Encoding of episodic information through fast task-irrelevant perceptual
learning
Virginie Leclercq
a,b, Christophe C. Le Dantec
a, Aaron R. Seitz
a,⇑ aDepartment of Psychology, University of California – Riverside, 900 University Avenue, Riverside, CA 92521, USA b
INSHEA, Grhapes (EA 7287), Suresnes, France
a r t i c l e
i n f o
Article history:Received 15 May 2013
Received in revised form 13 September 2013 Available online 23 September 2013 Keywords:
Perceptual learning Encoding Remember/know
a b s t r a c t
The mechanisms guiding our learning and memory processes are of key interest to human cognition. While much research shows that attention and reinforcement processes help guide the encoding process, there is still much to know regarding how our brains choose what to remember. Recent research of task-irrelevant perceptual learning (TIPL) has found that information presented coincident with important events is better encoded even if participants are not aware of its presence (seeSeitz & Watanabe, 2009). However a limitation of existing studies of TIPL is that they provide little information regarding the depth of encoding supported by pairing a stimulus with a behaviorally relevant event. The objective of this research was to understand the depth of encoding of information that is learned through TIPL. To do so, we adopted a variant of the ‘‘remember/know’’ paradigm, recently reported byIngram, Mickes, and Wixted (2012), in which multiple confidence levels of both familiar (know) and remember reports are reported (Experiment 1), and in which episodic information is tested (Experiment 2). TIPL was found in both experiments, with higher recognition performance for target-paired than for distractor-paired images. Furthermore, TIPL benefitted both ‘‘familiar’’ and ‘‘remember’’ reports. The results of Experiment 2 indicate that the most confident ‘‘remember’’ response was associated with episodic information, where participants were able to access the location of image presentation for these items. Together, these results indicate that TIPL results in a deep enhancement in the encoding of target-paired information.
Ó2013 Elsevier Ltd. All rights reserved.
1. Introduction
Memory is a limited resource (Miller, 1956). We are unable to encode and store all the information present in the environment, and such exhaustive memorization would lead to difficulties in effectively utilizing stored information quickly and effectively to guide behavior. While people often want a memory system that follows their direction and stores information that they deem important, such as ‘‘my keys are on the dresser’’, it is well known that memory is not so obedient, ‘‘where are my keys and I can’t get that Barney song out of my head’’. While much research shows that attention and reinforcement processes help guide the encod-ing process (Broadbent, 1958; Cowan, 1988; Craik et al., 1996; Seitz, Lefebvre, et al., 2005; Seitz & Watanabe, 2003, 2005), there is still much to know regarding how our brains choose what to remember.
Recent research has found that stimuli presented at temporally-coincident times with important events are better encoded even if participants are not aware of their presence (seeSeitz & Watanabe, 2009). For example, stimuli presented with a task-target are better
learned than those presented with task distractors (Dewald, Sinnett, & Doumas, 2011; Leclercq & Seitz, 2012a, 2012b, 2012c, 2012d; Lin et al., 2010; Swallow & Jiang, 2010, 2011). This effect was called the task irrelevant perceptual learning (TIPL). TIPL has been observed in different learning paradigms. It has been studied in detail in the case of low-level perceptual learning (Seitz & Watanabe, 2005; see alsoSeitz & Watanabe, 2009for a review), and more recently for perceptual memories with the study of a fast form of TIPL (fast-TIPL) (Leclercq & Seitz, 2012a, 2012b, 2012c, 2012d). According to these studies, learning and memory is supe-rior for stimuli that are correlated with important events whether or not these stimuli have been deemed ‘‘relevant’’ to the behavior. However a limitation of existing studies of fast-TIPL is that they provide little information regarding the depth of encoding sup-ported by pairing a stimulus with a behaviorally relevant event. For example, the superior memorization can be accounted for either because some features of the target-paired images are more salient (Perceptual Learning account), because the target-paired images are more familiar (Familiarity account), or because the tar-get-paired images contained episodic information (Episodic ac-count). In the Perceptual Learning account viewers may not remember the target-paired images, per se, however, they report images as being familiar when some features of the images seem 0042-6989/$ - see front matterÓ2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.visres.2013.09.006 ⇑ Corresponding author. Fax: +1 (951) 827 3985.
