Hemispheric performance in object-based attention

Copyright 2004 Psychonomic Society Inc 84

Psychonomic Bulletin amp Review2004 11 (1) 84-91

In 1980 Posner and colleagues suggested that selectivevisual attention could be likened to a mental ldquospotlightrdquothat is committed to specific spatial locations (Posner1980 Posner Snyder amp Davidson 1980) Stimuli thatappear within the ldquobeamrdquo of the attentional spotlight areldquoilluminatedrdquo and processed more efficiently than stim-uli at unattended locations The behavioral result is thatresponse times (RTs) andor response errors are reducedfor stimuli at attended versus unattended positions

A few years later a seminal paper by Duncan (1984)reported that visual selective attention could also becommitted to specific objects as well as to specific loca-tions Duncan found that errors increased when subjectswere required to make a single judgment about two ob-jects as compared with when they were required to maketwo judgments about a single object Because the two ob-jects shared the same location in space a purely space-based theory of attention predicted that the two objectsshould have been processed as easily as one object Thefact that there was a cost in attending to two objects at thesame location demonstrated that selection could be ob-ject based as well as space based

More recent evidence has suggested that different brainregions may subserve space-based and object-based at-tention For instance in a particularly influential studyEgly Driver and Rafal (1994) measured both space-based

This work was supported by grants to AK from the the HumanFrontier Science Program the Natural Sciences and Engineering Re-search Council of Canada and the Michael Smith Foundation forHealth Research We thank David Balota John Wixted John Duncanand an anonymous reviewer for their comments Correspondenceshould be directed to A Kingstone 2136 West Mall Department ofPsychology University of British Columbia Vancouver BC V6T 1Z4Canada (e-mail alankingstoneubcca)

Hemispheric performance in object-based attention

MONICA A VALSANGKAR-SMYTHUniversity of Alberta Edmonton Alberta Canada

CHRISTA-LYNN DONOVANUniversity of British Columbia Vancouver British Columbia Canada

SCOTT SINNETTUniversity of Barcelona Barcelona Spain

MICHAEL R W DAWSONUniversity of Alberta Edmonton Alberta Canada

and

ALAN KINGSTONEUniversity of British Columbia Vancouver British Columbia Canada

The goal of the present study was to investigate whether object-based attention effects differ acrossthe cerebral hemispheres Previous research has suggested that object-based attention is preferentiallylateralized to the left hemisphere (Egly Driver amp Rafal 1994 Egly Rafal Driver amp Starrveveld 1994)However work by Vecera (1994) has suggested that these previous studies may have failed to obtain apure measure of object-based attention The present study applied modified versions of Duncanrsquos(1984) seminal object-based attention paradigm Subjects were typically presented with one target ob-ject to a single visual field (one-object display) two target objects to the same visual field (two-objectunilateral display) or two target objects to different visual fields (two-object bilateral display) In allthree experiments response accuracy was higher for the one-object displays than for the two-objectdisplays Most important this object-based cost was especially severe when selection of two target el-ements was isolated to the right visual field (left hemisphere) We confirmed that this effect was spe-cific to object-based attention in three different ways Experiment 1 manipulated stimulus distance asrecommended by Vecera Experiment 2 ensured that target selection was based on nonspatial attrib-utes and Experiment 3 used overlapping displays as in Duncan (1984) Collectively the data are in ac-cord with previous conclusions that object-based attention is a specialized form of orienting subservedby lateralized cortical brain mechanisms However contrary to previous research it appears that it isthe right hemisphere and not the left hemisphere that is preferentially biased for committing object-based attention to elements in the visual environment

