UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Reflexive and volitional orienting in children and adults : pointing to a new future Ristic , Jelena 2003

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2003-0343.pdf [ 2.42MB ]
Metadata
JSON: 831-1.0091108.json
JSON-LD: 831-1.0091108-ld.json
RDF/XML (Pretty): 831-1.0091108-rdf.xml
RDF/JSON: 831-1.0091108-rdf.json
Turtle: 831-1.0091108-turtle.txt
N-Triples: 831-1.0091108-rdf-ntriples.txt
Original Record: 831-1.0091108-source.json
Full Text
831-1.0091108-fulltext.txt
Citation
831-1.0091108.ris

Full Text

R E F L E X I V E A N D V O L I T I O N A L O R I E N T I N G I N C H I L D R E N A N D A D U L T S : P O I N T I N G T O A N E W F U T U R E by J E L E N A RISTIC . B.A., The Un ive rs i t y of Br i t i sh Co l umb i a , 2000 A THESIS S U B M I T T E D I N P A R T I A L F U L F I L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F M A S T E R O F A R T S i n T H E F A C U L T Y O F G R A D U A T E STUDIES (Department of Psychology; Cogn i t i ve Systems Program) W e accept this thesis as con fo rming to the requ i red standard T H E U N I V E R S I T Y O F BRIT ISH C O L U M B I A Ju ly 2003 © Jelena Rist ic, 2003. In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of The University of British Columbia Vancouver, Canada DE-6 (2/88) 11 A B S T R A C T Trad i t iona l ly , vo l i t iona l attention has been s tud ied i n the Posner cu ing pa rad igm by us ing central spat ia l ly predict ive arrows as attentional cues. The important assumpt ion unde r l y i ng this methodo logy has been that ar rows (and other central attentional cues) orient attention spat ia l ly on ly if they predict where the target is l i ke ly to appear. Recent studies, however , indicate that this fundamenta l assumpt ion unde r l y i ng the classic methodo logy is mistaken. Several studies n o w report that a range of central d i rect ional cues, i n c lud ing arrows, can tr igger ref lexive or ient ing effects i n both young ch i ld ren and adults when they do not predict where the target is l i ke ly to appear. This fact raises the quest ion of whether past studies, u s i ng pred ic t ive central ar rows as attentional cues, were measur ing (1) vo l i t iona l attent ion because the arrows were predict ive, (2) ref lexive attention because the ar rows were direct ional , or (3) some combinat ion of vo l i t iona l and ref lexive attention. Th is issue was invest igated i n two studies. The first s tudy is presented i n Chapter 1. This invest igat ion tested adults w i t h (1) predict ive a r row cues (2) nonpred ic t ive a r row cues, to get a pure measure of ref lexive or ient ing, and (3) predict ive nond i rect iona l cues, to get a pure measure of vo l i t iona l or ient ing. The results of this first s tudy showed that the magn i tude of or ient ing observed w i t h pred ic t ive ar row cues was a lways larger than the s u m of pure ref lexive and vo lunta ry or ient ing, suggest ing that the tradi t ional measure reflects an interact ion between ref lexive and vo l i t iona l attention. The second study, presented in Chapter 2, tested ch i ld ren between the ages of 3 and 6 i n a condi t ions comparable to those i n the first s tudy w i t h adults. The results showed that pure ref lexive or ient ing was adult- l ike, but vo l i t iona l or ient ing was not. Y o u n g ch i ldren, un l i ke adults, cou ld sustain vo l i t iona l attention for on l y a brief per iod Ill of t ime. Moreover , and also un l i ke adults, ref lexive and vo l i t iona l or ient ing appeared to be add i t ive rather than interactive. It is suggested that col lect ively these f ind ings are consistent w i t h the fields' current unders tand ing of the maturat ion of b ra in regions thought to mediate ref lexive and vo l i t iona l or ient ing. i v T A B L E O F C O N T E N T S Abstract i i Table of Contents i v L is t of Tables v i L i s t of F igures v i i Acknow ledgements v i i i Ded ica t ion ix C H A P T E R 1 1 Introduct ion 2 Exper iment 1 7 Me t hod • 7 Part ic ipants 7 Appa ra tus and S t imu l i 7 Des ign 9 A r r o w Cues 10 N u m b e r Cues 10 Procedure 11 Results 12 D iscuss ion 17 Exper iment 2..... 17 Me t hod 18 Part ic ipants 18 Appara tus , S t imu l i , Des ign, and Procedure 18 Results 19 v D iscuss ion 22 Genera l D i scuss ion 22 References 25 C H A P T E R 2 28 Introduct ion 29 Exper iment 1 31 Me t hod 32 Part ic ipants 32 Appa ra tus and S t imu l i 33 Des ign 34 Procedure 35 Results 37 D iscuss ion 42 References 45 LIST O F T A B L E S v i C H A P T E R 1 Table 1. M e a n response times, s tandard deviat ions, and error rates for Exper iment 1 13 Table 2. M e a n response times, standard deviat ions, and error rates for Exper iment 2 19 C H A P T E R 2 ' Table 1. M e a n response t imes and standard deviat ions for adults 37 Table 2. M e a n response t imes and standard deviat ions for ch i ld ren 38 Table 3. Overa l l error and eye movement rates •. 38 v i i LIST O F F I G U R E S C H A P T E R 1 F igure 1. Sample Sequence of Events 9 F igure 2. Exper iment 1 Results 14 F igure 3. Exper iment 2 Results 20 C H A P T E R 2 F igure 1. Sample Sequence of Events 34 F igure 2. Exper iment 1 Results 39 V l l l A C K N O W L E D G E M E N T S I w o u l d l ike to express sincere grat i tude to m y superv isor A l a n K ingstone for the except ional suppor t he p rov i ded me w i th , bo th on a scientif ic and personal level. Next , many thanks to Jatinder (Bobby) S i dhu for a l l the daycare trips; Bobby 's f r iendship and personal sacrif ice t ru ly made the research conducted w i t h y oung ch i ld ren (reported i n Chapter 2) effortless and enjoyable. The research reported here was suppor ted b y graduate fe l lowsh ips f r om the Na tu ra l Sciences and Eng ineer ing Research Counc i l of Canada (NSERC) , the M i chae l Sm i th Foundat ion for Hea l th Research (MSFHR) , and a Un ive r s i t y of Br i t i sh Co l umb i a graduate award . Add i t i o na l suppor t came f r om grants awa rded to A l a n K ingstone f r om the H u m a n Front iers Science P rog ram (HFSP), N S E R C , and M S F H R . to my parents Sonja and Pavle 1 C H A P T E R 1 2 Introduct ion Behav iora l studies of covert spat ial attention to date indicate that two qual i tat ive ly different forms of attentional or ient ing occur i n response to two different attentional cues (e.g., Posner, 1980; Jonides, 1981; Mu l l e r & Rabbitt, 1989). Ref lexive or ient ing occurs i n response to salient events i n the env i ronment. Vo l i t i ona l or ient ing, on the other hand , is generated i n accordance w i t h an observer's goals and expectations. Trad i t iona l ly , behav iora l markers of ref lexive or ient ing were exper imenta l ly examined us ing abrupt unexpected per iphera l events whereas the behav iora l index of vo l i t iona l attention was examined b y man ipu la t ing central spat ia l ly in format ive symbol i c cues. In the per iphera l cu ing task part ic ipants are typ ica l ly presented w i t h brief luminance transients, serv ing as attentional cues, appear ing r andom l y (p=. 5) i n one of the two possib le target locations. The observers' task is to press a key w h e n the target appears on the screen at one of these two locations. In order to prof i le tempora l characteristic of the target response, the t ime between the cue onset and the target onset is typ ica l ly va r ied (St imulus Onset A synch rony or SO A ) between 0 and 1000 ms. The results of such a task general ly indicate that targets appear ing at the locat ion where the cue was presented w i t h i n 0-300 ms are detected faster than targets appear ing at the noncued locat ion. Howeve r , when the cue-target t ime de lay exceeds 300ms, the opposi te effect is seen: targets appear ing at the uncued locat ion n o w produce faster response t imes (RTs) than the targets appear ing at the p rev ious l y cued locat ion. This reversal of the attentional effect has been labeled " Inh ib i t ion of Re tu rn " (IOR), denot ing the idea that attention was in i t ia l ly al located to the cued locat ion (hence the shorter RTs ear ly on) and is n o w be ing inh ib i ted i n return ing to the cued locat ion at longer S O A intervals (e.g., Posner, 1980; Posner & Cohen , 1984). This b iphas ic RT pattern at the cued locat ion marked b y ear ly RT faci l i tat ion fo l l owed by IOR at longer cue-target delays has 3 tradi t ional ly been thought to reflect the behav iora l s ignature of automat ic attentional or ient ing (e.g., Rafal , Calabres i , Brennan & Sciolto, 1989; Posner & Cohen , 1984). In contrast, vo l i t iona l or ient ing is thought to reflect contro l led rather than automatic process ing i n that it requires the deve lopment of an expectancy set that is based on the in format ion p rov i ded b y the cue (Mu l le r & Rabbitt, 1989; Jonides, 1981). A s such, attentional cues typ ica l ly used to elicit vo l i t iona l or ient ing manipu late internal expectations about a target's occurrence b y us ing mean ingfu l symbo l i c s t imu l i for d i rect ing attention (e.g., Jonides, 1981). In this task, part ic ipants are asked to fixate a central a r row cue that has predict ive va lue about a target's occurrence. That is, the target appears at the locat ion ind icated by a r row d i rect ion on 80% of a l l the trials. The task is to detect the target appear ing either at the cued or l i ke ly locat ion (p=. 8) or at the uncued or un l i ke l y locat ion (p=. 2). The results show that targets appear ing at the l i ke ly locat ion are a lways detected faster than targets appear ing at an unexpected locat ion w i t h the effect increasing and reaching an asymptote at approx imate ly 500ms S O A (Mu l l e r & Rabbitt, 1989; Jonides, 1981). The differences i n the behav iora l index of ref lexive and vo l i t iona l attention movements have suggested that this d i cho tomy may not reflect differences i n al locat ion of resources w i t h i n a s ingle attentional mechan i sm (Posner, 1980; Jonides, 1981). A s such, it was suggested that ref lexive and vo l i t iona l attention operate i n an add i t ive fashion w i t h automat ic (reflexive) and contro l led (vol i t ional) attentional be ing di f ferent ia l ly susceptible to manipu lat ions of cognit ive l oad and expectancy in fo rmat ion (e.g., Mu l l e r & F ind lay , 1988). Ref lex ive or ient ing was f ound to be h igh l y resistant to in format ion p rov i ded b y the cue; it cou ld not be prevented g iven the proper t r igger ing st imulus. A t the same t ime, vo lunta ry attention was found to be susceptible to the effects of a secondary task as we l l as to the automatic effects p roduced by the sensory propert ies of the cue (e.g., Mu l l e r & Rabbitt, 1989). This difference between 4 ref lexive and vo l i t iona l attention has been re inforced b y a wea l th of neuro imag ing studies report ing that ref lexive and vo l i t iona l or ient ing, as def ined b y the classic behav iora l parad igms, engage dist inct neura l circuits (e.g., Posner & Petersen, 1990; Posner, 1992; Corbetta, M i e z i n , Shu lman & Petersen, 1993; Hopf inger , Buonocore & M a n g u n , 2000; Corbetta & Shu lman, 2002). Conve rg ing evidence indicates that poster ior par ieta l lobe and frontal cortex (more speci f