Signal Validity Protects a Target Stimulus From the Attentional Blink

Journal of Experimental Psychology:Learning, Memory, and Cognition2009, Vol. 35, No. 2, 408 – 422© 2009 American Psychological Association0278-7393/09/$12.00 DOI: 10.1037/a0014525Attentional Changes During Implicit Learning: Signal Validity Protects aTarget Stimulus From the Attentional BlinkEvan J. Livesey, Irina M. Harris, and Justin A. HarrisThis document is copyrighted by the American Psychological Association or one of its allied publishers.This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.University of SydneyParticipants in 2 experiments performed 2 simultaneous tasks: one, a dual-target detection task within arapid sequence of target and distractor letters; the other, a cued reaction time task requiring participantsto make a cued left–right response immediately after each letter sequence. Under these rapid visualpresentation conditions, it is usually difficult to identify the 2nd target when it is presented in temporalproximity of the 1st target—a phenomenon known as the attentional blink. However, here participantsshowed an advantage for detecting a target presented during the attentional blink if that target predictedwhich response cue would appear at the end of the trial. Participants also showed faster reaction timeson trials with a predictive target. Both of these effects were independent of conscious knowledge of thetarget–response contingencies assessed by postexperiment questionnaires. The results suggest thatimplicit learning of the association between a predictive target and its outcome can automaticallyfacilitate target recognition during the attentional blink and therefore shed new light on the relationshipbetween associative learning and attentional mechanisms.Keywords: predictive learning, attentional blink, signal validity, implicit learningindirect and inferred through observations that the learned behavior is generally consistent with the predictions of these models.Partly for this reason, learning theorists have conventionallyadopted broad definitions of attention and attentional change,which in its most general sense simply refers to any change in theprocessing of a CS.Learning a relationship between a conditioned stimulus (CS)and an outcome that it predicts is often assumed to be accompaniedby changes in attention. Some models of associative learning (e.g.,Kruschke, 2001; Mackintosh, 1975) propose that changes in attention are dictated by the relative utility of the various predictivesignals that one might extract from presented stimuli: Those features that are relatively good predictors of an outcome attractattention, whereas relatively poor predictors lose attention. Learning about the signal validity of a CS, the extent to which it signalsthe occurrence of a relevant outcome, thus results in a change inthe processing of that CS during later learning episodes. This ideahas received support from a wide variety of animal and humanexperiments (see Le Pelley, 2004, for a recent review). Much ofthe evidence in support of these proposed attentional changes hasemerged from studies of predictive or discrimination learning, inwhich the principal behavioral measure is the rate at which discrimination accuracy increases or associations between events areconditioned. Such evidence cannot easily separate changes inlearning rate from other changes in performance. Thus evidencefor a particular attentional mechanism, or even a general theoretical principle about attention and learning, has typically beenLearning and Attentional ChangeThere is a diverse range of “attentional” processes that mightchange as a consequence of learning about a stimulus. Theseinclude overt attentional changes, such as orienting responses (e.g.,Sokolov, 1963) or changes in gaze direction (Deubel & Schneider,1996; Kowler, Anderson, Dosher, & Blaser, 1995) in response toa CS becoming meaningful, which have a direct impact on thephysical sampling of the stimulus. Even where the locus of attention diverges from gaze direction (e.g., Posner, 1980), covertchanges in spatial attention might well operate in a similar fashion,as subjects may preferentially process information from a regionwhere a CS is expected or has recently occurred. Attentionalchange may also refer to changes in the relative share of limitedcapacity resources allocated to processing stimulus features according to their utility, in the sense that stimuli may compete forattention even in the absence of changes in stimulus sampling atthe sensory level. This form of selective attention underpins avariety of theories of discrimination learning and assumes thatthere are limits on the quantity of stimulus information that can beencoded, or learned about, at any given time (e.g., Sutherland &Mackintosh, 1971). In contrast, learning the predictive validity ofa