RIGHT-HEMISPHERE OBJECT-BASED ATTENTION 85

and object-based attention in a single paradigm Parietallesion patients were presented with displays that con-sisted of two outline rectangles positioned either aboveand below central fixation or to the right and left of cen-tral fixation Attention was drawn to the end of one of therectangles by an abrupt peripheral flash (called a cue)Then a target requiring a detection response was pre-sented at one of three possible locations at the cued lo-cation at the noncued end of the cued rectangle or at anoncued end of the noncued rectangle It was expectedthat RTs to a target at a noncued location would be longerthan RTs to a target at a cued location because of the ad-ditional time required to shift attention from the cued lo-cation to the noncued location The question was whethera space-based attentional effect (shifting attention withinthe cued object [cued location vs noncued location withinthe same rectangle]) would differ from an object-basedattentional effect (shifting attention between the cued ob-ject and the noncued object [cued location vs noncued lo-cation on different rectangles]) Results suggested thatright-hemisphere lesion patients exhibited a space-basedattention deficit and left-hemisphere lesion patients exhibited an object-based attention deficit These resultswere interpreted as indicating that the left and the rightcerebral hemispheres are differentially specialized forspace-based (right-hemisphere) and object-based (left-hemisphere) attentional orienting

Egly Rafal Driver and Starrveveld (1994) conducteda similar experiment with a split-brain patient The par-adigm was the same as that in Egly Driver and Rafalrsquos(1994) study except that now there were always four rec-tangles presented to the subject two in each visual fieldWhen the targets were presented in the left visual field(LVF right hemisphere) there were no significant dif-ferences in the RT cost of shifting attention within an ob-ject and between objects When targets were presented inthe right visual field (RVF left hemisphere) howeverthe split-brain patient was much slower to respond whenshifts of attention between objects were required thanwhen shifts of attention within an object were requiredThe authors concluded that this study dovetailed with theevidence from the parietal lesion patientsmdashthat is theright hemisphere was specialized for space-based atten-tion and the left hemisphere was specialized for object-based attention

Vecera (1994) however has questioned this interpre-tation of the studies of Egly and colleagues (Egly Driveramp Rafal 1994 Egly Rafal et al 1994) SpecificallyVecera found that the object-based attention effect inthese studiesmdashshifting attention between a cued and anoncued rectanglemdashwas extremely sensitive to changesin the distance that separated the two rectangles

The implication of Vecerarsquos (1994) finding should notbe underestimated If the object-based attention effectsreported by Egly and colleagues (Egly Driver amp Rafal1994 Egly Rafal et al 1994) are sensitive to spatialmanipulations the hemispheric differences reported inEgly and colleaguesrsquo patient studies may merely reflect

differences in space-based attentional orienting and havelittle to do with object-based attentional orienting Thusthe question remains open as to whether the brain mech-anisms subserving object-based and space-based atten-tion are represented differentially between the cerebralhemispheres The goal of the present study was to ad-dress this question

EXPERIMENT 1

The task was for healthy subjects to judge whether ob-jects in two different displays were the same or different(see Figure 1) In the target display one or two objectswere presented briefly and then masked For two-objectdisplays the items could both be in the same visual field(two-object unilateral display) and therefore project tothe same hemisphere or the two objects could be in dif-ferent visual fields (two-object bilateral display) andtherefore project to different hemispheres In the finaldisplay a probe item was always presented at the loca-tion of one of the target objects Half the time the probematched the previous object and half the time it differedFrom previous research (eg Baylis amp Driver 1993Duncan 1984 1993 Enns amp Kingstone 1997 Vecera ampFarah 1994) we expected that response accuracy wouldbe reduced when the subjects were required to attend totwo objects in the initial display as compared with whenthey had to attend only to a single objectmdashthat is thereshould be a two-object cost The critical question waswhether this object-based attention effect would be thesame or different between the hemispheres To confirmthat our object-based attention effect was not an artifactof space-based attentional orienting we applied the sametest as that in Vecera (1994) and manipulated the dis-tance between objects

MethodSubjects Thirty-two undergraduate psychology students were

tested All had normal or corrected-to-normal vision and receivedcourse credit for their participation

Apparatus This experiment was conducted on a Macintosh 66computer The stimuli were presented on a 14-in Apple color mon-itor (set to black and white) at a viewing distance of approximately57 cm Responses were collected from keyboard buttonpresses