ica l ly f ronta l eye f ie ld region) are i nvo l ved i n contro l l ing the d i rect ion of covert vo l i t iona l attention as we l l as the maintenance of an expectancy set. In contrast, the circuit that invo lves ventra l frontal cortex and temporopar ieta l junct ion shows sustained act ivat ion, as measured b y fMRI , for detect ion of abrupt s t imul i . Because of the d i f fer ing maturat iona l rates of frontal and par ietal cortices (e.g., Johnson, 1997), marked deve lopmenta l differences of ref lexive and vo l i t iona l attention have been observed such that fu l l y - funct ion ing ref lexive or ient ing is present short ly after b i r th (Johnson, Posner & Rothbart, 1991) wh i l e vo l i t iona l or ient ing reaches adult- l ike funct ion ing much later, poss ib ly at a round 8 years of age (e.g., Brodeur & Enns, 1997). The key theoretical assumpt ion unde r l y i ng classic behav iora l parad igms of spatial attention is that central a r row cues engaged vo l i t iona l or ient ing mechan ism on ly because they predict spat ia l ly where the target w i l l appear (Mu l l e r & Rabbitt, 1989; Rafa l et a l , 1989; Jonides, 1981). In other words , w i thout pred ic t ive in format ion about the target, a r rows were regarded as ineffective i n e l i c i t ing any or ient ing response, and cou ld not, b y v i r tue of the s t imulus characteristics, tr igger a shift i n attention. This v i ew stems largely f r om a classic s tudy by Jonides (1981, Exper iment 2) that requi red subjects to search a br ief ly presented array of letters for the target letter (L or R). Before the array appeared, a central a r row cue, r andom ly po in t ing to one of the target locations, was f lashed momentar i l y at f ixat ion. Results ind icated that if subjects were to ld to ignore the arrow, or ient ing to the cued locat ion was absent compared to the cond i t ion where 5 ar rowhead was flashed at a per iphera l locat ion, suggest ing that a nonpred ic t ive central a r row cue does not tr igger ref lexive attention. Several recent studies, however , indicate that this acceptance of a nu l l result was mistaken. Rist ic, Fr iesen, and K ingstone (2002) asked part ic ipants to detect targets appear ing to the left or r ight of a central ar row cue. Important ly, the a r row was un informat ive of the target locat ion such that it was r andom ly po in t ing to either of the two possible target locations. Cont ra ry to the tradi t ional result, part ic ipants were a lways faster to detect targets appear ing at the cued compared to uncued locations even though they unders tood that a r row direct ion had no predict ive mean ing. Rist ic et al (2002) observed faster RTs for cued targets at a range of cue-target intervals (195,600, and 1000 ms), demonstrat ing that un in format ive central a r row cues tr iggered a ref lexive shift i n spat ia l attention. Interest ingly, a nd contrary to the tradi t ional f ramework, the attentional or ient ing index was ma rked b y a pro longed faci l i tat ion for cued targets w i thout be ing accompanied b y inh ib i t ion at longer S O A intervals. It is important to note that this is not an isolated instance repor t ing ref lexive or ient ing i n response to central ly presented direct ional un in format ive cues. S imi lar f ind ings us ing central a r row s t imu l i were obta ined i n two other studies (Homme l , Pratt, Co l za to & God j in , 2001; T ipples, 2002) as we l l as i n other invest igat ions emp loy ing other central d i rect ional cues such as eye gaze d i rect ion of both schematic faces (e.g., Fr iesen and K ingstone, 1998) and images of real faces (e.g., D r i ve r et al , 1999; Langton & Bruce, 1999), head or ientat ion (Dr iver et al , 1999; Langton, 2000), and f inger po in t ing (Langton & Bruce, 2000). Taken together, these recent results demonstrate that a range of nonpred ic t ive but direct ional attention cues presented at central f ixat ion, i nc lud ing arrows, re l iab ly tr igger a ref lexive shift of spatial attention, w i t h the effect marked ly different effect f r om the tradi t ional ref lexive effect, i.e., ref lexive or ient ing i n is long-l i ved and it is not accompanied by IOR. 6 Since the in t roduct ion of the cu ing pa rad igm i n ear ly 1980s, the poss ib i l i ty of a central attentional cue tr igger ing ref lexive attention was abandoned, and, as a consequence never ser iously entertained, a premise that was largely based on f ind ings reported b y Jonides in 1981 (see K ingstone, Smi lek, Rist ic, Fr iesen & Eas twood, i n press for a rev iew). The recent studies, however , show ing that spat ia l ly nonpred ic t ive but d irect ional s t imu l i p roduce shifts i n ref lexive attention, raise the poss ib i l i ty that ref lexive attention may have been contr ibut ing to the effects observed i n the past w i t h spat ia l ly pred ic t ive direct ional s t imul i . In l ight of this poss ib i l i ty , the most important quest ion n o w becomes whether the classic behav iora l pa rad igms u t i l i z i ng spat ia l ly in format ive a r row s t imu l i were est imat ing the contr ibut ion of vo l i t iona l attention alone, as or ig ina l ly thought (because the a r row is predict ive), ref lexive attention (because a r row is direct ional), or perhaps some un ique combinat ion of ref lex ive and vo l i t iona l responses that result f r om an interact ion generated b y cue d irect ional i ty and its in format ion. This test represents a cruc ia l examinat ion of the va l i d i t y of bo th the classic exper imenta l pa rad i gm and the accumulated results that were, over the past years, generated by it. In two experiments reported here we invest igated this fundamenta l quest ion whereby or ient ing responses for spat ia l ly predict ive and nonpred ic t i ve d irect ional (arrow) and nondi rect iona l (digit) attentional cues were dissociated. If the classic l ine of thought is correct, than we shou ld observe l itt le or no difference between the vo l i t iona l or ient ing effects generated i n response to spat ia l ly pred ic t ive d i rect ional cues compared to spat ia l ly pred ict ive nondi rect iona l central cues. If, i n contrast, the tradi t ional pa rad igm was not measur ing vo l i t iona l or ient ing alone, one w o u l d predict marked differences i n the attentional effect p roduced by a r row and d ig i t cues. 7 Exper iment 1 Exper iment 1 was des igned to measure attentional or ient ing i n response to spat ia l ly pred ict ive and nonpred ict ive central cues. Spat ia l predict iveness was var ied across two different cues types: d irect ional a r row cues and nond i rect iona l d ig i t cues. A s noted, recent studies demonstrate that central a r row cues tr igger ref lexive shift of attention even w h e n ar row direct ion does not re l iably pred ict the target locat ion (e.g., Rist ic et a l , 2002). The a im here was to emp loy a central cue that w o u l d not have inherent d irect ional i ty and thus w o u l d not engage ref lexive attention. Howeve r , w h e n this cue was g iven predict ive value, it w o u l d p rov ide a pure measure of vo l i t iona l or ient ing. The cr i t ica l compar i son here is whether vo l i t iona l or ient ing tr iggered b y such nondi rect iona l cues matches the vo l i t iona l or ient ing effect p roduced b y the classic spat ia l ly pred ic t ive a r row cues. Me t hod Part ic ipants For ty eight (48) undergraduate students part ic ipated i n the exper iment i n exchange for monetary compensat ion (24 part ic ipated i n each of the two cue type condit ions). A l l observers were naive to the purpose of the exper iment and reported no rma l or corrected-to-normal v i s ion 1 . Appa ra tus and S t imu l i S t imu lus presentat ion and t im ing was contro l led b y VScope 1.2.7 software (Rensink, 1995) runn ing on the 6100/66 Power Mac in tosh computer . S t imu l i were presented on a 15-inch A p p l e color moni tor set to black and whi te , operat ing at 65Hz screen refresh rate. 1 Seven participants were excluded from the initial sample comprised of 55 participants. Because the task was a simple detection task, we adopted a stringent criterion based on which all observers who made more than 5% errors across both cue conditions were excluded from the analysis. 8 The s t imu l i and sample t im ing sequence are i l lustrated i n F igure 1. A l l s t imul i were black l ine d raw ings presented on whi te background. A r r o w cues were created by comb in ing a straight l ine (2.1° long) w i t h an a r rowhead and an ar rowta i l (wi th each 45-degree or iented l ine measur ing 1° i n length) attached to both ends of the l ine (e.g., <-<). The who le a r row measured 3. 3° of v i sua l angle i n length as measured f r om the tip of the a r rowhead to the end of the arrowta i l . D ig i t cues (3, 6, and 9) were 3. 3° i n height and 2° i n w i t h the except ion of number 1 wh i c h was created us ing a capital letter I (3.2° l ong and 0.5° w ide) . The number cues were created us ing the Geneva font of 100 points i n size. Both a r row and number cues were pos i t ioned such that the center of the ar row and the center of the d ig i t was a lways a l igned w i t h the center of the screen. A t the start of each tr ial, a central f ixat ion po int subtend ing 1°, compr i sed of two perpendicu lar intersect ing l ines (each 1° i n length), appeared at the center of the screen. The target was a lways a black asterisk (measur ing .9°) that appeared w i t h eccentricity of 6.5° as measured f r om the center of the cue to the center of the target. The center of the target was a l igned w i t h the center of both ar row and number cues a long the hor izonta l and vert ical axes. Directional Cue Nondirectional Cue E 675 ms Cue Onset 100, 300, 600, 900 SOA Target Onset until response or 2700ms Figure 1 illustrates the stimuli and timing sequences for directional (arrow) and nondirectional (digit) cue conditions. Identical stimuli and presentation sequence were used in both Experiment 1 and Experiment 2. A straight line or a fixation point appeared on the screen for 675 ms. Then, an arrow pointing left, right, up or down, or a central number cue (1, 3, 6, 9) appeared on the screen. The target appeared either to the left, right, up or down after 100, 300, 600 or 900 ms. Both the cue and the target remained on the screen until response was made or for 2700ms, whichever came first. Intertrial interval was 525 ms. Note that the stimuli are not drawn to scale. Des ign Cue type (arrow or number) was var ied between subjects so that each group responded to a single central cue (either a r row or number) . Both central cues were var ied as either spat ia l ly predict ive (probabi l i ty of the target occur r ing at the cued locat ion =. 8) or spat ia l ly nonpred ic t ive (probabi l i ty of the target occurr ing at the cued locat ion p=. 25) of the target locat ion. Cue predict iveness was var ied w i t h i n subjects such that each part ic ipant responded to both predict iveness condit ions. Cue predict iveness order was counterbalanced across part ic ipants such that half the 10 part ic ipants i n each group (N=12) received spat ia l ly pred ic t ive cues first and the other half (N=12) received spat ia l ly nonpred ic t ive cues first. A r r o w Cues. O n every tr ia l , a central cue cou ld be directed to the left, r ight, up , or d o w n const i tut ing four possib le target locations (left, r ight, up , down) . In the nonpred ic t ive condi t ion, the target appeared w i t h equal probab i l i ty at a l l four locations (p=. 