E-mail address:[email protected](A.R. Seitz).
Contents lists available atScienceDirect
Vision Research
more salient than some baseline. In the Familiarity account (Tulving, 1985), the target-paired images may be better encoded in a semantic memory, without any episodic information regarding the encoding experience (i.e. no memory of the screen-location of the image when encoded). Finally, in the Episodic account (Tulving, 1985), there may be some memory of the encoding episode (e.g. remembering screen-location of the image or where it was within the image stream).Tulving (1985)suggested that memory types could be dissociated through the questions used to probe the memory; such requiring the observer to report whether they are ‘‘familiar’’ with or ‘‘remember’’ an object. Such approaches are commonly used to dissociate between memory systems in the brain and have led to the dual-process theories of recognition memory (Atkinson & Juola, 1973, 1974; Hintzman & Curran, 1994; Jacoby, 1991; Jacoby & Dallas, 1981; Mandler, 1980; Wixted, 2007; Yonelinas, 1994). While the dual-process theories are con-troversial (Donaldson, 1996; Dunn, 2004, 2008) and conclusions based on dissociations of memory reports must be considered carefully, using multiple memory reports with confidence scales can provide a useful approach to understand the memory processes.
Accordingly, we chose to adopt a method fromIngram, Mickes, and Wixted (2012)as a useful framework to better understand the depth of memory that could be elicited through fast-TIPL (Leclercq & Seitz, 2012a). We first conducted Experiment 1 to understand the effect of the target-pairing on the memory and then replicated these results in Experiment 2 where we also tested for episodic information associated with remembered items. Both experiments showed that fast-TIPL boosted both remember and familiar judg-ments for target-paired items compared to distractor-paired items, and results of Experiment 2 indicated that fast-TIPL can facilitate encoding of episodic information associated with target-paired items.
2. Experiment 1
2.1. Methods 2.1.1. Participants
Seventy-five participants took part to the first experiment, but only 63 participants (41 females, 22 males; ages 18–36) were in-cluded in the data analysis. Participants were exin-cluded (n= 12) either because they failed to respond on the majority of trials in the Rapid Serial Visual Presentation (RSVP) task or because they had more false-positives than hits on the Image Recognition Task. Of note, when all participants are included in the analyses, none of the statistical effects change in significance. Participants gave writ-ten informed consent to participate in this experiment, which was approved by the Human Research Review Board of the University of California, Riverside. All participants reported normal or cor-rected-to-normal visual acuity and received course credit and financial compensation for the 40-min session.
2.1.2. Apparatus and stimuli
An Apple Power Mac G4 running Matlab (Mathworks, Natick, MA) and Psychtoolbox Version 3 (Brainard, 1997; Pelli, 1997) was used for stimulus generation and experiment control. Stimuli were presented on a 2200 CRT monitor with resolution of
16001200 and a refresh rate of 100 Hz. Participants sat with their eyes approximately 60 cm from the screen. The backgrounds of all displays were a mid-gray. Display items consisted of 700700 pixel (18.3 degrees of visual angle) photographs depict-ing natural or urban scenes from eight distinct categories (i.e., mountains, cityscapes, etc.). Images were obtained from the LabelMe Natural and Urban Scenes database (Oliva & Torralba,
2001) at 250250 pixels of resolution, then up-sampled to 700700 pixels of resolution.
2.1.3. Procedure
During this experiment, the participants were required to per-form successively a letter detection task and then an image detec-tion task.