stimulus may permit faster processing of that stimulus withoutexhausting limited resources, and this may occur in a way that isnot directly driven by capacity limitations and, consequently, doesnot require selective processing. Models that assign independentparameters to each stimulus to represent their attentional weightingEvan Livesey, Irina M. Harris, and Justin A. Harris, School of Psychology, University of Sydney, Sydney, Australia.This research was supported by Australian Research Council (ARC)Grant DP0771154 to Justin A. Harris, an ARC Australian PostdoctoralResearch Fellowship to Evan Livesey, and an ARC Queen Elizabeth IIResearch Fellowship to Irina M. Harris. The authors would like to thankBob Boakes, Luis Jimenez, and Andy Wills for comments on an earlier´version of this article.Correspondence concerning this article should be addressed to Evan Livesey, School of Psychology, University of Sydney, Griffith Taylor BuildingA19, Sydney NSW 2006, Australia. E-mail: evanl@psych.usyd.edu.au408This document is copyrighted by the American Psychological Association or one of its allied publishers.This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.ATTENTIONAL CHANGES DURING IMPLICIT LEARNINGessentially take this view, even if the rules governing attentionalchange are competitive or based on a comparison with the utilityof other stimuli.The Mackintosh (1975) model can be seen as an example of thislatter approach. However, these processes generally imply thatattention has the opportunity to influence the extent to which apreviously learned association manifests in the performance ofsome response, as well as influence new learning. In contrast,Mackintosh (1975) suggested that, without any convincing evidence to the contrary, the predictive utility of a stimulus should beassumed to affect only the rate at which that stimulus is learnedabout (i.e., its associability) rather than assume any further changesin performance that are normally associated with attention. Furthermore, there is a question of whether changes in associabilityare governed by the predictive validity of the CS (Mackintosh,1975) or the predictability of the unconditioned stimulus (US; e.g.,Pearce & Hall, 1980), both of which account for some but not allof the relevant findings in animal learning (Le Pelley, 2004). Thefact that these processes are not necessarily mutually exclusive,and may have opposing or additive effects on learning in differentcircumstances, makes the task of determining the relationshipbetween attention and learning all the more challenging.Recently, researchers have begun to focus on concomitant measures of stimulus processing to better characterize the attentionalchanges that occur during human predictive or category learning.These have included measures of gaze duration directed towardcompeting visual stimuli, which have generally shown (albeitunder different task requirements and testing different hypotheses)that participants spend more time fixating on informative stimulithat are relevant to the task outcomes than on redundant or irrelevant stimuli (Kruschke, Kappenman, & Hetrick, 2005; Rehder &Hoffman, 2005; Wills, Lavric, Croft, & Hodgson, 2007). Thesestudies demonstrate that learning about the significance of a stimulus has a directly observable effect on the overt physical samplingof that stimulus.Studies using neurophysiological measures have also suggestedthat predictive learning in humans is accompanied by changes inthe neural mechanisms involved in stimulus processing. For instance, Wills et al. (2007) found differences in the event-relatedpotentials evoked by the onset of stimuli that had previously beeneither uniquely predictive of a surprising outcome or completelyredundant to the occurrence of an expected outcome. Using functional magnetic resonance imaging, Ploghaus et al. (2000; see alsoO’Doherty, Dayan, Friston, Critchley, & Dolan, 2003; Turner etal., 2004) also found neural substrates that appeared to codeabsolute prediction error, that is, the mismatch between expectedoutcome and actual outcome regardless of whether that outcome isoverpredicted or underpredicted. On the whole, these studies provide evidence of changes in stimulus processing dictated, in somefashion, by the association between CS and US. It is hoped that anaccumulation of evidence from a range of measures that are notsolely dependent on the learned response will paint a clearerpicture of the attentional processes that change as a consequence oflearning in humans.The present study investigated the relationship between predictive learning and attention but with two important differences fromprevious work. The first of these was to use an independentbehavioral measure of stimulus processing sensitive to temporal409dynamics of visual processing. The second was the use of aspeeded response task in which