25). A l l possible cue direct ions and target locations were d is t r ibuted equal ly throughout the exper iment. In the predict ive cond i t ion however , the target appeared at the locat ion to where the a r row was po in t ing i n 80% of a l l trials. In rema in ing 20% of trials targets appeared equa l ly often among three rema in ing locat ions (6.67% for each of the three locations). Tr ia ls i n wh i c h the target appeared at the locat ion to where the ar row was po in t i ng are labeled as cued target trials and the tr ials i n wh i c h the target appeared at any of three other possible locations are labeled as uncued target trials. N u m b e r Cues. For the number cues one of the four possib le digits (1 ,3 ,6 ,9) appeared at the center of the screen at the beg inn ing of each tr ial . W h e n number cues were un in format ive of the target locat ion, the targets appeared w i t h equal probabi l i ty (p=. 25) at a l l four possib le locations (left, r ight, up , down) regardless of wh i c h central d ig i t appeared on the screen. Howeve r , w h e n the number cue was pred ic t ive of the target locat ion, the exper iment was set up so that number 3 pred ic ted the target occurr ing on the r ight, number 9 the target occurr ing on the left, number 1 the target occurr ing up, and number 6 the target occurr ing down . A l l four cues were presented equal ly often throughout the exper iment. M i r r o r i n g the setup emp loyed for a r row cues, spat ia l ly pred ic t ive d ig i t cues correct ly ind icated target locat ion on 80% of a l l tr ials. In the rema in ing 20% of the trials the target appeared r andom ly at one of the three rema in ing locations (6.67% per location). For a l l cue types (arrows and numbers) and cue predict iveness (nonpredict ive or predict ive) condit ions, four cue-target onset delays (100,300, 600 and 900 ms) were • 11 var ied equa l l y among a l l possib le trials. In a l l condi t ions part ic ipants were asked to per fo rm a speeded target detection response by press ing the spacebar key on the keyboard w i t h the index f inger of their preferred hand . Add i t i ona l l y , i n approx imate ly 6% of a l l tr ials w i t h i n each of the four condit ions the target was not presented on the screen. These catch trials were d ispersed randomly across a l l possib le cue direct ions and were inc luded to ensure that part ic ipants responded to target onset and not its ant ic ipated appearance. Procedure The start of every tr ial was s ignaled b y a 675ms presentat ion of a f ixat ion cross i n the center of the computer screen. Then, a central cue (either a r row po in t ing left, r ight, up , or d own , or one of the four number cues) appeared on the screen. The target demand ing a s imple detection response appeared at one of the four target locations after 100,300,600, or 900 ms. The tr ial terminated on response or after 2700ms, wh ichever came first. The intertr ia l interval was set at 525 ms. React ion T ime (RT) was measured f r om target onset and it was based on execut ion of the keyboard responses. Each cue type x cue predict iveness cond i t ion (arrow nonpred ic t ive; a r row predict ive; number nonpred ic t ive; and number predict ive) was compr i sed of 480 exper imental trials d is t r ibuted over 8 b locks of 60 trials. Thus, each part ic ipant completed a total of 960 trials, 480 i n response to each cue predict iveness cond i t ion. Part ic ipants were seated i n a d im l y lit r oom centered w i t h respect to the computer screen and the keyboard at approx imate distance of 57 cm. Before the commencement of the experiment, observes were s hown a p icture of a typ ica l exper imenta l tr ia l . They were in fo rmed about the type of the central cue, its possible direct ions (or values), and possible target locations. Depend i ng on the cue predict iveness cond i t ion, observers were in fo rmed that the cues were either nonpred ic t ive or pred ict ive of the target locations. They were expl ic i t ly to ld, and 12 understood, the probabi l i t ies of the target occurrence for either a r row or number central cues. Ten practice trials were r un before the first testing block, and the experimenter offered to answer any questions after the practice r un was completed. Part ic ipants were instructed h o w to init iate testing blocks and were i n fo rmed about the durat ion of the exper iment and their total t ime commitment . A l l part ic ipants were asked to respond as fast and as accurately as they cou ld and to ma inta in central f ixat ion throughout the experiment. Results Incorrect key presses, ant ic ipat ions (RT<100 ms), t imed-out responses (RT>1000ms), and false a larms (responding w i t h a keypress w h e n the target was not present) were classif ied as errors and were exc luded f r om the analysis. For the a r row central cues, ant ic ipat ions accounted for 0.47% of a l l trials wh i l e t imed-out responses accounted for 0.26% of a l l target-present trials across both predict iveness condit ions. Ove ra l l false a la rm rate was 0.62%. For d ig i t central cues, ant ic ipat ions accounted for 0.53%, t imed out trials for 0.26%, and key press errors for 0.004% of a l l target present trials. False a la rm rate was 1.02%. A s each type of error accounted for less than 2% of a l l trials, errors were not ana lyzed further. M e a n error rates for each of the S O A by va l id i ty cond i t ion are presented in Table 1. 13 Arrow Number Condition nonpredictive predictive nonpredictive predictive M SD %E M SD %E M SD %E M SD %E 100 ms SOA Cued 342.7 52.6 0 349 51.0 .003 344.8 38.9 .002 355.2 50.6 .001 Uncued 348.7 50.5 .002 369 57.3 .004 345.9 42.8 .002 360 51 .006 300 ms SOA Cued 324.1 51.7 .009 316.8 48.6 .014 326 41.9 .017 323.4 42.2 .013 Uncued 334.5 52.4 .018 358.8 48 .017 324.9 38.7 .02 340.5 46.3 .011 600 ms SOA Cued 305.1 43.9 .004 304.8 42.1 .007 307.7 43.1 .006 300.9 39 .009 Uncued 321.5 47.1 .006 350.6 52.7 .004 310.1 36.4 .006 330.5 49.2 .011 900 ms SOA Cued 321.3 50.1 .007 317.0 44.3 .006 322.9 45 .004 313.4 44.2 .006 Uncued 330.7 47.6 .005 354.9 46.3 .011 318.6 40 .006 340.1 47 .004 Tablel. Mean response times (RTs), standard deviations, and error rates for Experiment 1. The correct RT means are i l lustrated i n Figures l a and l b for the a r row and digit central cues as a funct ion of S O A and cue va l id i ty . A r r o w cues were effective i n tr igger ing or ient ing both as nonpred ic t ive and predict ive of a target pos i t ion. Rep l i cat ing prev ious results (e.g., Jonides, 1981; Rist ic et a l , 2002), on average, part ic ipants responded faster to targets at cued compared to noncued locations. N u m b e r cues, on the other hand, were effective on ly w h e n they pred ic ted target locations. 14 A l t h ough both types of pred ict ive cues (arrow and digit) p roduced signif icant or ient ing effects, the cu ing effect (uncued RT minus cued RT) was m u c h larger for the predict ive a r row cue. F igure 2c i l lustrates the cu ing effects for a l l types of cues. A: Directional Cue 375 358 . Q; E i - 341 c o t; ro 324 CU cC c ro 307 <u Z 290 nonpredictive predictive arrow Q arrow \ \ v . • V 1 1 1 1 V I I I I legend uncued cued 100 300 600 900 SOA 100 300 600 900 SOA B: Nondirectional Cue 375 358 . — ' CU E i - 341 c o '•6 ro 324 cu cC c to 307 cu T. 290 nonpredictive predictive number number 2, 0 • \ N \ \ • v . V —I 1 1 1 V I I I ! legend uncued cued 100 300 600 900 SOA C: Cuing Effects 50 100 300 600 900 SOA • o cu D u I •a v U c 3 legend predictive arrow predictive number nonpredictive arrow nonpredictive number 100 300 600 SOA 900 Figure 2 shows results from Experiment 1. Figure 2a shows mean response times (RTs) plotted as a function of SOA and validity for nonpredictive and predictive directional arrow cues. Figure 2b shows mean RTs for nonpredictive and predictive central digit cues. Figure 2c plots the difference between uncued and cued RTs from all four conditions presented in Figures 2a and 2b. 15 These observations were first conf i rmed by 2 separate w i thin-subjects A N O V A s w i t h cue predict iveness, S O A , and va l id i ty , conducted on each of the cue type separately 2. S ignif icant ma i n effects of both S O A and va l i d i t y were observed for both ar row [F (3, 69)= 42.6, p< .0001; F (1, 23)= 101.47, p< .0001] and number center cues [F (3, 69)= 39.42, p< .0001; F (1, 23)= 16.43, p< .0001 respectively] ind ica t ing a preparatory set for increasing S O A intervals, and overa l l faster RTs for cued compared to uncued targets, respectively. The cue predict iveness x va l id i t y [all Fs> 4.0, ps< .05] interact ion was s ignif icant i n both analyses, as we l l as a three-way interact ion between cue predict iveness, S O A , and va l id i ty [arrow: F (3, 69)= 2.82, p<. 05; digit: F (3, 69)= 4.21, p< .01]. Thus, the va l i d i t y effect dif fered w i t h respect to S O A and two predict iveness condi t ions for each cue type as i l lustrated i n F igure 2c. Next , i n order to examine potent ia l differences between the two cues, cue type (arrow, number) was inc luded as a between -subject var iable i n a separate four -way between-wi th in A N O V A . Signif icant ma i n effects of va l i d i t y [F (1, 46)= 99.29, p<. 0001], S O A [F (3,138)= 80.19, p<. 0001], and cue predict iveness [F (1,46)= 6.83, p<. 05] were observed. The highest order interact ion that reached signif icance was a three-way interact ion between cue predict iveness, SOA , and va l i d i t y [F (3,138)= 6.4, p<. 001] ind icat ing that or ient ing effect var ied dif ferent ly across S O A for spat ia l ly predict ive and spat ia l ly nonpred ic t ive cues. Two two-way interactions between cue predict iveness and 2 Although the order of cue predictiveness presentation was counterbalanced across subjects for both cue types, we included it as a between-subject variable in a preliminary analyses conducted on each of the cue type separately. No order effects interacting with validity were observed. For the arrow cues a four-way between (order)-within (cue predictiveness, SOA, and validity) omnibus A N O V A conducted on mean RT with returned a significant interaction between cue predictiveness and order [F (1, 22)= 12.46, p<. 01] reflecting the issue that both predictive and nonpredictive cues were overall responded to faster when received second with the effect being bigger for the predictive cues. No other effects or interactions involving presentation order were significant [all Fs< 2.6, all ps>.06]. The same analysis was performed on digit central cues. The only significant effect involving order was again an interaction with cue predictiveness [F (1, 22)= 7.69, p< .05] reflecting that again cue predictiveness condition received second was overall responded to faster. Importantly, an interaction involving validity and order was not present [all Fs<l]. 16 va l i d i t y [F (1,46)= 110.05, p<. 0001] and S O A and va l i d i t y [F (3,138)= 11.8, p<. 