2.1.3.1. Letter detection task. In each trial, a stream of 10 images was successively presented in the middle of the screen. Each im-age was presented 133 ms, followed by a blank ISI of 367 ms for a SOA of 500 ms (Fig. 1). A gray aperture (1 degree of visual angle and luminance of 92 cd/m2) was presented in the center of each image, thus centered in the middle of the screen. Each image was presented with a letter (0.75 degree of visual angle, Font Cou-rier, Size 32) in the middle of the gray aperture. This letter could be a distractor (black letter; luminance of 0.25 cd/m2) or a target (white letter; luminance of 250 cd/m2). Each letter had the same onset and offset times as the image with which it was paired. In each trial, 1 image out of the 10 presented was paired with a white target letter; the others 9 images were paired with black distractor letters. The white target letter could appear in position 3 to 8. The type of stimulus that an image coincided with (e.g. a target or a distractor) was held constant across the experiment. For one set of 120 images, 20 images were paired with the white letters (tar-gets) and the remaining 100 images were paired with black letters (distractors), also presented from position 3 to 8 (to control for primacy and recency effects). Image assignment to target and dis-tractor was random for each participant. A set of 80 filler images, for which no target was presented, was used with letters pre-sented in positions 1, 2, 9 and 10. These fillers images were not tested in the image recognition test. Each image was presented 10 times during the entire experiment. Participants were asked to memorize the identity of the white letter and the images. At the end of each trial, participants pressed the key corresponding to the white letter. Participants performed a practice block of 12 trials. Each participant was then tested for a total of 200 trials, in 10 blocks of 20 trials. Breaks were given between blocks and subjects had to press the space bar on the keyboard to start the next one.
2.1.3.2. Image recognition task.At the end of the experiment, partic-ipants performed an image recognition task. Eighty images were presented to the participants: the 20 images paired with the target, 20 images paired with the distractors (randomly assigned for each participant) and 40 new images never presented in the experi-ment. One image was presented at a time until participants made their response. For each image, participants were asked to make an old/new decision about this image. To do so we used a rating scale with six levels (Ingram, Mickes, & Wixted, 2012; Experiment 2) illustrated inFig. 2. A response of 1–3 was used to indicate their confidence that the image was new, while a response of 4–6 was used to indicate their confidence that the image was old. In other words, 1 indicated the highest confidence that the image was new, and 6 indicated the highest confidence the image was old. Old ratings of 4–6 were further parsed by familiarity and remem-ber options (4F, 5F, and 6F; 4R, 5R, and 6R) where F means familiar and R means remember. Responses were recorded on the number pad and stickers were placed over the numbers 4–9 with 4–6 used for the R-scale and 7–9 for the F-scale. This scale provided a visual indication for the participants that remember judgments can also be made without the highest confidence. Participants made old– new decisions by clicking on a digit corresponding to their re-sponse. For the R or F responses, participants were told that they have to respond R if they can remember some qualitative informa-tion about the item (such as recollecting what you thought about
when the item was first presented and the item order during the presentation), and to respond F if they were unable to recall details about its occurrence. At the beginning of the experiment, partici-pants were told to memorize the images, and that an image recog-nition task would be performed at the end of the experiment.
2.2. Results
The performance on the white letter detection task was quite good with average performance of 96.8% (±0.3% SE). This demon-strates that participants properly attended to the center of the screen and complied with the instructions of the letter detection task.
For the image recognition task, we first examined overall accu-racy at reporting whether an image was ‘‘new’’ or ‘‘old’’ (Fig. 3). We found that recognition of target-paired images (68 ± 2.5% SE) was significantly better (p< .0001; paired t-test) than that of distractor-paired images (44.5 ± 1.9%) with correct rejections of new images at 62.7% (±1.4%). These data are consistent with previous findings demonstrating an advantage for target-paired images in fast-TIPL (Leclercq & Seitz, 2012a, 2012b, 2012c, 2012d).
We next asked whether the advantage for target-pairing could be linked to an advantage in familiarity, remembering or both. To address this we examined the percentage of response for each re-sponse type (Fig. 4) for target-paired and distractor paired images.
Fig. 1.Design of Experiment 1. Participants had to memorize the identity of the white letter while also memorizing the images presented in RSVP (10 images per trial) and to rapidly press the correct key-letter in the keyboard when the screen ‘‘what was the white letter?’’ appeared.