learning is entirely incidental.The Attentional BlinkWe examined visual processing of target CSs under conditionsof rapid serial visual presentation (RSVP) and, in doing so, used aphenomenon commonly known as the attentional blink (AB;Raymond, Shapiro, & Arnell, 1992). The AB refers to the observation that identification of the second of two targets in an RSVPstimulus stream is usually impaired if it appears about 200 –500 msafter the presentation of the first target. The spacing of the twotargets in terms of their serial position in the RSVP sequence iscritical for the occurrence of the AB. With relatively simple andfamiliar visual stimuli, such as alphanumeric characters, the AB isoften found to be strongest when the targets occur close together intime and when there is at least one intervening distractor betweenthe targets (Olivers, van der Stigchel, & Hulleman, 2007).Several studies have shown that the semantic qualities of stimuliin the AB are at least partially processed even when target detection is inaccurate (Shapiro, Driver, Ward, & Sorensen, 1997;Visser, Merikle, & Di Lollo, 2005), and some studies also indicatethat the AB is sensitive to the learned significance of stimuliappearing as targets and distractors. For instance, familiar ormeaningful stimuli, such as one’s own name (Shapiro, Caldwell, &Sorensen, 1997) or famous faces (Jackson & Raymond, 2006), areprotected to some degree from the AB, as their identification is lessimpaired than similar but less familiar stimuli. Smith, Most, Newsome, and Zald (2006) also showed that a stimulus recently associated with an aversive burst of white noise can automaticallyinduce an AB-like impairment when used as a distractor in anRSVP task. Thus there is some evidence that the learned significance of a stimulus can affect the severity of, or even induce, anAB, although precisely what aspects of learning result in ABchanges in any given paradigm is yet to be explored systematically. Frequency of occurrence of targets within a paradigm hasbeen explored systematically: A high-probability target causes lessimpairment on the recognition of a subsequent target (i.e., areduced AB) compared with a low-probability target, suggestingthat an expected stimulus is more easily processed (Crebolder,Jolicœur, & McIlwaine, 2002). However, beyond the effects offamiliarity and frequency, the influence of stimulus significance onthe AB remains unclear.Implicit LearningWe employed a speeded response task in which learning proceeded incidentally, with participants given no instruction or feedback about the to-be-learned information. Participants completed alengthy series of dual-task trials, each comprising an RSVP sequence of target and distractor letters followed immediately by aresponse cue—a circle appearing on either the left or right side ofthe computer screen. On each trial, the participant observed theletter sequence, trying to detect the two target letters (which weredistinguishable by their color), then responded as quickly as possible to the left–right response cue with a corresponding key press.Having performed this speeded response, the participant thenreported the two letter targets. On a proportion of trials, a particular letter appeared during the RSVP sequence, and its appearanceThis document is copyrighted by the American Psychological Association or one of its allied publishers.This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.410LIVESEY, HARRIS, AND HARRISwas always followed by the same response cue. For instance, theappearance of letter P might always be followed by the leftresponse cue. The consistent pairing of P and left presents anopportunity for the speeded response to be performed faster thanwould normally be possible. That is, learning that P signals the leftresponse should improve the speed of responding on those trialswhere P appears.A response priming paradigm of this nature can essentially bedescribed as a form of Pavlovian conditioning, in which a conventional US is replaced by an imperative stimulus that requires avoluntary speeded response (Perruchet, Cleeremans, & Destrebecqz, 2006). In this case, a predictive letter CS is consistentlypaired with a response cue US. A speeded left or right key press ismade in response to the appearance of this US, so that correctresponding is not contingent on learning the CS–US relationship,but responses may nonetheless be primed by the appearance of theCS if learning takes place. Pavlovian conditioning in animals hasbeen an important test bed for attentional models of associativelearning and has yielded several phenomena that strongly suggestchanges in attention to the CS (e.g., in relation to unblocking;Dickinson, Hall, & Mackintosh, 1976; Dickinson & Mackintosh,1979; Holland, 1984, 1988; Mackintosh & Turner, 1971). Overtattentional changes during Pavlovian conditioning, such