0001] demonstrat ing the s imi lar trend were also signif icant. Howeve r , the interact ion between cue predict iveness, va l id i ty , and cue type was nonsigni f icant [p>. 2] ind icat ing that, overa l l , the pattern of or ient ing in response to non in format ive and informat ive cues was not statist ical ly different between direct ional and nond i rect iona l attentional cues. W h e n nondi rect iona l number cues were var ied independent ly of the target pos i t ion, no differences i n RTs were observed for cued compared to uncued targets. In contrast, w h e n number cues were made predict ive of the target locat ion, a s ignif icant cu ing effect at longer S O A intervals was observed, demonst ra t ing a result consistent w i t h behav iora l index of vo l i t iona l or ient ing (e.g., Mu l l e r & Rabbitt, 1989). A recent s tudy (Fisher, Castel , D o d d & Pratt, 2003) reported that numer i ca l l y l ow central ly presented number cues (1 and 2) tr igger an automatic shift of attention to targets appear ing i n the left v i sua l f ie ld whereas numer i ca l l y h i gh numbers (8 and 9) induce a shift of spat ia l attent ion towards the r ight v i sua l f ie ld. To examine whether this effect was operat ing i n the present data, we ana lyzed mean RTs i n response to spat ia l ly nonpred ic t ive d ig i t cues (1,3, 6,9) as a funct ion of S O A (100,300,600, and 900 ms) and target pos i t ion (left, r ight, up , or down) . N o s ignif icant interact ion i n vo l v i ng cue type and target pos i t ion was observed [F (9, 207) = 1.79, p > .05] ind ica t ing that indeed this spat ial bias was not operat ing i n our data and as such d i d not inf luence the magni tude of vo l i t iona l or ient ing w h e n d ig i t cues were man ipu la ted as spat ia l ly predict ive. Overa l l , the results f r om Exper iment 1 indicate that w h e n var ied as spat ia l ly pred ict ive of the target pos i t ion, both ar row and number cues p roduced signif icant cu ing effects for targets appear ing at the pred ic ted locations. In contrast to number cues, on l y d irect ional a r row tr iggered a ref lexive shift of attent ion w h e n it was un in format ive of the target pos i t ion, and w h e n i t was made spat ia l ly pred ic t ive the s ize of or ient ing effect surpassed that of symbol i c in format ive d ig i t cues. 17 D iscuss ion W h e n cue direct ional i ty and cue predict iveness are exper imenta l ly dissociated, marked differences i n magn i tude of vo l i t iona l or ient ing i n response to direct ional and nondi rect iona l attentional cues emerge: Pred ict ive a r row cues p roduced larger or ient ing effects than predict ive nondirect iona l d ig i t cues. Howeve r , w h e n or ient ing i n response to the two cues was compared i n a between-subject analys is no s ignif icant interact ion between cue type, cue va l id i ty , and predict iveness was observed. Th is result raises two possible explanat ions. First, it is indeed possible that or ient ing i n response to nondi rect iona l and direct ional predict ive cues does not differ, a f i nd ing that w o u l d val idate t radi t ional cu ing studies. Howeve r , this does not agree w i t h our observations as i l lustrated in F igure 2c. A n alternative explanat ion is that the cue type was inc luded as a between-subject factor and the present exper iment lacked the necessary power to detect a s ignif icant difference between the magni tudes of or ient ing to predict ive d irect ional and predict ive nondirect iona l cues. Exper iment 2 was conducted in order to address this alternative. Exper iment 2 The results f r om Exper iment 1 demonstrated that the magn i tude of or ient ing effect tr iggered b y predict ive ar row cues a lways exceeded or ient ing effects el ic ited b y an uncontaminated measure of vo l i t iona l or ient ing p roduced b y symbo l i c d ig i t cues. Howeve r , this magn i tude difference was not rel iable. A l t h o u g h the data f r om Exper iment 1 are conv inc ing, the fact remains that the results d i d not show statistical difference between the two cues when tested i n an omn ibus A N O V A . In add i t ion, even if a rel iable difference i n or ient ing magn i tude between the pred ic t ive a r row and dig i t cues had been observed, the fact remains that cue type was man ipu la ted between subjects. A s such there is a real poss ib i l i ty that any potent ia l differences i n or ient ing might have reflected group differences rather than differences i n attentional or ient ing. 18 To address both these issues, we examined whether rel iable difference w o u l d be observed w h e n the effects are measured ent ire ly w i t h i n the same group of part ic ipants. Me t hod Part ic ipants The data f r om add i t iona l forty eight (48) undergraduate students f r om the Un ive rs i t y of Br i t i sh Co l umb i a were inc luded i n the analysis. A l l part ic ipants were b l i nd to the purpose of the exper iment and none had part ic ipated i n any prev ious condit ions. Test ing was d i v i d ed over two sessions i n dura t ion of less than one hour each that were conducted on separate days. Appara tus , S t imu l i , Des ign , and Procedure Exper imenta l parameters were kept ident ica l to those of Exper iment 1 unless expl ic i t ly stated. A l l part ic ipants completed al l four cue type x cue predict iveness condit ions. Cue type (arrow or number) was counterbalanced w i t h i n subjects across sessions such that ha l f the part ic ipants received (N=24) a r row cues first and the other half received number cues first (N=24). Order of cue predict iveness (nonpredict ive; predict ive) presentat ion was counterbalanced both between and w i t h i n subjects such that half of a l l observers (N=24) received pred ic t ive cues fo l l owed b y nonpred ict ive cues i n the first session and the other half (N=24) received the opposi te order i n their first session. Counterba lanc ing of cue type and cue predict iveness was complete ly crossed such that each group of 12 part ic ipants received a dist inct combinat ion of a r row and number cues between sessions as we l l as two cue predict iveness orders w i t h i n sessions. C u e type was kept constant w i t h i n a single session i n order to match exper imenta l man ipu la t i on w i t h that used i n Exper iment 1. Each part ic ipant completed a total of 1920 exper imenta l trials, 480 i n response to each cue type x cue predict iveness condi t ion. 19 Results A s i n Exper iment 1, ant ic ipat ions (RT<100) / t imed-out responses (RT>1000 ms), and false a larms were exc luded f r om the analysis. Errors occurred on less than 1% of al l target trials wh i l e false a larms occurred on less than 2.1% no target trials i n each of the cue predict iveness condit ions. Errors were not ana lyzed further. M e a n correct RTs for al l cue predict iveness condit ions are s hown i n Table 2 and i l lustrated i n F igure 3a for a r row cues and F igure 3b for number cues. Arrow Number Condition nonpredictive predictive nonpredictive predictive 100 ms SOA Cued Uncued 300 ms SOA Cued Uncued 600 ms SOA Cued Uncued 900 ms SOA Cued Uncued M SD %E 343.2 352.1 323.2 332 306.8 318.5 313 325.1 64.2 .007 61.8 .009 59.6 .027 57.8 .027 46 .008 48 .01 48.5 .004 44.2 .012 M SD %E 342.3 56.7 366.7 72 311.5 49.7 354.6 55.3 299.3 41.5 337.7 44 308.2 40.1 339.4 48.3 .004 .006 .027 .029 .011 .007 .007 .01 M SD %E 347.4 63.5 .007 343.9 57.1 .01 324.7 55.6 .022 324 49.3 .029 308.6 44.9 .011 313.6 45.9 .009 316.2 46.3 .011 319.8 45.5 .011 M SD %E 347 57 .009 348 58.3 .005 321.5 50.2 .025 334.1 57.1 .033 307.4 46.2 .014 322.1 45.2 .019 311.1 41 .013 325.6 39.3 .01 Table 2. Mean response times (RTs), standard deviations, and error rates for Experiment 2. 20 A: Directional Cue 375 £ 3 5 8 cu j ! 341 c o '"g 324 | 307 290 nonpredictive arrow predictive arrow l egend uncued cued 100 300 600 9 0 0 S O A 100 300 600 900 S O A B: Nondirectional Cue 375 358 . — ' cu E 341 i -c o tj 324 I D CU a: c I D 307 CU 290 nonpredictive number predictive number \ V • • • I yv-° V-• i i i l egend uncued cued 100 300 600 900 S O A C: Cuing Effects 50 •a cu u I T 3 CU 3 < J C 3 100 300 600 900 S O A l egend predictive arrow predictive number nonpredictive arrow nonpredictive number -®-100 300 600 900 Figure 3 shows results rlrjirh Experiment 2. Figure 3a shows mean response times (RTs) plotted as a function of SOA and validity for nonpredictive and predictive directional arrow cues. Figure 3b shows mean RTs for nonpredictive and predictive central digit cues. Figure 3c plots the difference between uncued and cued RTs from all four conditions presented in Figures 3a and 3b. A s i l lustrated i n F igures 3a and 3b the results m i r ro red those obta ined i n Exper iment 1. These observations were examined us ing a four -way within-subjects A N O V A w i t h cue type (arrow, number), cue predict iveness (nonpredict ive, predict ive), 21 S O A (100,300, 600, and 900 ms), and va l id i t y (cued, uncued) i nc luded as factors 3. The results ind icated that, on average, observers were faster to detect targets appear ing after longer cue-target delays [F (3,141)= 59.4, p< .0001] as we l l as that overa l l cued targets were detected faster than uncued targets [F (1,47)= 167.78, p< .0001]. T w o two-way interactions between cue type and va l id i t y [F (1,47) = 89.67, p< .0001] and cue predict iveness and va l id i ty [F (1, 47)= 97.08, p<. 0001] were h igh l y signif icant. A three-w a y interact ion between cue predict iveness, S O A , and va l i d i t y was also s ignif icant [F (3,141)= 4.62, p< .01] ref lect ing that the cu ing effect d i f fered across S O A intervals but on ly w h e n attentional cues were made spat ia l ly pred ict ive of the target. Important ly, so was the three-way interact ion between cue type, cue predict iveness, and va l i d i t y [F (1, 27)= 19.23, p< .0001] ind ica t ing that spat ia l ly pred ic t ive a r row cues p roduced s igni f icant ly larger or ient ing effects than the predict ive number cues. A s i n Exper iment 1, i n order to ver i fy that spat ia l ly nonpred ic t i ve number cues d i d not p roduce any RT faci l i tat ion for any of the possible target posit ions, the data f r om un in format ive d ig i t cond i t ion were ana lyzed as a funct ion of S O A (100, 300, 600, and 900 ms), number cue (1 ,3 ,6 ,9) , and target pos i t ion (left, r ight, up , down) . The analysis ind icated that, once again, cue type x target pos i t ion interact ion was not 3 As in Experiment 1, order of cue presentation and cue predictiveness was included as a between-subject variable in a preliminary analysis. First, to assess any possible effects of cue type presentation order, mean correct RT scores were subjected to a between (cue order: arrow, number) within (cue type, cue predictiveness, SOA, validity) A N O V A . The analysis returned no significant main effect of order [F<1] and significant interactions of cue order x cue type [F (1, 46)= 9.1, p<. 01] and cue type, cue validity, and cue order [F (1, 46)= 5.66, p<. 