Stimuli and Procedure Figure 1 illustrates the sequence ofstimulus events presented in a given trial The initial display signaledthe start of a trial and consisted of a black central fixation point withfour black location markers on a gray background For half of thesubjects these markers were located 4ordm from central fixation (nearcondition) and for the remaining half of the subjects the markerswere located 8ordm from fixation ( far condition) In both the near andthe far conditions the location markers were positioned on the fourcorners of an imaginary square centered on fixation The subjectswere instructed to keep their eyes on the fixation point at the start ofeach trial and to withhold any eye movements until the end of thetrial The duration of this initial display was 700 msec The next dis-play (the target display) was composed of one or two horizontal orvertical black ovals The ovals subtended 09ordm 3 07 ordm of visual angleand were presented in the location markers for 100 150 or 200 msec(each duration was equiprobable and randomly selected) Immedi-ately following this display was a 180-msec display consisting of

86 VALSANGKAR-SMYTH DONOVAN SINNETT DAWSON AND KINGSTONE

four squares with a pattern of thick white and black oblique linesThese pattern masks subtended 26ordm 3 23ordm of visual angle and werecentered on each of the four location markers The final display (theprobe display) was similar to the second display except that onlyone black oval was presented This probe always appeared in thesame location as a black oval in the target display Half the time theprobe matched the orientation of the target black oval that had pre-ceded it and half the time the probe mismatched it

The subjectrsquos task was to decide whether the probe matched ormismatched the target If the probe matched the target and the probewas in the LVF the subject pressed the ldquozrdquo key with the left handIf the probe matched the target and the probe was in the RVF thesubject pressed the ldquordquo key with the right hand When a responsewas executed the probe was extinguished and after an intertrial in-terval of 1350 msec the next trial began If the probe did not matchthe target no response was to be made On these trials the probewas extinguished after 1995 msec and after an intertrial interval of1350 msec the next trial began

A single object two objects in the same visual field (two-objectunilateral display) and two objects in different visual f ields (two-object bilateral display) were equally likely and were selected ran-domly from trial to trial On single-object displays the position ofthe target occurred at random and with equal probability in each ofthe four possible locations For two-object unilateral displays LVFand RVF presentations were equiprobable and randomly selected

For two-object bilateral displays top bottom and diagonal fieldpresentations were equiprobable and randomly selected For two-object displays the probe item appeared randomly and with equalprobability at one of the target locations On single-object displaysthe probe always occurred at the location of the target In all casestarget and probe orientations were equiprobable and randomly se-lected and whether the probe orientation was the same as or dif-ferent from the target orientation was equiprobable and varied ran-domly from trial to trial

Each subject received 20 practice trials followed by nine blocksof 64 trials Approximately 1 h was required for the subject to com-plete the 696 trials (20 practice trials plus 576 test trials) The sub-jects were instructed to respond as accurately as possible Speedwas not emphasized

ResultsResponse accuracy (proportion correct) was subjected

to an analysis of variance with object display (one-objecttwo-object unilateral or two-object bilateral) displaytime (100 150 or 200 msec) and target visual field (leftor right) as within-subjects factors and display distance(near or far) as a between-subjects factor Performance inall the conditions is presented in Table 1

Figure 1 Example of the sequence of events in Experiment 1 on a two-object bilateral display trial Each trialbegan with the presentation of a central fixation point with four location markers After 700 msec either one ortwo horizontal or vertical black ovals would appear within the location markers for 100 150 or 200 msec (targetdisplay) A 180-msec masking display was then presented The final (probe) display was similar to the target dis-play except that only one black oval was presented in the same location as a black oval in the target display Thetask was to indicate whether or not the probe and the target orientations matched

RIGHT-HEMISPHERE OBJECT-BASED ATTENTION 87

Analysis revealed main effects for object display[F(260) 5 6410 p 0005] display time [F(260) 54524 p 0005] and target field [F(130) 5 2563 p 0005] reflecting the fact that response accuracy im-proved when there was only one object when the displaytime was lengthened and when the target was in the LVFThere was no main effect of display distance [F(130) 5009 p 1] There was however a f ield 3 distanceinteraction [F(130) 5 541 p 05] indicating that theoverall LVF advantage increased when elements wereplaced further afield Display distance had no other effecton performance In particular there was no interactionbetween object display and display distance (all Fs 1)suggesting that the object effects reflected object-basedattention and were not merely an artifact of space-basedattention (Vecera 1994)