as orienting responses made by rats to the CS, also appear to conform topredicted changes in CS processing (Kaye & Pearce, 1984). However, it is not clear whether incidental learning in humans cansustain changes in attention in quite the same fashion. Attentionalstudies looking specifically at the effects of learning stimulusrelevance have employed intentional learning, in which feedbackis dependent on the responses made to the CS and participantsactively seek solutions to the task at hand in a trial-and-errorfashion. In such studies, attending to relevant stimulus dimensionsor previously predictive stimuli is of some obvious benefit to theparticipant, such as allowing improved performance (e.g., Rehder& Hoffman, 2005) or decreasing attentional load without sacrificing performance (e.g., the blocking studies by Kruschke et al.,2005). But should one expect to see more automatic changes in CSprocessing that accompany associative learning even if thosechanges have no obvious or direct benefit to the human subject? Amore specific and potentially more controversial variant of thesame question might be to ask whether attentional changes accompany implicit learning of CS–US contingencies.Implicit learning typically describes learning that occurs incidentally, in the absence of conscious hypothesis testing and conscious rule abstraction, and yields knowledge that does not necessarily require conscious thought processes (e.g., Shanks & St.John, 1994; for a recent summary, see Perruchet, 2008). To date,studies of the relationship between implicit learning and attentionhave been concerned with very different aspects of attention, suchas attentional load and the effects of instructed selective attention.The former concerns the effect of cognitive capacity limitations onimplicit learning (Frensch, Buchner, & Lin, 1994; Jimenez &´Mendez, 1999; Shanks & Channon, 2002; Shanks, Rowland, &´Ranger, 2005), whereas the latter examines whether voluntaryselective attention to a particular task determines what can belearned implicitly (Jiang & Chun, 2001; Jiang & Leung, 2005;Jimenez & Mendez, 1999). Indeed there is some disagreement´´over whether implicit learning should be defined as learning thatdoes not load on limited attentional resources, rather than in termsof awareness or conscious intent (e.g., Cleeremans, 1997; Frensch,Lin, & Buchner, 1998; Shanks & Channon, 2002). In contrast, thepresent experiments primarily investigated the effect of implicitlearning on attentional processing of the CS, particularly in relation to the AB. Nonetheless, the effect of attentional selection ofthe CS on learning was also examined by comparing learning totargets and distractors. As discussed below, these comparisonsbear some resemblance to studies of the effect of voluntary selective attention on implicit learning.Although most implicit learning studies that employ reactiontime (RT) as a key performance measure concern the learning ofcomplex sequences (e.g., Willingham, Nissen, & Bullemer, 1989),we have employed a Pavlovian conditioning approach with asimple CS–US contingency, in general agreement with the rationale outlined by Boakes, Roodenrys, and Barnes (1995). In particular, conscious knowledge of a simple CS–US relationship canbe assessed with a high degree of confidence and sensitivity, andthe learning of the association itself is theoretically tractable,which is particularly important in this case for relating learning toattentional change. We return to the issue of contingency awareness in the General Discussion. For the time being, it will sufficeto say that we intend to demonstrate changes in attention accompanying the acquisition of associative priming in conditions thatmake it unlikely that participants acquire conscious knowledge ofthe relevant contingencies that might otherwise sustain a consciousintent to search for the predictive CSs. In this sense, the learningthat accompanies the attentional changes observed in the followingexperiments may be considered implicit in nature.Experiment 1Experiment 1 investigated whether learning could be shown toa predictive target and a predictive distractor and to gauge howlearning might affect target detection accuracy across differencesin serial position separating the two target letters (i.e., the lagbetween T1 and T2). A simple motor priming effect was taken asthe critical evidence for learning; that is, whether participantsresponded faster on trials that contained a target or distractor thatsignaled which response would be required than on control trialswith no predictive item. With this in mind, the choice of appropriate controls was extremely important because the frequency ofpresentation of items in RSVP has a clear effect on the ease withwhich an item is processed (e.g., Crebolder et al., 2002). Frequency differences could affect both target detection