Fig. 2.Rating scale used in Experiments 1 and 2. A rating of 6 corresponds to the highest level of confidence that the image was presented in the images stream (old) and is further broken down by a remember–familiar judgment. Ratings of 4 and 5 reflect lower levels of confidence and also require a remember–familiar judgment. A rating of 1 reflects the highest level of confidence that the image was not previously presented in the images stream (new). R means Remember; F means Familiar.
Fig. 3.Overall recognition performance for the image recognition task in Exper-iment 1. Error bars represent within subject standard error of the mean.
Fig. 4.Percentage of response in Experiment 1 for each response type (from 1 to 6) for target-paired and distractor-paired images. Error bars represent within subject standard error of the mean.
A three-way repeated measures ANOVA on Confidence (4– 6)Pairing (Target vs. Distractor)Knowledge (Familiar vs. Remember) showed significant main effects for all three conditions (F(62, 1) = 23.9, p< .0001; F(62, 1) = 6.0, p= .017; F(62, 1) = 14.5, p= .0003, respectively) as well as significant interactions at all lev-els (allp6.0001). Following this up with separate 2-way ANOVAs on Confidence (4–6)Pairing (Target vs. Distractor) separately for Familiar and Remember reports showed main effects for Confi-dence and Pairing for both Familiar and Remember reports (all p6.0001), but while there was an interaction between Confidence and Pairing for Remember reports (F(124, 2) = 20.7,p< .0001) there was no such interaction for Familiar reports (F(124, 2) = 0.84, p= .43). While the effects of target-pairing on Familiarity was lar-gely consistent across confidence level, the effect of target-pairing on Remember ratings was only apparent for the highest confidence rating (p< .0001, for rating of 6 compared to distractor-paired images). These results suggest that TIPL advantages memory re-ports across the memory scale and support for the Episodic Account.
While we find the greatest effect of fast-TIPL on the Remember-6 responses, without an additional test, it is unclear whether par-ticipants really recalled any episodic information related to those images. To test this, Experiment 2 was conducted, where 2 images were presented with each letter (one on the left and one on the right) and participants were given a surprise test image location after making the old/new judgments.
3. Experiment 2
3.1. Methods 3.1.1. Participants
Seventy-five participants took part in Experiment 2, but only 52 participants (24 females, 28 males; ages 18–29) were included in the data analysis. Participants were excluded (n= 23) either be-cause they failed to respond on the majority of trials in the RSVP task or because they had more false-positives than hits on the Im-age Recognition Task. The failure-rate was higher than in Experi-ment 1 because the image recognition task was much more difficult, with two images (half the size as in Experiment 1) pre-sented at once, at a smaller size and in the periphery. Of note, when all participants are included in the analyses none of the statistical effects change with the exception that the interaction between Pairing and Memory type falls below significance. Participants gave written informed consent to participate in this experiment, which was approved by the Human Research Review Board of the Univer-sity of California, Riverside. All participants reported normal or cor-rected-to-normal visual acuity and received course credit and financial compensation for the 40-min session.
3.1.2. Apparatus and stimuli
Same as described in Experiment 1. 3.1.3. Procedure
3.1.3.1. Letter detection task. Same as described in Experiment 1, but two images were paired with each character: one to the left and one to the right (Fig. 5). Thus, instead of 10 images, 20 images were presented in each trial. Image size was 512512 pixels (13.6°of visual angle), and each pair of images was separated by a visual angle of 3°, leading to a global visual angle of 30.2°.
As in Experiment 1, for a set of 120 images, 20 were paired with the target, 10 were always presented on the left and the other 10 were always presented on the right. The other 100 images were presented with the distractors, 50 images were presented only on the left side and the other 50 always presented on the right side.
Finally, 40 on the 80 filler images were presented on the left and the other 40 on the right. In this experiment each image was pre-sented 20 times. Each participant was tested for a total of 200 trials presented in 10 blocks. Breaks were given between blocks and sub-jects had to press the space bar on the keyboard to start the next one.
