2010). The attentional boost effect: Transient increases in attention to one task enhance performance in a second task. Cognition, 115(1), 118–132.

74.

Swallow

K. M.

Jiang

Y. V.

(2011). The role of timing in the attentional boost effect. Attention, Perception, & Psychophysics, 73(2), 389–404. https://doi.org/10.3758/s13414-010-0045-y

75.

Swallow

K. M.

Jiang

Y. V.

(2013). Attentional load and attentional boost: A review of data and theory. Frontiers in Psychology, 4, Article 274. https://doi.org/10.3389/fpsyg.2013.00274

76.

Swallow

K. M.

Jiang

Y. V.

(2014). Perceptual load and attentional boost: A study of their interaction. Journal of Experimental Psychology: Human Perception and Performance, 40(3), 1034–1045. https://doi.org/10.1037/a0035312

77.

Thunell

Thorpe

S. J.

(2019). Memory for repeated images in rapid-serial-visual-presentation streams of thousands of images. Psychological Science, 30(7), 989–1000.

78.

Tulving

Murray

(1985). Elements of episodic memory. Canadian Psychology, 26(3), 235–238.

79.

Tulving

Thompson

D. M.

(1973). Encoding specificity and retrieval processes in episodic memory. Psychological Review, 80(5), 352–373.

80.

Turker

Swallow

K. M.

(2018). Relational memory is enhanced in the attentional boost effect. Journal of Vision, 17(10), 876.

81.

Uncapher

M. R.

Otten

L. J.

Rugg

M. D.

(2006). Episodic encoding is more than the sum of its parts: An FMRI investigation of multifeatural contextual encoding. Neuron, 52(3), 547–556.

82.

Walker

Ciraolo

Dewald

Sinnett

(2017). Differential processing for actively ignored pictures and words. PLOS ONE, 12(1), Article e0170520.

83.

Wixted

J. T.

(2007). Dual-process theory and signal-detection theory of recognition memory. Psychological Review, 114(1), 152–176. https://doi.org/10.1037/0033-295X.114.1.152

84.

Xue

Dong

Chen

Mumford

J. A.

Poldrack

R. A.

(2010). Greater neural pattern similarity across repetitions is associated with better memory. Science, 330(6000), 97–101.

85.

Yonelinas

A. P.

(1994). Receiver-operating characteristics in recognition memory: Evidence for a dual-process model. Journal of Experimental Psychology: Learning, Memory and Cognition, 20(6), 1341–1354. https://doi.org/10.1037/0278-7393.20.6.1341

86.

Yonelinas

A. P.

(2001). Consciousness, control, and confidence: The 3 Cs of recognition memory. Journal of Experimental Psychology: General, 130(3), 361–379.

87.

Yonelinas

A. P.

(2002). The nature of recollection and familiarity: A review of 30 years of research. Journal of Memory and Language, 46(3), 441–517. https://doi.org/10.1006/jmla.2002.2864

88.

Yonelinas

A. P.

Aly

Wang

W.-C.

Koen

J. D.

(2010). Recollection and familiarity: Examining controversial assumptions and new directions. Hippocampus, 20(11), 1178–1194.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.90 MB