05] indicating that again both cue types were responded to faster when they were received second and that larger effect were observed for arrow cues regardless of whether they were received first or second. Next, to examine whether validity effects varied with order of cue predictiveness, it (nonpredictive, predictive) was entered as between-subject variable in three-way between within A N O V A conducted on each of the two cues types separately including cue predictiveness, SOA, and validity as within-subject factors. Again, for both arrow and number cues, no main effects of cue predictiveness order were observed [both Fs<l]. The only significant interaction was observed in arrow condition between cue validity and order [F (1, 46)= 4.7, p<. 05] reflecting the fact that overall invalid trials were overall slower when predictive cues were received second. No other interactions involving cue predictiveness and order of the presentation were significant [all Fs< 2.3, all ps>.085]. 22 significant [F (9, 423)= 1.4, p>. 15] demonstrating that any spatial bias induced by perception of numerically low and high numbers was not present in our data. As such, once again, we are confident that orienting effects observed in response to spatially predictive digit cues accurately reflects the contribution of endogenous orienting alone. The results obtained in Experiment 2 replicate those from Experiment 1. Spatially nonpredictive directional arrow cues were once again effective in eliciting reflexive orienting while spatially nonpredictive digit cues were not. In addition both predictive arrow and number cues were effective in eliciting volitional orienting. Spatially predictive arrow cues produced the largest orienting effects as illustrated in Figure 3c. Discussion Experiment 2 manipulated cue type and cue directionality within the same group of observers. This manipulation allowed for a direct comparison of the magnitudes of the orienting effects generated by our four conditions (arrow nonpredictive; arrow predictive; number nonpredictive; number predictive). Thus, any differences between the orienting magnitudes were now independent of group differences because Experiment 2 was carried out within-subjects. The results mirrored those of Experiment 1, and indicated that a spatially predictive arrow cue, that was traditionally used in a range of behavioral and neuroimaging studies of human volitional attention, triggers a shift of spatial attention that is reliably larger than the unconfounded measure of voluntary orienting measured by spatially informative symbolic digit cues. General Discussion As outlined in the introduction, for over twenty years, the Posner paradigm has been widely used in research investigations and predictive arrow cues have almost exclusively been used to elicit voluntary orienting responses. However, several recent findings (e.g., Hommel et al, 2001; Tipples, 2002; Ristic et al, 2002) demonstrate that central arrow cues, by a virtue of their inherent directionality, trigger a reflexive shift of 23 spat ia l attention. T w o exper iments reported here examined the differences between the behav iora l index of vo l i t iona l attention w h e n tr iggered b y spat ia l ly pred ict ive central d i rect ional cues, such as arrows, and w h e n tr iggered b y pu re l y symbol i c nondirect ional in format ive cues, that on their o w n do not produce a shift i n attent ion automatical ly. In Exper iment 1 we dissociated or ient ing i n response to spat ia l ly pred ict ive direct ional and nondi rect iona l attentional cues. The data demonstrated differences between the magni tudes of vo l i t iona l or ient ing when tr iggered b y the cue that, b y the v i r tue of its s t imulus characteristics, does not trigger a ref lexive attentional shift and the or ient ing effect tr iggered b y the predict ive ar row cues. Exper iment 2 repl icated this in i t ia l result, and also demonstrated that a difference i n magn i tude of attentional or ient ing i n response to d irect ional a r row and nondi rect iona l d ig i t cues was signif icant. To address the in i t ia l quest ion per ta in ing to the va l i d i t y of the classic paradigms, the data f r om the two exper iments indicate that the or ient ing i n response to spat ia l ly in format ive a r row cues does not equal either the ref lexive or pure vo lunta ry effect. Instead, the large or ient ing effect el ic i ted by spat ia l ly pred ic t ive a r row cues appears to be un ique such that the magn i tude of this or ient ing effect surpasses bo th ref lexive and vo l i t iona l effects. In other words , its magn i tude cou ld not be accounted for s imp ly by the add i t i on of the ref lexive and vo luntary or ient ing effects (see F igures 2c and 3c). A l t h ough this interpretat ion represents a p laus ib le account for the data, it is st i l l unclear whether this large or ient ing effect arises as a result of a un ique interact ion between ref lexive and vo l i t iona l or ient ing systems or whether it represents an isolated effect seen on ly w i t h spat ia l ly in format ive direct ional cues, such as arrows. If the effect is largely due to the interact ion between the two or ient ing mechanisms, one w o u l d predict that the s imi lar results w o u l d emerge w h e n other attentional cues that tr igger ref lexive or ient ing w h e n spat ia l ly un in format ive and vo lun ta ry or ient ing when presented as in format ive of the target are emp loyed (e.g., abrupt per iphera l onsets). 24 Add i t i ona l l y , since we measured or ient ing response to two specif ic s t imu l i on ly, it is u n k n o w n whether the s imi lar superadd i t ive behav iora l result cou ld be observed for other informat ive d irect ional cues presented at central f ixat ion, such as eye gaze d i rect ion or head or ientat ion. These questions are left for future studies. Perhaps the most important imp l i ca t ion of f ind ings reported here concerns the va l i d i t y of past exper imenta l results reported i n the attentional l i terature over the past two decades. O u r results demonstrate two very important points. First, w e replicate prev ious results ind icat ing that ref lexive attentional shifts occur w h e n central nonpred ic t ive cues are presented. A s such these f ind ings raise concerns about the or ig ina l d is t inct ion between ref lexive and vo luntary attentional systems; a theoretical f ramework that was g rounded i n dist inct behav iora l pa rad igms emp loy i ng per iphera l onsets and informat ive central s t imu l i as attentional cues. Second, our data demonstrate that past invest igat ions of vo l i t iona l attent ion that used pred ic t ive central a r row s t imu l i were i n fact not assessing the contr ibut ion of vo l i t iona l attent ion alone, but rather a specif ic or ient ing response el ic i ted b y a predict ive d i rect ional cue (i.e., arrow). Mo r e general ly, these data cast doubt on the va l id i ty of the classic Posner pa rad igm, wh i ch rout ine ly used predict ive a r row cues, and, consequently, the accumulated knowledge about the propert ies of contro l led attentional processes m u c h of wh i c h has been g rounded i n this t radi t ional exper imental procedure. For example, the classical pa rad igms have been over the years typ ica l ly adopted as an exper imenta l default for neuropsycho log ica l (e.g., Rafa l & Robertson, 1995), neu ro imag ing (Mangun & H i l l y a rd , 1990; Corbetta et al , 1993; Hopf inger , Buonocore & M a n g u n , 2000; Corbetta & Shu lman, 2002), and deve lopmenta l (e.g., Brodeur & Enns, 1997; Go ldberg , Mau r e r & Lewis , 2001) invest igat ions, and as such have been the foundat ion on w h i c h m u c h of the current know ledge about h u m a n attentional processes and the resu l t ing scientif ic theories of attentional operat ions have been bui l t . 25 References Brodeur , D, A . & Enns, J, T. (1997). Cover t v i sua l or ient ing across the l i fespan. Canad i an Tournal of Exper imenta l Psycho logy, 51, 20-35. Corbetta, M. , M i e z i n , F.M., Shu lman, G.L., & Petersen, S.E. (1993). A PET s tudy of v isuospat ia l attention. Journal of Neurosc ience, 13,1202-1226. Corbetta, M & Shu lman, G . L. (2002). Cont ro l of goal-directed and s t imu lus-dr iven attention i n the bra in. Na tu re Reviews: Neurosc ience, 3, 201-215. Dr iver , J., Dav i s , G., R icc iarde l l i , P., K i d d , P., Maxwe l l , E., & Baron-Cohen, S. (1999). Gaze percept ion tr iggers v isuospat ia l or ient ing b y adults i n a ref lexive manner. V i sua l Cogn i t i on , 6,509-540. Fisher, M , H. , Caste l , A , D., D o d d , M , D., & Pratt, J. (2003). Perce iv ing numbers causes spat ia l shifts of attention. Na tu re Neurosc ience, 6, 555-556,. Fr iesen, C . K., & K ingstone, A . (1998). The eyes have it!: Ref lex ive or ient ing is tr iggered by nonpred ic t ive gaze. Psychonomic Bu l le t in and Rev iew, 5,490-495. Goldberg,. M . C, Maure r , D, & Lewis , T, L. (2001). Deve lopmenta l changes i n attention: the effects of endogenous cue ing and of distractors. Deve lopmenta l Science, 4, 209-219. H o m m e l , B., Pratt, J., Co lzato, L. & God i jn , R. (2001). Symbo l i c contro l of v i sua l attention. Psycho log ica l Science, 12,360-365. Hopf inger , J. B., Buonocore, M. , H. , & M a n g u n , G., R. (2000). The neura l mechanisms of t op -down attentional control . Na tu re Neurosc ience, 3 (3), 284-291. Johnson, M , H . (1997). Deve lopmenta l Cogn i t i ve Neurosc ience. Cambr idge , M A : B lackwe l l Publ ishers. Johnson, M . H , Posner, M . I, & Rothbart, M . K. (1991). Componen t s of v i sua l or ient ing i n ear ly infancy: Cont ingency learning, ant ic ipatory l ook ing and disengaging. Journal of Cogn i t i ve Neurosc ience, 3, 335-344. 26 Jonides, J. (1981). Vo lun ta r y versus automatic contro l over the mind ' s eye's movement . In J. B. L o n g and A . D. Badde ley (Eds.), A t ten t ion and Performance IX (pp. 187-203). H i l l sda le , N J : E r l baum. K ingstone, A , Smi lek, D, Rist ic, J, Fr iesen, C . K, & Eas twood , J, D. (in press). A t ten t ion researchers! It is t ime to look at the real wo r l d . Cur ren t Di rect ions i n Psycho log ica l Science. Langton, S. R. H. , & Bruce, V . (1999). Ref lex ive social or ient ing. V i s ua l Cogn i t ion , 6, 541-567. Langton , S, R, H . (2000). The mutua l inf luence of head and gaze or ientat ion i n the analysis of social attention direct ion. Quar te r l y l ou rna l of Exper imenta l Psycho logy, 53A, 825-845. Langton, S, R, H . & Bruce, V . (2000). Y o u must see the point: Au toma t i c process ing of cues to the d i rect ion of social attention. l ou rna l of Exper imenta l Psychology: H u m a n Percept ion and Performance, 26, 747-757. M a n g u n , G . R, & H i l l y a r d , S. A . (1990). E lect rophys io log ica l studies of v i sua l selective attent ion i n humans , In A . B. Schiebel & A . F. Wechs ler (Eds.) Neu rob io l ogy and H i ghe r Cogn i t i ve Func t ion (pp. 271-295). N e w York , N Y : The Gu i l f o r d Press. Mu l l e r , H , J. & F ind lay , J, M . (1988). The effect of v i sua l attent ion on per iphera l d i sc r im inat ion thresholds i n s ingle and mu l t ip le element d isp lays. A c t a Psychologica, 69,129-155. Mu l l e r , H . J., & Rabbitt, P. M . A . (1989). Ref lex ive and vo lun ta ry or ient ing of v i sua l attention: t ime course of act ivat ion and resistance to in ter rupt ion. l ou rna l of Exper imenta l Psychology: H u m a n Percept ion and Performance, 15,315-330. Posner, M . I. (1980). Or ien t ing of attention. Quar te r l y l ou rna l of Exper imenta l Psycho logy, 32,3-25. 