accuracy andthe speed with which a subsequent response is executed. Therefore, control trials contained a nonpredictive target or distractor,matched in each case to the frequency of presentation (and allother temporal characteristics of presentation) of the predictiveletters. As shown in Table 1, trials containing a predictive T2target (Tp) and trials containing a predictive distractor item (Dp)each made up 20% of the overall number of trials. Trials with thecontrol targets and distractors, referred to as Tf and Df, respectively (i.e., frequency matched controls), each occurred on another20% of trials. In Experiment 1, presentation of Tp was alwaysfollowed by a left response cue on the cued reaction time (CRT)task, whereas presentation of Dp was always followed by a rightresponse cue. In contrast, Tf and Df were followed by left and rightCRT cues 50% of the time and as such were completely nonpredictive. Because Tp always signaled left in this experiment, onlyATTENTIONAL CHANGES DURING IMPLICIT LEARNINGTable 1Trial Design for Experiment 1 Showing the Trial Sequence forEach of the Five Trial TypesTrial typeThis document is copyrighted by the American Psychological Association or one of its allied publishers.This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.TpTfDpDfRandomRSVP sequence#####ddddddddddT1T1T1T1T1ddddddddddddddddddddd T2 d # cued T2 d # cued T2 d # cued T2 d # cued T2 d # cue411Responses were made via a standard computer keyboard (the twoControl keys served as the left and right CRT response keys).Audio feedback for CRT responses was delivered through headphones.Left–right cue100% left50% left–50% right100% right50% left–50% right50% left–50% rightNote. In each rapid serial visual presentation (RSVP) sequence, d denotesa distractor presented in white, and T1 and T2 denote the two red targetletters. The four bold letters (T2 on Tp and Tf trials, d on Dp and Df trials)represent four preallocated letters that remained constant throughout theexperiment. All other letters (T1 and remaining T2 and d letters) wererandomly chosen on each trial (without replacement) from the pool ofremaining letters. The cue represents the left or right response cue appearing immediately after the hash mark (#) at the end of the RSVP sequence.T2 appeared equally often in serial Positions 8, 9, and 10 throughout theexperiment (as did the meaningful distractor on Dp and Df trials). T1appeared in Position 3 for the first 60 blocks and then equally often inPositions 3, 6, and 7 for the final 45 blocks. Each trial type occurredequally often (once per randomized block of trials).the Tf trials on which a left response was required were used tocalculate mean RTs. This avoids any issue of overall biases towardmaking left responses more quickly or slowly than right responses,which could produce artifactual RT differences between Tp andTf. Similarly, only Df trials with right responses were used tocompare with Dp trials.The design of the experimental trials allowed concurrent assessment of response speed on the CRT task and report accuracy of thetargets in the RSVP sequence. A postexperiment questionnaire,used to gauge the general level of conscious knowledge of thecontingencies between the predictive items and responses, followed immediately after the completion of the final experimentaltrial.MethodParticipantsThirty-four students at the University of Sydney participated inthe experiment in return for course credit. All were naive to theaims of the experiment. Exclusion criteria were established to ruleout participants who performed very poorly on either of the targetdetection or speeded response tasks. If a participant correctlyreported less than 40% of T1 targets, gave the incorrect CRTresponse on more than 20% of trials, or failed to respond within1,000 ms on more than 20% of correct CRT responses, the participant’s data were discarded. These criteria were used for bothexperiments. In Experiment 1, 2 participants were excluded forone or more of these criteria. All analyses were conducted on theremaining 32 participants.ApparatusParticipants were tested individually in a dimly lit room. Theexperiment was run on a Dell OptiPlex desktop computer with17-in. (43.18-cm) cathode ray tube monitor running at 85-Hzrefresh rate. Participants sat approximately 50 cm from the screen.Stimuli and DesignEach RSVP sequence consisted of serial presentation of 10uppercase letters (Arial font; point size ϭ 72), each appearing inthe center of the computer screen. Each sequence began and endedwith an additional hash (#) visual mask. Each letter appeared forapproximately 106 ms and was immediately replaced by the nextletter in the sequence. In each sequence, the two targets were red,and distractors were white, all against a black background. All…

Order your essay today and save 30% with the discount code: KIWI20