3.1.3.2. Image recognition task. Same as described in Experiment 1, but after each ‘old’ response, subjects were required to report if the image had been presented on the left or the right by pressing the ‘left-arrow’ or the ‘right-arrow’ key. Trials in which an incorrect key, or response timeout (after 10 s) were ignored in the analyses (1.9% of trials were removed from the analysis for this reason). At the beginning of the experiment, participants were told to memo-rize the images, and that an image recognition task would be per-formed at the end of the experiment, however they were not instructed that the side was an important information.
3.2. Results
Performance on the white letter detection task was quite good with average performance of 96.5% (±0.4% SE), which was highly similar to the performance observed in Experiment 1. These data demonstrate that participants properly attended to the center of the screen and complied with the instructions of the letter detec-tion task.
For the image recognition task, we first examined overall accu-racy at reporting whether an image was ‘‘new’’ or ‘‘old’’ (Fig. 6). We found that recognition of target-paired images (59.6 ± 2.5% SE) was significantly better (p< .0001; pairedt-test) than that of distractor-paired images (45.0 ± 1.9%) with correct rejections of new images at 61.9 ± 2.0%. These data are highly consistent with those from Experiment 1, with the only notable difference being a lower-rec-ognition rate for the target-paired images, however, this can be ex-plained by the fact that in Experiment 2 two images were presented at a time, at a smaller size and in a peripheral location compared to Experiment 1.
We next asked whether the advantage for target-pairing could be linked to an advantage in familiarity, remembering or both. To address this we examined the percentage of report for each re-sponse type (Fig. 7). A three-way repeated measures ANOVA on Confidence (4–6)Pairing (Target vs. Distractor)Knowledge (Familiar vs. Remember) showed significant main effects for Pair-ing and Knowledge (F(51, 1) = 14.5, p< .0004; F(51, 1) = 102.0, p< .0001) but not for Confidence. However, here there was an interaction between Confidence and Knowledge (F(102, 2) = 3.9, p< .022) with no other interaction near significance (allpP.36). Following this up with separate 2-way ANOVAs on Confidence (4–6)Pairing (Target vs. Distractor) separately for Familiar and Remember reports showed main effects for Confidence and Pairing for both Familiar and Remember reports (allp6.01), but no inter-actions. These results suggest that while there are differences be-tween confidence ratings bebe-tween Familiar and Remember reports that this effect is not significantly mediated by target-pair-ing. Of note, while Remember 6 reports were much lower than found in Experiment 1 (discussed above), as hypothesized, these were still significantly higher for target-paired compared to dis-tractor-paired images (p< .022).
The key purpose of Experiment 2 was to ascertain whether any episodic information was present when participants reported that they ‘‘remembered’’ the images. To accomplish this we asked par-ticipants to report the location on the screen (left or right) at which each image appeared. Notably, participants were not informed un-til after the end of the exposure period that they would be tested on the location of image appearance. InFig. 8, we can see that localization accuracy for target-paired images rated with a 6 was
much better than that for similarly rated distractor-paired images or images rated to with lower confidence. An two-way repeated measures ANOVA on Confidence vs. Pairing showed no main ef-fects (pP.35) but a significant interaction (F(102, 2) = 10.9, p= .0001). Furthermore at-test between target-paired and distrac-tor-paired items rated with a 6 shows a significant effect (p< .008). These results provide further support for the Episodic Account, where episodic information is associated with remembered but not for familiar images and that this effect is boosted through TIPL.
4. Discussion
The objective of this research was to understand the depth of encoding of information that is learned in TIPL. To do so, we adopted a memory scale recently reported by Ingram, Mickes, and Wixted (2012), in which multiple confidence levels of both familiar and remember reports are reported (Experiment 1), and Fig. 5.Design of Experiment 2. Participants had to memorize the identity of the white letter while also memorizing the images presented in RSVP (20 images per trial) and to rapidly press the correct key-letter in the keyboard when the screen ‘‘what was the white letter?’’ appeared.