27 Posner, M , I. (1992). A t tent ion as a cognit ive and neura l system. Cur rent Direct ions i n Psycho log ica l Science, 1,11-14. Posner, M . I., & Cohen , Y. (1984). Components of V i s ua l Or ient ing . In H . Bouma & D. G . Bowhu i s (Eds.), A t tent ion and Performance X (pp. 531-556). H i l l sda le , NJ : E r l baum. Posner, M . I, & Petersen, (1990). The attention system of the h u m a n bra in. A n n u a l Rev iew of Neurosc ience, 13, 25-42. Rafa l , R, D., Ca labres i , P, A. , Brennan, C , W., & Sciolto,T, K. (1989). Saccade preparat ion inhib i ts reor ient ing to recently attended locations. Tournal of Exper imenta l Psychology: H u m a n Percept ion and Performance, 15, 673-685. Rafal , R. & Robertson, L. (1995). The neuro logy of v i sua l attention. In M . S. Gazzan iga (Ed.), The Cogn i t i ve Neuros icences (pp. 625-648). Cambr idge , M A : M I T Press. Rensink, R. A. , (1995). VScope 1.2.7, M i c r oPsy ch Software, Vancouver , Br i t i sh Co l umb ia , Canada , copyr ight 1991 - 1995. Ristic, J., Fr iesen, C. K., & K ingstone, A . (2002). A r e eyes special? It depends on h o w you look at it. Psychonomic Bu l le t in and Rev iew, 9,507-513. T ipp les, J. (2002). Eye gaze is not un ique: Au tomat i c or ient ing i n response to non in format ive arrows. Psychonomic Bu l le t in and Rev iew. 28 C H A P T E R 2 29 Introduct ion Deve lopment of attentional processes p lays a large role i n ear ly organizat ion of behav ior such that it inf luences early control of arousal levels, distress management, self-regulat ion as we l l as the organizat ion of ear ly social interactions (e.g., Rothbart, Posner & Ros icky, 1994; Posner & Rothbart, 2000). Behav iora l invest igat ions of h u m a n attention indicate that wh i l e attention general ly moves i n conjunct ion w i t h eye movements (overt attention), it can also be al located to an. object or a pos i t ion i n space independent ly of a shift i n gaze pos i t ion (covert attention) (e.g., K l e i n , K ingstone, Pontefract, 1992; K l e i n & Shore, 2000). Both covert and overt attent ion can be tr iggered either external ly by s t imulus propert ies (e.g., transient abrupt-onset events) or by internal goals and expectations of an observer (Mu l l e r & Rabbitt, 1989; Posner, 1980; Jonides, 1981). Deve lopmenta l studies conducted to date indicate that these two attentional systems (i.e., exogenous and endogenous, respectively) exhibit differential maturat iona l rates, and as such p lay a large role i n the deve lopment of relevant emergent behaviors (e.g., H o o d , A t k i n son & Bradd ik , 1998). Deve lopment of ref lexive or exogenous attention has typ ica l ly been invest igated in the exper imenta l parad igms man ipu la t ing abrupt per iphera l onsets, m i r ro r i ng those emp loyed w i t h adults (e.g., Posner, 1980). W h e n ref lexive or ient ing is examined us ing manua l response t ime (RT) measures i n a cross-sectional sample of ch i ldren, between ages of 5 and 18, the results indicate that ref lexive or ient ing responses exhibit l ittle or no comparat ive change over the l i fespan (e.g., Brodeur, Tr ick & Enns, 1997). In contrast, studies of vo l i t iona l or endogenous attention indicate the presence of marked deve lopmenta l differences in the control of attention. Invest igations of covert vo luntary attent ion suggest that adul t - l ike contro l of attentional a l locat ion is observed on ly at about 8 years of age (e.g., Brodeur, Tr ick & Enns, 1997). These ma rked behav iora l differences i n a l locat ion of covert ref lexive and vo l i t iona l attention have largely been 30 attr ibuted to different maturat iona l rates of the unde r l y i ng b ra in mechan isms where ref lexive attention is thought to be contro l led b y earl ier deve lop ing subcort ical (e.g., super ior col l iculus) and cort ical (parietal lobe) structures whereas the control of vo l i t iona l attention is accompl ished b y later deve lop ing frontal cort ical areas (e.g., Johnson, 1997). Recent ly, Rist ic, Fr iesen & K ingstone (2002) reported that ch i ld ren as young as 4 years of age w i l l shift their attention ref lexively to where a spat ia l ly nonpred ic t ive a r row is directed. A group of ch i ld ren (mean age 4. 5 years) and a compar i son group of adults detected per iphera l targets appear ing to the left or r ight of a central spat ia l ly nonpred ic t ive a r row cue. At tent iona l or ient ing was samp led at tempora l intervals of 195,600, and 1000 ms. The results revealed that, overa l l , bo th groups responded faster to cued compared to uncued targets. This effect appeared less than 200 ms after the a r row cue was presented and it persisted, and grew i n magn i tude, as the cue-target S O A was extended to approx imate ly 1000 ms. Th is f i nd ing contrasts w i t h the on ly other invest igat ion that has s tud ied the effect of central a r row cues on attentional or ient ing i n such y oung ch i ldren. Brodeur and Enns (1997, Exper iment 1) asked three groups of ch i ldren (mean ages of 6,8, and 10 years) to ident i fy a target (X or O) that appeared either near to, or far f rom, a central a r row cue. This a r row cue correct ly ind icated the target hemi f ie ld o n 80% of the trials (i.e., it was spat ia l ly predict ive). Or i en t ing effects were examined at S O A s of 133,250 and 450 ms. Wh i l e adults showed signif icant or ient ing effects (Uncued-Cued RT) for a l l cue-target intervals, ch i ld ren showed signif icant or ient ing effects for 133 ms S O A only. In contrast to the Rist ic et a l . (2002) study, the cu ing effect for the youngest g roup of ch i ldren dec l ined as the S O A increased, d isappear ing altogether b y the 500 ms S O A . Based on these data, the authors argued that ch i ldren 6 years of age were able to orient attention vo l i t iona l l y i n response to the predict ive nature of a central a r row cue, but on ly for a 31 short per iod of t ime. Subsequent studies showed that after age 8, l i ke adults, ch i ldren are able to sustain vo l i t iona l or ient ing for longer durat ions (Goldberg, Mau re r & Lewis , 2001). Insofar as these two studies can be attr ibuted to ref lexive and vo l i t iona l attention respectively, the imp l i ca t ion is that for y oung ch i ld ren ref lexive or ient ing is rap id and long- last ing and vo l i t iona l or ient ing is r ap id and short- l ived (the latter interpretat ion be ing favored b y Brodeur & Enns). Howeve r , as we have s hown i n Chapter 1, there are very good reasons to quest ion the interpretat ion of studies that have emp loyed pred ic t ive a r row cues, as it was used b y Brodeur & Enns (1997). The attention effect they observed w i t h a pred ic t ive a r row cue cou ld reflect ref lex ive attention, vo l i t iona l attention, or some combinat ion of the two. Nevertheless, the fact remains that the Brodeur & Enns (1997) s tudy suggests that pred ic t ive a r r ow cues p roduce a very different pattern of results than wha t was observed for adults i n Chapter 1. Specif ical ly, where the attention effect for predict ive arrows grew for adults as the cue-target interval extended across t ime, the effect dec l ined for ch i ld ren i n the Brodeur and Enns ' (1997) invest igat ion. Whether this effect shou ld be attr ibuted to short- l ived vo l i t iona l attention, or a combinat ion of ref lexive and vo l i t iona l attention that is very different f r om what was observed for adults, is very much an open quest ion. In Chapter 2 we addressed this issue d i rect ly by emp loy ing the logic used successful ly i n Chapter 1, whereby we dissociated or ient ing responses of y oung ch i ld ren to spat ia l ly predict ive and nonpred ic t ive d i rect ional (arrow) and nondi rect iona l (shape) central attentional cues. Exper iment 1 In order to examine this issue, first we presented a g roup of y oung ch i ldren and a group of adults w i t h a cu ing task i n wh i c h central d i rect ional a r row cues served as f ixat ion s t imul i . Further, to assess whether the or ient ing tr iggered b y spat ia l ly 32 predict ive a r row cues indeed reflected vo luntary or ient ing, i n a subsequent man ipu la t i on we compared or ient ing i n y oung ch i ld ren and adults i n response to central geometr ic shape cues. Instead of us ing d ig i t cues as emp loyed i n Chapter 1, here we u t i l i zed central geometric shape (circle or square) as a symbo l i c target locat ion predictor i n order to equate fami l iar i ty w i t h the cue between y oung ch i ld ren and adults. Th is nondi rect iona l cue man ipu la t i on represents a cruc ia l test of whether the attentional or ient ing i n response to predict ive a r row cues i n ch i ld ren is ind icat ive of vo lunta ry or ient ing or not. Both d i rect ional (arrow) and nondi rect iona l (shape) cues were presented as either nonpred ic t ive (p=. 5) or pred ict ive (p=. 8) of the target locat ion. A l l man ipu la t ions were carr ied out between subjects such that each cue type by cue predict iveness cond i t ion was carr ied out on a separate group of part ic ipants i n each age group. In order to ensure that we were i n fact examin ing covert or ient ing responses eye movements were mon i tored i n a l l exper imental condit ions. Me t hod Part ic ipants A total of 60 ch i ld ren and 60 undergraduate students part ic ipated i n the present study. Fifteen ch i ld ren and 15 adults were assigned to each of the cue type x cue predict iveness condi t ions (arrow nonpredict ive; a r row predict ive; shape nonpredict ive; shape predict ive). The ages for the ch i ld ren i n each of the four condi t ions were: (1) nonpred ic t ive a r row cond i t ion (9 males) 3 years, 5 months to 6 years, 2 months (mean age 4 years, 5 months); (2) predict ive ar row cond i t ion (8 males) 4 years, 11 months to 5 years, 11 months of age (mean age 5 years, 3 months); (3) nonpred ic t ive shape condi t ion (7 males) 3 years, 5 months to 5 years, 4 months (mean age 4 years, 3 months); (4) predict ive shape cond i t ion (9 males) 3 years, 4 months to 4 years, 9 months (mean age 4 years, 4 months). A l l ch i ld ren were recruited f r om local Vancouver Daycare Centers 33 w i t h the in fo rmed consent of the parents and the daycare centers. Undergraduate students were recrui ted f r om U B C Psycho logy Subject Poo l , and a l l observers completed the exper iment i n exchange for course credit. Appa ra tus and S t imu l i The s t imu l i were presented on a Mac in tosh Powerbook 3200c laptop computer connected to an external keyboard. The s t imu l i were presented on a 12-inch L C D color screen set to black and whi te . VScope 1.2.7 software (Rensink, 1995) was used to control s t imulus presentat ion and record response latencies and accuracies. Part ic ipants responded by press ing the spacebar key on the external keyboard , pos i t ioned i n front of the computer, wh i c h was marked w i t h red tape to r em ind ch i ld ren of correct response key. Eye movements were mon i to red on-l ine b y the exper imenter us ing an external m i r ro r pos i t ioned above the computer screen. The s t imu l i and sequence of events are presented i n F igure 1. A l l s t imu l i were black l ine d raw ings presented on a wh i te background. The a r row st imulus was created by attaching an a r rowhead and an arrowta i l (each 45-degree or iented l ine measured 0.6°) to the both ends of a hor i zonta l l ine measur ing 1.3° i n length. The a r row st imulus was 2.5° l ong as measured f r om the tip of the ar rowhead to the end of the arrowta i l , and it was a lways pos i t ioned at the center of the screen. In a nond i rect iona l cue condi t ion, one of two geometr ica l shape cues, l ine drawings of a circle and a square, served as central f ixat ion s t imul i . The circle and square both measured 1.9° i n length and height. A sma l l central cross (subtending 1° v i sua l angle), served as a f ixat ion po int i n the nondi rect iona l cue condit ions. The response target was a black asterisk subtend ing 0.7°. The target a lways appeared 5° to the left or r ight of center as measured f r om the center of the cue to center of the target. 