The interaction between interval and object was alsosignificant [F(4120) 5 882 p 0005] As is indicatedin Table 1 this was due to the fact that the performanceimprovement that was produced when display time waslengthened was much greater for two-object displaysthan for one-object displays presumably because per-formance was near ceiling for the one-object displayeven at the shortest display duration

The only other significant effect was an object 3 fieldinteraction [F(260) 5 2363 p 0005] As is illustratedin Figure 2 this interaction reflects the fact that there wasno difference between visual fields for one-object displays[F(130) 5 030 p 1] but that there was an advantagefor the LVF in two-object displays Planned contrastsshowed that the LVF advantage was highly significant for

the two-object unilateral display [F(130) 5 9460 p 0005] and was marginally significant for the two-objectbilateral display [F(130) 5 557 p 05] Planned con-trasts also revealed that performance for the two-object bi-lateral displays was higher than that for two-object unilat-eral displays This effect was marginally significant forthe LVF [F(130) 5 601 p 05] and was highly signif-icant for the RVF [F(130) 5 9637 p 0005]

DiscussionThe goal of the present study was to investigate

whether object-based attention effects differ between thecerebral hemispheres To test whether our effects werespecific to object-based attention we manipulated stim-ulus distance as was recommended by Vecera (1994)Our results revealed that response accuracy was higherfor one-object than for two-object displays and that thistwo-object cost did not interact with manipulations tostimulus distance suggesting that it is truly an object-based attention effect In addition the two-object costwas less pronounced when two items were presented be-tween visual f ields rather than within the same field(two-object unilateral display) indicating that there is aperformance benefit when both hemispheres commit at-tention to objects

The most intriguing f inding however was that thetwo-object cost was extremely severe when items wereisolated to the RVF (left hemisphere) This finding sug-gests that the left hemisphere does not have a preferen-tial bias for object-based attention an interpretation thatis at odds with the conclusions of Egly Driver and Rafal

Table 1Mean Response Accuracies for Experiments 1 2 and 3 as a Function of Target Visual Field

(LeftRight) and Three Target Durations (eg 100 150 and 200 msec in Experiment 1)

Experiment 1Left Right

Condition 100 msec 150 msec 200 msec 100 msec 150 msec 200 msec

NearOne-object unilateral 94 94 96 95 95 95Two-object unilateral 79 83 85 70 77 82Two-object bilateral 82 87 89 82 85 87

FarOne-object unilateral 93 95 96 91 94 94Two-object unilateral 79 85 91 66 72 78Two-object bilateral 81 87 89 77 84 87

Experiment 2Left Right

60 msec 105 msec 150 msec 60 msec 105 msec 150 msec

One-object unilateral 88 92 92 85 90 94Two-object unilateral 77 84 87 73 79 83Two-object bilateral 75 86 90 79 82 90

Experiment 3Left Right

60 msec 120 msec 180 msec 60 msec 120 msec 180 msec

One object 67 78 84 71 79 81Two objects 59 68 70 59 62 67

88 VALSANGKAR-SMYTH DONOVAN SINNETT DAWSON AND KINGSTONE

(1994) and Egly Rafal et al (1994) Indeed in contrastto this previous work the present data suggest that theleft hemisphere is particularly poor at committing atten-tion selectively to multiple elements in its visual field

There are several reasons though to question this in-terpretation of Experiment 1 First distance was manip-ulated between groups which may have weakened ourability to capture variations in object costs at differentstimulus eccentricities Consistent with this possibility

is the fact that in the two-object unilateral condition theadvantage of the LVF over the RVF was larger in the farcondition (13) than in the near condition (06) althoughthis variation did not produce a higher order interactionSecond the number of objects presented on any giventrial was confounded with the number of stimulated lo-cations In other words in the two-object condition spa-tial attention had to be divided across two items whereasin the one-object condition it did not Thus our hemi-