Fig. 6.Overall recognition performance for the image recognition task in Exper-iment 2. Error bars represent within subject standard error of the mean.
Fig. 7.Percentage of response in Experiment 2 for each response type (from 1 to 6) for target-paired and distractor-paired images. Error bars represent within subject standard error of the mean.
Fig. 8.percentage of correct response for the location task represented for each type of response from 4 to 6 for the target-paired remembered and target-paired familiar. Error bars represent within subject standard error of the mean.
in which episodic information is tested (Experiment 2). TIPL was found in both experiments, with higher recognition performance for target-paired than for distractor-paired images. Furthermore, TIPL benefitted both reports of ‘‘familiar’’ and ‘‘remember’’. The re-sults of Experiment 2 indicate that the remember response ‘‘6’’ was associated with episodic information, where participants were able to access the location of image presentation for these items. To-gether, these results indicate that TIPL results in a deep enhance-ment in the memorization of target-paired information.
While the current experiment utilized an experimental para-digm related to the debate over the dual process theories of mem-ory of recognition (Atkinson & Juola, 1973, 1974; Dunn, 2004, 2008; Wixted, 2007; Yonelinas, 1994), the results are equivocal regarding the existence of separate memory systems. Notably the benefit of target-pairing was found both for familiar and remem-bered images and this benefit was found across the rating scales. Given the broad benefit of TIPL across the memory scale, we cannot conclude that TIPL differently affects one aspect of memory com-pared to another. Furthermore it is possible that there is some spread across the scale is due to procedural details of our experi-ment where weakly ‘‘remembered’’ items were rated as familiar or vice-versa. For example, after the first trial of the location task in Experiment 2, participants may have changed their criteria and been more cautious in making Remember responses. While we cannot rule out contamination of one scale with responses more appropriate to the other, or the contribution of meta-cogni-tive effects such as response strategy, the fact that episodic infor-mation was only found for the most confident responses suggests that there is some relationship between confident Remember re-sponses and episodic information for those tokens. Still, overall, our results are equally consistent with separate memory systems that are each benefited by TIPL (Wixted, 2007) or a single memory system where memory differs by strength (Donaldson, 1996; Dunn, 2004, 2008).
Our results are informative regarding what types of information can be learned through TIPL. TIPL has been most studied in low-le-vel perceptual learning and has been found for a variety of fea-tures; such as motion processing (Watanabe et al., 2002), orientation processing (Franko, Seitz, & Vogels, 2010; Nishina et al., 2007), critical flicker fusion thresholds (Seitz, Nanez, et al., 2005, 2006), contour integration (Rosenthal & Humphreys, 2010), auditory formant processing (Seitz et al., 2010), and phonetic pro-cessing (Vlahou, Protopapas, & Seitz, 2012; Vlahou, Seitz, & Protop-apas, 2009). However, while there are a number of reports of fast-TIPL enhancing memory (Dewald, Sinnett, & Doumas, 2011; Lecl-ercq & Seitz, 2012a, 2012b, 2012c, 2012d; Lin et al., 2010; Swallow & Jiang, 2010, 2011) these studies have not addressed what is memorized in these tasks. For example, it is possible that prior studies of fast-TIPL relied upon participants being more sensitive to the features of the target-paired images and relying upon this to inform their reports of which images were new or old. The pres-ent study addresses this concern directly by showing that not only do participants report strong memories of, at least some of, the tar-get-paired images, but that they even remember episodic informa-tion (such as locainforma-tion of presentainforma-tion) of the best-remembered images.
More generally, our results show that TIPL is a general phenom-enon that boosts encoding/retention for information that is paired at times of behavioral relevance (Seitz & Watanabe, 2009). As dis-cussed above, TIPL benefits encoding of differing types of informa-tion ranging from basic stimulus features, to, now, episodic memory. This research is consistent with behaviorally relevant stimuli causing a momentary boost to encoding rate (Seitz & Dinse, 2007) that benefits whatever encoding processes are active at that point in time. Attention can either facilitate (Leclercq & Seitz, 2012b) or inhibit (Choi, Seitz, & Watanabe, 2009) encoding in a
selective manner depending upon the relevance of those stimuli to the participant. As a whole this research shows that one needs to both understand the reinforcement that a participant experi-ences and how attention is directed to understand what informa-tion will be remembered.