34 CD E Directional Cue > — > * 675 ms Cue Onset 100 or 900 ms SOA Target Onset until response or 2300ms Nondirectional Cue + Figure 1 illustrates stimuli and timing sequence for directional (arrow and nondirectional (geometric shape) cue conditions. A straight line or a fixation point appeared on the screen for 675 ms. Then, an arrow pointing left or right, or central circle or square appeared on the screen. The target appeared centered across horizontal meridian, either on the left or right side of the cue after 100 on 900 ms. Both the cue and the target remained on the screen until response was made or for 2700ms, whichever came first. Intertrial interval was 516 ms. Note that the stimuli are not drawn to scale. Design For the nonpredictive cue condition, the target location was unrelated to the cue. For instance, in the arrow cue condition, the left or right direction of the arrow was randomly determined as was the left or right location of the target. Thus, the target could appear at the cued location (location pointed at by the arrow) or at the uncued location (location not pointed at by the arrow) with a probability equal to chance (p=. 5). Similarly, for nondirectional shape cue, the cue was either a circle or a square, and the location of the target was unrelated to the type of shape cue. For the predictive arrow cue condition, the arrow direction was again chosen at random, but now the target 35 appeared at the locat ion cued by the a r row most of the t ime (p=. 8) and occasional ly at the uncued locat ion (p=. 2). S imi lar ly , for the shape cue cond i t ion, one shape indicated that the target was l i ke ly to appear on the left (p=. 8) and the other shape indicated that the target was l i ke ly to appear on the r ight (p=. 8). The order of w h i c h shape indicated wh i c h side was counterbalanced across part ic ipants. The t imecourse of attentional a l locat ion was examined b y samp l ing the performance at two cue-target s t imulus onset asynchronies (SOAs): 100 and 900 ms. The two delay intervals were d istr ibuted equal ly throughout the exper iment. In add i t ion, i n approx imate ly 6% of a l l trials the target was not presented. These catch trials were r andom ly seeded throughout the experiment. Each part ic ipant, except the two ch i ldren w h o completed on l y one block in response to nonpred ic t ive shape, completed a total of 100 exper imenta l trials that were d i v i ded i n two b locks of 50 trials each. A l l part ic ipants, except one ch i ld , completed the two b locks of trials i n succession. A practice b lock of about 10 tr ials was r un at the beg inn ing of the exper imental session. A fai lure to emphas ize response speed (mean detection RT greater than 500 ms for adults and 900 ms for chi ldren) resulted i n 1 adul t and 2 ch i ld ren be ing replaced. Procedure A l l trials began w i t h the presentat ion of a straight l ine or a f ixat ion cross for 675 ms. Then, i n the a r row condi t ion, an ar rowhead and an ar rowta i l appeared at both ends of the l ine to create an a r row po in t ing left or r ight. In the nond i rect iona l cue condi t ion, center shape, either a circle or a square, appeared at the center of the screen. A f ter 100 or 900 ms the target appeared on either left or r ight side of the screen. Each tr ia l terminated on response, or after 2300 ms wh ichever came first. The intertr ia l interval was 516 ms. 36 A l l ch i ld ren were tested at a daycare faci l i ty. T w o exper imenters were present w i t h the ch i ld ren at a l l t imes. One experimenter sat beside the ch i l d and ensured that each part ic ipant was respond ing w i t h the proper key. The other exper imenter was pos i t ioned beh ind the part ic ipant and recorded eye movements b y observ ing part ic ipants eyes i n the mi r ro r and the exper imental sequence on the screen. O n every tr ia l , the exper imenter judged whether an eye movement h a d occurred and if it d i d the exper imenter recorded the tr ial number on wh i c h the saccade was made. Eye movements that were made either i n ant ic ipat ion of the target pos i t ion or saccades that were made to the target w h e n it appeared on the screen were recorded. A f ter agreeing to take part i n the study, the ch i ld ren sat i n front of the computer and were centered w i t h respect to the screen and keyboard (v iew ing distance approx imate ly 57 centimeters). The ch i ld ren were then to ld h o w the s tudy was go ing to proceed. In a l l condit ions ch i ldren were asked to "catch the snowf lake" as fast as they cou ld w h e n it appeared on the screen by press ing the red key on the keyboard. They were to ld that they w o u l d see either an ar row or circle or square presented at the center of the screen before the snowf lake appeared. In the nonpred ic t ive cue condit ions, the ch i ld ren were in fo rmed that a r row direct ion (or shape) d i d not indicate where the snowf lake w o u l d appear. In the predict ive cue condit ions, the ch i ld ren were in formed that a r row d i rect ion w o u l d i n fo rm them about snowf lake 's locat ion most of the time. S imi la r ly , for the pred ic t ive shape cue cond i t ion ch i ld ren were i n fo rmed about the predict ive re lat ionship between the cue and the l i ke ly target locat ion. The experimenter ensured that a l l ch i ld ren understood instruct ions fu l l y before commenc ing the exper iment. Undergraduate students were tested i n the laboratory. Exper imenta l procedure and instruct ions para l le led those emp loyed w i t h ch i ldren, w i t h one except ion that on ly one exper imenter was present i n the testing r oom at a l l t imes. 37 Results M e d i a n RT was calculated for each part ic ipant. The interpart ic ipant mean RTs for both age groups are i l lustrated i n F igure 2 and presented i n Table 1 for adults and Table 2 for ch i ldren. Incorrect key presses, t imed-our responses, and false alarms counted as errors and were removed f rom the analysis. Table 3 presents overa l l error rates as we l l as saccadic eye movement rate 1. Arrow Shape Condition* nonpredictive predictive nonpredictive predictive M SD M SD M SD M SD 100 ms SOA Cued 385.5 67.3 415.2 93.3 400.5 83.2 379.2 68.5 Uncued 389.9 53.6 442.9 95.9 397.9 82 391.8 74.6 900 ms SOA Cued 337.9 47.7 343.6 66.2 358.7 47 336 47.4 Uncued 351.1 54.4 389 99 353.5 47.5 352.1 52.1 Table 1. Mean RTs and standard deviations for adults. *Since there was no true "cued" and "uncued" target conditions in nonpredictive shape condition, corresponding numbers represent mean RTs for targets appearing on the left and right side location for the two SOA intervals. 1 In order to ensure that we were measuring covert attention, number of recorded eye movements made either to the target or in anticipation of the target was included a covariate in preliminary analyses. Saccadic eye movements did not account for significant variation present in the data, and as such did not change reported results. A l l analyses reported here show actual p values without eye movement rate included as a covariate. 38 Arrow Shape Condition* nonpredictive predictive nonpredictive predictive M SD M SD M SD M SD 100 ms SOA Cued 680.6 153 685.4 153.5 795.4 211.7 726.4 145.3 Uncued 686.8 136.6 724.2 168.5 780 207 773 208.7 900 ms SOA Cued 603.9 93.7 585.2 93.6 671.3 175 643.6 128.3 Uncued 630.8 118.74 618.3 146.3 711.7 212.6 646.1 149 Table 2. Mean RTs and standard deviations for children. *Since there was no true "cued" and "uncued" target conditions in nonpredictive shape condition, corresponding numbers represent mean RTs for targets appearing on the left or right side locations for the two SOA intervals. Adults Children Condition Arrow TO IK FA E M TO IK FA EM Nonpredictive . 45% 0 1. 12% 2. 5% 3. 4% . 64% 25.8% 5% Predictive 0 0 2. 2% 3. 4% 3. 3% . 14% 19. 3% 8% Shape Nonpredictive 0 0 2. 2% 2% 4. 4% .2% 24% 6.1% Predictive 0 0 0 3. 5% 1. 7% . 28% 25. 6% 7. 3% Table 3. Error rates and eye movement rates for all conditions. TO (timed out responses), IK (incorrect key presses), FA (false alarms), E M (eye movements). 39 ADULTS CHILDREN A:Directional Cue 5 0 5 r 4 5 7 4 0 9 3 6 1 3 1 3 2 6 5 nonpredictive predictive arrow arrow \ • • 1 1 • I 1 0 0 9 0 0 SOA 1 0 0 9 0 0 SOA B:Nondirectional Cue 5 0 5 4 5 7 <u E i-c o ti ID CU a: c ID CU 4 0 9 3 6 1 3 1 3 2 6 5 nonpredictive predictive shape shape s • i i 1 0 0 9 0 0 1 0 0 9 0 0 SOA SOA Cuing Effects 8 2 0 r 7 7 2 cu E 7 2 4 i-c o '•rj 6 7 6 ID CU a: c ID 6 2 8 CU 5 8 0 nonpredictive predictive arrow arrow \ \ \ \ N \ * "u . \ 1 0 0 9 0 0 SOA 1 0 0 9 0 0 SOA uncued cued 8 2 0 g 7 7 2 cu j ! 7 2 4 c o ti ID & 6 7 6 cu 5 8 0 nonpredictive predictive shape shape \ \ • i I • 1 0 0 9 0 0 SOA 1 0 0 9 0 0 SOA 1 0 0 9 0 0 SOA predictive arrow predictive shape —M— nonpredictive arrow Figure 2 shows results from Experiment 1. Results from two age groups are presented on the left (adults) and on the right (children). Figure 2a shows mean response times (RTs) plotted as a function of SOA and validity for nonpredictive and predictive directional arrow cues for both age groups. Figure 2b shows mean response times for nonpredictive and predictive central shape cues, with only an SOA effect plotted for nonpredictive shape condition. Figure 2c plots the difference between uncued and cued RTs for both adults (left panel) and children (right panel). 40 The results for the nonpred ic t ive a r row cond i t ion (see F igure 2a) were ana lyzed us ing a three-way A N O V A where age group (chi ldren and adults) was inc luded as a between-subject factor, and S O A and va l id i t y were within-subject factors. The analysis ind icated that a l l ma i n effects were signif icant: group [F (1,28)= 88.92, p< .0001], S O A [F (1,28)= 28.02, p<. 0001], and va l id i ty [F (1,28)= 4.43, p<. 05]; and that no interactions were s ignif icant [all ps>. 