Figure 2 Mean response accuracy in Experiments 1 2 and 3 as a function of tar-get visual field and number of objects to be selected

RIGHT-HEMISPHERE OBJECT-BASED ATTENTION 89

spheric differences in the unilateral condition mightmerely reflect the right hemispherersquos ability to dividespatial attention between presented items as comparedwith the left hemisphere Such an interpretation wouldbe consistent with the position that the right hemisphereis superior to the left hemisphere in orienting spatial at-tention (eg Davis amp Schmit 1973)

In Experiments 2 and 3 we examined whether object-based attention effects would continue to be lateralized tothe right hemisphere when these concerns were addressed

EXPERIMENT 2

In order to eliminate any potential confound betweenthe number of items in the visual field and the number ofpossible target items we presented objects at all fourpossible target locations on every trial As in Experi-ment 1 target items were colored black but now they co-occurred with nontarget items that were colored whiteThe subjects were instructed to attend only to the blackobjects The task was again to report whether the probeand the target objects matched Note that because therewas an object in every location and target candidates dif-fered from nontarget candidates solely on the basis of anonspatial object attribute (color contrast) target selec-tion had to be object based rather than space based

Second and as a particularly rigorous test of our ini-tial findings we replaced our Experiment 1 oval-shapedstimuli with letter stimuli One might argue that the oval-shaped stimuli we had used were for some unspecifiedreason better suited for a nonlinguistic right hemispherethan for a language-based left hemisphere and that thisis why we obtained a right-hemisphere advantage in Ex-periment 1 In Experiment 2 we eliminated this concernbecause if anything letter stimuli should be preferredby the language-based left hemisphere (Fecteau Ennsamp Kingstone 2000)

MethodMethodological details were the same as those in Experiment 1

except where indicated Twenty new undergraduate students weretested The display sequence was the same with the exception thatblack or white Zs or Ns (13ordm 3 13ordm) were presented in all four ofthe location markers 6ordm away from central fixation Possible targetletters were colored black and nontargets were colored white Dis-play durations were now 60 100 and 150 msec since pilot workhad revealed that response performance was near ceiling with theslightly longer range of display durations in Experiment 1 In all otherrespects the task was the same as before For example if the probeletter did not match the target letter no response was made If theprobe and the target letters were the same however either a right ora left key was pressed depending on the field of the probe letter

ResultsThe data were analyzed as in Experiment 1 Perfor-

mance in all the conditions is presented in Table 1 Maineffects for object display [F(238) 5 3060 p 001]display time [F(238) 5 4167 p 001] and target field[F(119 5 489 p 05] were all significant demon-strating that as in the first experiment response accu-

racy improved when there was only one object when thedisplay time was lengthened and when the target was inthe LVF Therefore despite the fact that the number ofitems presented on each trial was held constant in Ex-periment 2 and letter stimuli had replaced the oval stim-uli the right hemisphere continued to outperform the lefthemisphere

Both the field 3 interval [F(238) 5 070 p 05]and the interval 3 object [F(476) 5 208 p 05] in-teractions were nonsignificant Importantly and as isshown in Figure 2 the field 3 object performance pat-tern was almost identical to that in Experiment 1 Thisinteraction nudged significance [F(238) 5 280 p 07] and planned comparisons confirmed that as in Ex-periment 1 there was a highly significant LVF advan-tage for the two-object unilateral displays [F(118) 51116 p 005] Planned contrasts also showed that asbefore there was no significant two-object bilateral versustwo-object unilateral advantage for the LVF [F(118) 50610 p 5] but that this effect was highly significantfor the RVF [F(118) 5 15740 p 001]

DiscussionThe primary motivation for Experiment 2 was to de-

termine whether the object effects and hemispheric dif-ferences observed in Experiment 1 would reemergewhen a possible confound between object and locationnumber was controlled and letter stimuli were presentedThe same effects were found with the key finding beingthat the two-object cost was particularly severe when twotarget items were isolated to the RVF (left hemisphere)Because (1) items were presented in every location onevery trial (2) target selection was based on a nonspatialtarget attribute and (3) stimulus form (ie letters)should favor left-hemisphere processing it is reasonablethat the results in Experiments 1 and 2 be attributed to anobject-based selection advantage in the right hemisphere