5. Conclusion
The present study shows that at behaviorally relevant times, such as recognizing a target, coincident information is better mem-orized. This boost to memory enhances both the extent to which participants gain familiarity with items as well as their ability to remember episodic information associated with these items, such as their location. This suggests that task-irrelevant learning bene-fits processes beyond just those of stimulus feature processing.
Acknowledgement
This study was funded by NSF (BCS-1057625) to ARS. The authorswould like to thank Vivian Tran who helped run partici-pants in thesestudies.
References
Atkinson, R. C., & Juola, J. F. (1973). Factors influencing the speed and accuracy of word recognition. In S. Kornblum (Ed.), Attention and performance IV (pp. 583–612). New York: Academic Press.
Atkinson, R. C., & Juola, J. F. (1974). Search and decision processes in recognition memory. In D. H. Krantz, R. C. Atkinson, & P. Suppes (Eds.),Contemporary developments in mathematical psychology (pp. 243–290). San Francisco: Freeman.
Brainard, D. H. (1997). The psychophysics toolbox.Spatial Vision, 10(4), 433–436. Broadbent, D. E. (1958).Perception and communication. London: Pergamon Press. Choi, H., Seitz, A. R., & Watanabe, T. (2009). When attention interrupts learning:
Inhibitory effects of attention on TIPL.Vision Research, 49(21), 2586–2590. Cowan, N. (1988). Evolving conceptions of memory storage, selective attention, and
their mutual constraints within the human information-processing system. Psychological Bulletin, 104, 163–191.
Craik, F. I. M., Govoni, R., Naveh-Benjamin, M., & Anderson, N. D. (1996). The effects of divided attention on encoding and retrieval processes in human memory. Journal of Experimental Psychology: General, 125, 159–180.
Dewald, A. D., Sinnett, S., & Doumas, L. A. A. (2011). Conditions of directed attention inhibit recognition performance for explicitly presented target-aligned irrelevant stimuli.Acta Psychologica, 138, 60–67.
Donaldson, W. (1996). The role of decision processes in remembering and knowing. Memory & Cognition, 24, 523–533.
Dunn, J. C. (2004). Remember–know: A matter of confidence.Psychological Review, 111(2), 524–542.
Dunn, J. C. (2008). The dimensionality of the remember–know task: A state-trace analysis.Psychological Review, 115(2), 426–446.
Franko, E., Seitz, A. R., & Vogels, R. (2010). Dissociable neural effects of long term stimulus-reward pairing in macaque visual cortex. Journal of Cognitive Neuroscience, 22(7), 1425–1439.
Hintzman, D. L., & Curran, T. (1994). Retrieval dynamics of recognition and frequency judgments: Evidence for separate processes of familiarity and recall. Journal of Memory and Language, 33, 1–18.
Ingram, K. M., Mickes, L., & Wixted, J. T. (2012). Recollection can be weak and familiarity can be strong.Journal of Experimental Psychology: Learning, Memory, and Cognition, 38, 325–339.
Jacoby, L. L. (1991). A process dissociation framework: Separating automatic from intentional uses of memory.Journal of Memory and Language, 30, 513–541. Jacoby, L. L., & Dallas, M. (1981). On the relationship between autobiographical
memory and perceptual learning.Journal of Experimental Psychology: General, 3, 306–340.
Leclercq, V., & Seitz, A. R. (2012a). Fast task-irrelevant perceptual learning is disrupted by sudden onset of central task elements.Vision Research, 15(61), 70–76 (Epub 2011 July 24).
Leclercq, V., & Seitz, A. R. (2012b). Enhancement from targets and suppression from cues in fast task-irrelevant perceptual learning. Acta Psychologica, 141(1), 31–38.