It is still possible though to contrive a spatial atten-tion explanation for the right-hemisphere advantage ontwo-object unilateral displays Such an account mightposit that although the selection of a target is based onnonspatial attributes shifting attention from one targetto the other requires spatial reorienting and that it is thisact of reorienting that is superior in the right hemisphereTo address directly spatial explanations of this type weran a final experiment that was closely modeled on theoriginal Duncan (1984) paradigm in which overlappingtarget items were used It has been demonstrated repeat-edly that in this paradigm spatial attention accounts arenot viable Thus if a right-hemisphere advantage wasfound to persist we could attribute this lateralization toobject-based attention

EXPERIMENT 3

Two overlapping target items were presented at ran-dom to either the LVF or the RVF The subjects were re-quired to attend to only one of the two target items (one-

90 VALSANGKAR-SMYTH DONOVAN SINNETT DAWSON AND KINGSTONE

object blocks) or to both items (two-object blocks) As inthe previous two experiments at the end of each trial thesubjects were probed for their correct knowledge on oneof two target attributes

MethodMethodological details were the same as those in the previous

experiments except where indicated Twenty new undergraduatestudents were tested The display sequence is illustrated in Figure 3The f ixation preview display was presented for 500 msec Over-lapping target itemsmdasha box and a linemdashsubtending 55ordm 3 30ordmwere then presented randomly to the left or right of center by 9ordm(fixation to middle of overlap) Each target item contained two at-tributes that varied randomly from trial to trial For the box the gapmight be on the top or the bottom and its concave sides were eithercurved or pointed For the line it sloped down from top left to bot-tom right or vice versa and was dotted or dashed Display durationswere now 60 120 and 180 msec A target mask composed of allpossible display items jittered by 025ordm was then presented for120 msec at the location of the target items A probe display wasthen presented in the same field as the target items and mask Thisprobe display consisted of two items positioned side by side thatdiffered only on the attribute that was to be judged When the gapon the box was to be judged rectangles with the gap on the top and

bottom were presented when the sides of the box were to be judgedthe concave sides were curved on one rectangle and pointed on theother When the angle of the line was to be judged two solid lineswere presented side by side with one line sloping down from topleft to bottom right and the other sloping in the reverse directionWhen the texture of the line was to be judged the adjacent lineswere vertical with one dotted and the other dashed When the probedisplay was on the left the subjects made an unspeeded left-handedtwo-alternative choice using ldquozrdquo and ldquoxrdquo to indicate whether the tar-get had possessed the relevant attribute of the left probe or the rightprobe (in Figure 3 the correct response would be a ldquozrdquo response toindicate the left rectangle with the gap on the top) When the probedisplay was on the right the right hand was used to press ldquordquo or ldquordquoto indicate the left or the right probe respectively

In different blocks of trials the subjects were required to attendto the relevant attributes of the box or the line (one object) or boththe box and the line (two objects) The order of these conditionswas counterbalanced across subjects Each condition was com-posed of 184 trials divided equally between two blocks Forty-threepractice trials preceded testing in each condition

ResultsPerformance for one- and two-object displays are pre-

sented in Table 1 Main effects for object display [F(119) 5

Figure 3 Example of the sequence of events in Experiment 3 on a typical trial Each trial began with the presentation of acentral fixation point After 500 msec two overlapping objects would appear on the left or the right for 60 120 or 180 msecEach object a box and a slanted line had two attributes that were relevant if the object was probed gap location (topbottom)and side concaveness (pointedcurved) for the box slant and texture (dotteddashed) for the line The objects were masked for120 msec and then depending on the task one of the two relevant attributes for an object was probed in the same field as thetarget display The task was to indicate with an unspeeded response which of the probes matched the target attribute (eg theleft probe has a top gap as does the target in the present illustration)