Leclercq, V., & Seitz, A. R. (2012c). The impact of orienting attention in fast task-irrelevant perceptual learning. Attention, Perception, & Psychophysics, 74(4), 648–660.
Leclercq, V., & Seitz, A. R. (2012d). Fast-TIPL occurs for salient images without a memorization requirement in men but not in women.PLoS ONE, 7(4), e36228. 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), e1000337.
Mandler, G. (1980). Recognizing: The judgment of previous occurrence. Psychological Review, 87, 252–271.
Miller, G. A. (1956). The magical number seven, plus or minus two: Some limits on our capacity for processing information.Psychological Review, 63, 81–97. Nishina, S., Seitz, A. R., Kawato, M., & Watanabe, T. (2007). Effect of spatial distance
to the task stimulus on task-irrelevant perceptual learning of static Gabors. Journal of Vision, 7(13), 1–10.
Oliva, A., & Torralba, A. (2001). Modeling the shape of the scene: A holistic representation of the spatial envelope.International Journal of Computer Vision, 42(3), 145–175.
Pelli, D. G. (1997). The VideoToolbox software for visual psychophysics: Transforming numbers into movies.Spatial Vision, 10(4), 437–442.
Rosenthal, O., & Humphreys, G. W. (2010). Perceptual organization without perception: The subliminal learning of global contour.Psychological Science, 21(12), 1751–1758.
Seitz, A. R., & Dinse, H. R. (2007). A common framework for perceptual learning. Current Opinion Neurobiology, 17(2), 148–153.
Seitz, A. R., Lefebvre, C., Watanabe, T., & Jolicoeur, P. (2005). Requirement for high-level processing in subliminal learning.Current Biology, 15(18), R753–R755. Seitz, A. R., Nanez, J. E., Holloway, S. R., & Watanabe, T. (2005). Visual experience can
substantially alter critical flicker fusion thresholds.Human Psychopharmacology Clinical and Experimental, 20(1), 55–60.
Seitz, A. R., Nanez, J. E., Holloway, S. R., & Watanabe, T. (2006). Perceptual learning of motion leads to faster flicker perception.PLoS ONE, 1(1), e28.
Seitz, A. R., Protopapas, A., Tsushima, Y., Vlahou, E. L., Gori, S., Grossberg, S., et al. (2010). Unattended exposure to components of speech sounds yields same benefits as explicit auditory training.Cognition, 115(3), 435–443.
Seitz, A. R., & Watanabe, T. (2003). Psychophysics: Is subliminal learning really passive?Nature, 422, 36.
Seitz, A. R., & Watanabe, T. (2005). A unified model for perceptual learning.Trends in Cognitive of Sciences, 9, 329–334.
Seitz, A. R., & Watanabe, T. (2009). The phenomenon of task-irrelevant perceptual learning.Vision Research, 49(21), 2604–2610.
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.
Swallow, K. M., & Jiang, Y. V. (2011). The role of timing in the attentional boost effect.Attention, Perception & Psychophysics, 73(2), 389–404.
Tulving, E. (1985). How many memory systems are there?American Psychologist, 40(4), 385–398.
Vlahou, E., Seitz, A. R., & Protopapas, A. (2009). Implicit learning of nonnative speech stimuli.Paper presented at the Acoustical Society of America, Portland, OR. Vlahou, E., Protopapas, A., & Seitz, A. R. (2012). Implicit training of nonnative speech
stimuli.Journal of Experimental Psychology: General, 141(2), 363–381. Watanabe, T., Nanez, J. E., Sr., Koyama, S., Mukai, I., Liederman, J., & Sasaki, Y. (2002).
Greater plasticity in lower-level than higher-level visual motion processing in a passive perceptual learning task.Nature Neuroscience, 5(10), 1003–1009. Wixted, J. T. (2007). Dual-process theory and signal-detection theory of recognition
memory.Psychological Review, 114, 152–176.
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, 1341–1354.