RIGHT-HEMISPHERE OBJECT-BASED ATTENTION 91

5145 p 001] display time [F(238) 5 6311 p 001] and target field [F(119) 5 451 p 05] were allsignificant This demonstrates that as in the previous ex-periments response accuracy improved when selectionwas for one object when the display time was length-ened and when items were presented to the LVF (righthemisphere)

As is shown in Figure 2 this LVFright-hemisphereadvantage was revealed when object selection across thetwo objects was required thus producing a highly sig-nificant field 3 object interaction [F(119) 5 6142 p 025] Because the two objects occupied the same loca-tion this object-based attention effect cannot be attrib-uted to a right-hemisphere advantage for spatial orient-ing Thus the present data provide conclusive evidenceand converge with our previous findings that object-based attention is preferentially lateralized to the righthemisphere

No other higher order effect was significant except fora field 3 time interaction [F(238) 5 465 p 05] re-flecting that the right-hemisphere performance advantageincreased in magnitude as display duration increased

DiscussionThe present experiment in conjunction with the pre-

vious two experiments provides compelling evidencethat the left hemisphere is particularly poor at commit-ting attention selectively to more than one object in thevisual field This is true even when target items are pre-sented alone in the visual field at different spatial loca-tions (Experiment 1) when they are letter stimuli andtarget selection cannot be based on spatial attributes (Ex-periment 2) and when items are overlapping in the samephysical space (Experiment 3) The results of each ex-periment converged on the conclusion that the effects re-flected object-based selection rather than space-basedselection Thus in agreement with Egly and colleagues(Egly Driver amp Rafal 1994 Egly Rafal et al 1994)our data show that object-based attention is a specializedform of orienting that is subserved by lateralized corti-cal brain mechanisms Contrary to the conclusions ofEgly and colleagues however our data indicate that theright hemispheremdashand not the left hemispheremdashis pref-erentially biased for committing object-based attentionto elements in the visual environment

To our knowledge the present investigation representsthe first examination of purely object-based attention ef-fects across the cerebral hemispheres The goal of futureresearch will be to isolate the specific brain mechanismssubserving object-based attention by testing patient popu-lations (eg focal lesion and split-brain patients) andorby functional neuroimaging of healthy individuals Weexpect that the paradigm utilized in the present investiga-tion should be amenable to these future lines of research

REFERENCES

Baylis G C amp Driver J (1993) Visual attention and objects Evi-dence for hierarchical coding of location Journal of ExperimentalPsychology Human Perception amp Performance 19 451-470

Davis R amp Schmit V (1973) Visual and verbal coding in the inter-hemispheric transfer of information Acta Psychologica 37 229-240

Duncan J (1984) Selective attention and the organization of visual in-formation Journal of Experimental Psychology General 113 501-517

Duncan J (1993) Coordination of what and where in visual attentionPerception 22 1261-1270

Egly R Driver J amp Rafal R D (1994) Shifting visual attentionbetween objects and locations Evidence from normal and parietal le-sion subjects Journal of Experimental Psychology General 123161-177

Egly R Rafal R Driver J amp Starrveveld Y (1994) Covert ori-enting in the split brain reveals hemispheric specialization for object-based attention Psychological Science 5 380-383

Enns J T amp Kingstone A (1997) Hemispheric coordination of spa-tial attention In S Christman (Ed) Cerebral asymmetries in sensoryand perceptual processes (pp 197-231) Amsterdam ElsevierNorth-Holland

Fecteau J H Enns J T amp Kingstone A (2000) Competition-induced visual field differences in search Psychological Science 11386-393

P osner M I (1980) Orienting of attention Quarterly Journal of Ex-perimental Psychology 32 3-25

P osner M I Snyder C R R amp Davidson B J (1980) Attentionand the detection of signals Journal of Experimental PsychologyGeneral 109 160-174

Vecera S P (1994) Grouped locations and object-based attentionComment on Egly Driver and Rafal (1994) Journal of Experimen-tal Psychology General 123 316-320

Vecera S P amp Farah M J (1994) Does visual attention select ob-jects or locations Journal of Experimental Psychology General123 146-160

(Manuscript received May 16 2000revision accepted for publication January 6 2003)