05]. These f ind ings reflect the fact that wh i l e ch i ldren were s lower overa l l to respond to targets than adults, the effects of S O A and cue cond i t ion were s imi lar for both age groups. Specif ical ly, RT dec l ined as S O A increased, ref lecting a classic foreper iod effect; and RT was shorter for cued than uncued locations at both S O A s ref lect ing the fact that attention was shifted ref lex ive ly to the locat ion indicated by a spat ia l ly nonpred ic t ive a r row cue as reported p rev ious ly b y Rist ic et a l . (2002). M e d i a n correct RTs for the two age groups that received pred ic t ive ar row cues were subjected to the same three-way A N O V A . This analys is also returned three signif icant ma i n effects: age group [F (1, 28)= 42.78, p<. 0001], S O A [F (1, 28)= 25.34, p< .0001], and va l i d i t y [F (1, 28)= 27.88, p< .0001]. N o interactions between group and va l id i t y were observed [all ps> .05]. A s the F igure 2a i l lustrates, bo th age groups p roduced s imi lar behav iora l responses w h e n presented w i t h pred ic t ive a r row cues. Wh i l e ch i ld ren were overa l l s lower, both age groups responded faster as the time between the cue onset and target presentat ion increased. Furthermore, both age were faster to respond to targets appear ing at the pred icted locat ion for both short and long S O A intervals. In summary , the data for the a r row cue condi t ions indicate that both ch i ldren and adults detected cued targets faster than uncued targets regardless of whether the central a r row cue was spat ia l ly pred ict ive or not. In contrast to nonpred ic t ive a r row cues, nonpred ic t ive shape cues were not effective i n e l ic i t ing ref lexive or ient ing i n either age group. Interpart ic ipant means of 41 med ian RTs calculated for each g roup of part ic ipants were subjected to the subsequent analyses for nondi rect iona l shape condi t ion. They are i l lustrated i n F igure 2b. A four -way A N O V A that i nc luded age group as a between-subject var iable, S O A , cue type (circle or square), and target pos i t ion (left or r ight), as within-subject var iables ind icated that again overa l l ch i ld ren responded s lower than adults [F (1,28)= 66.67, p<. 0001] and that on average both age groups responded faster as the S O A interva l lengthened [F (1, 28)= 14.02, p<. 001]. S O A x cue type x age group interact ion was also s ignif icant [F (1,28)= 4.92, p<. 05] ref lect ing that at the 900 ms S O A ch i ldren generated faster responses when the central cue was a circle than w h e n it was a square. Important ly , however , cue type d i d not interact w i t h target pos i t ion [F<1] ind icat ing that the shape cues d i d not tr igger any specif ic d i rect ional or ient ing effects i n either age group. Thus, shape cues represent a va l i d compar i son set against wh i c h any contr ibut ion of vo lunta ry attention can be assessed when the same cues are man ipu la ted as spat ia l ly predict ive of target locat ion. The data for both age groups are i l lustrated i n F igure 2b, show ing mean RTs as a funct ion of S O A . Data for pred ict ive shape cues were ana lyzed us ing a three-way A N O V A , w i t h age group as between-subject var iable, and S O A and va l i d i t y as within-subject variables. The analys is returned signif icant ma i n effects of g roup [F (1, 28)= 64.01, p<. 0001] and S O A [F (1, 28)= 46.69, p<. 0001] as we l l as a s ignif icant S O A x group interact ion [F (1,28)= 8.75, p< .01] ref lect ing a larger decl ine i n RTs across S O A s (the foreper iod effect) for ch i ldren. The va l i d i t y effect was marg ina l l y s igni f icant [F (1,28)= 4.15, p< .052] as was the three-way interact ion between age group, S O A , and va l id i ty [F (1, 28)= 3.3, p<. 07]. Four direct ional pa i red t-tests were conducted to compare mean RT for cued and uncued condit ions across both S O A intervals for adults and ch i ldren. The analysis ind icated that, indeed, s ignif icant or ient ing effects were present at both S O A s 42 of 100 and 900 ms for adults and on ly at an early S O A of 100 ms for ch i ld ren [all ps <. 05]. D iscuss ion There are several f ind ings wo r t h no t ing i n the present invest igat ion. First, the data w i t h nonpred ic t ive ar row cues repl icated those reported b y Rist ic, Fr iesen & K ingstone (2002), demonstrat ing again that ch i ld ren and adults a l ike w i l l shift their attention to the locat ion ind icated b y a nonpred ic t ive a r row cue. A l s o i n agreement w i t h Rist ic et a l (2002), it was observed that w i t h a nonpred ic t ive a r row cue the difference between cued and uncued target locations tends to g row for ch i ldren as the cue-target S O A is increased. Second, and i n agreement w i t h Brodeur and Enns (1997), the present s tudy found that w h e n the a r row cue was predict ive, a l though both ch i ld ren and adults p roduced s imi lar performance patterns, the effect of the cue dec l ined substant ia l ly across S O A s for ch i ldren, but the effect of the cue increased across S O A s for adults. Th i rd , b y compar ing these performances w i t h nonpred ic t ive and predict ive a r row cues against performance w i t h a predict ive nondi rect iona l shape cue, several important ins ights regard ing attentional or ient ing were revealed. For adults, the cu ing effect for a predict ive a r row cue was found to exceed wha t w o u l d be expected by add ing the ref lexive attentional effect p roduced b y a nonpred ic t ive a r row cue and the vo l i t iona l attentional effect p roduced by a predict ive shape cue. A s was reported p rev ious ly i n Chapter 1, ref lexive and vo l i t iona l components appear to interact to p roduce a larger cu i ng effect for pred ict ive ar rows than w o u l d be expected based on pure measures of the two components alone. In contrast, for ch i ld ren we found that the vo l i t iona l attention effect for a shape cue was s ignif icant on l y at the 100 ms, consistent w i t h the proposa l of Brodeur and Enns (1997) that this f o rm of or ient ing can on ly be sustained br ief ly b y ch i ld ren younger than 8 years of age. Moreover , and again in 43 contrast with adults, it appears that the cuing effect of a predictive arrow reflects precisely what would be expected by adding the reflexive attentional effect produced by a nonpredictive arrow cue and the volitional attentional effect produced by a predictive shape cue (see Figure 2c). Indeed, it appears that the cuing effect observed for children with a predictive arrow at the 900ms S O A reflects the residual presence of reflexive orienting, as there is no volitional component present at this long SOA for a predictive shape cue, and the cuing effect for a predictive arrow converges with the cuing effect of a nonpredictive arrow. To summarize, our data demonstrate that reflexive attention operates in a very similar manner for adults and children, but that endogenous attention operates and combines with reflexive attention differently between these two groups. Adults can orient attention endogenously for a sustained period of time; and when this form of orienting is combined with reflexive attention to a directional stimulus, the results is a superadditive interaction for predictive arrow cues. Children younger than 6 years, on the other hand, can also orient attention endogenously, but only for a brief period of time; and when this form of orienting is combined with reflexive attention to a direction stimulus, the two effects are additive. This finding suggests that the superadditive interaction with predictive arrow cues, that is observed with adults but not with children, may depend on the development of sustained, adult-like, volitional orienting. These similarities and differences between adults and children as a function of type of attentional orienting dovetail with the fields' current understanding of the maturation of the brain areas that subserve reflexive and volitional orienting. Research results to date indicate that because reflexive orienting is mediated by early developing subcortical and cortical brain areas, such as the superior colliculus and the parietal lobe, its presence can be behaviorally measured shortly after birth (e.g., Clohessy, Posner, Rothbart & Vecera, 1991; Johnson, Posner & Rothbart, 1991; Valenza, Simion & Umilta, 44 1994). Moreover , when performance of adults and y oung ch i ld ren is compared behav iora l l y i n a cu ing parad igm, l itt le or no differences are observed between the two groups of part ic ipants (Enns & Brodeur , 1989; Brodeur , Tr ick & Enns, 1997; Rist ic et al , 2002). In contrast to ref lexive or ient ing, vo luntary attention deve lops m u c h s lower and exhibits adul t - l ike propert ies on l y about 8 years of age (e.g., Brodeur , Tr ick & Enns, 1997; Go ldberg , Mau re r & Lewis , 2001). This agrees w i t h the not ion that contro l led attention is mediated by mechanisms res id ing w i t h i n the frontal cortex (e.g., Corbetta & Shu lman , 2002), a b ra in reg ion that matures more s l ow l y than the par ieta l cortex or the super ior co l l i cu lus (Johnson, 1997). Rud imenta ry contro l of vo l i t iona l attention can be observed i n infancy, w i t h newborns able to make ant ic ipatory saccades towards the interest ing objects i n the env i ronment (e.g., Johnson, Posner & Rothbart, 1991; H o o d , A t k i n s on & Bradd ik , 1998; Posner, 2001). Howeve r , when vo l i t iona l or ient ing is examined us ing attentional cu ing tasks, the results reported here and i n the l iterature, indicate that at age 6 years vo luntary or ient ing is st i l l too immatu re to p roduce anyth ing but short- l ived or ient ing effects that qu i ck l y decrease i n magn i tude over t ime (Brodeur & Enns, 1997). 45 References Brodeur, A. , D., Tr ick, M. , I, & Enns, J. T. (1997). Selective attent ion over the l i fespan. In Burack A . J. & Enns J. T. (Eds.). At tent ion, Deve lopment , and Psychopatho logy (pp. 74-97). N e w York: Gu i l f o r d Press. Brodeur , D, A . & Enns, J, T. (1997). Cover t v i sua l or ient ing across the l i fespan. Canad i an Tournal of Exper imenta l Psycho logy, 51, 20-35. C lohessy, A . B, Posner, M . I, & Rothbart, M , K., & Vecera (1991). The deve lopment of inh ib i t ion of return i n early infancy. l ou rna l of Cogn i t i ve Neurosc ience, 3, 345-351. Corbetta, M & Shu lman, G . L. (2002). Con t ro l of goal-directed and s t imu lus-dr iven attention i n the bra in . Na tu re Rev iews: Neurosc ience, 3, 201-215. Enns, J. T, & Brodeur , D, A . (1989). A deve lopmenta l s tudy of covert or ient ing to per iphera l v i sua l cues. l ou rna l of Exper imenta l C h i l d Psycho logy, 48,171-189. Go ldberg , M . C, Maurer , D, & Lewis , T. L. (2001). Deve lopmenta l changes i n attention: the effects of endogenous cue ing and of distractors. Deve lopmenta l Science, 4, 209-219. H o o d , B. M , A t k i n son , J, & Bradd ik , O. J. (1998). Selection-for-action and the deve lopment of or ient ing and v i sua l attention, In J. E. R ichards (Ed.), Cogn i t i ve neuroscience of attention: A Deve lopmenta l perspect ive (pp. 219-251). N e w Jersey: E r l baum. Johnson, M . H . (1997). Deve lopmenta l Cogn i t i ve Neurosc ience. Cambr idge , M A : B lackwe l l Publ ishers. Johnson, M . H , Posner, M . I, & Rothbart, M . K. (1991). Componen ts of v i sua l or ient ing i n ear ly infancy: Cont ingency learning, ant ic ipatory l ook ing and disengaging. Tournal of Cogn i t i ve Neurosc ience, 3,335-344. 46 Jonides, J. (1981). Vo lun ta ry versus automatic contro l over the mind ' s eye's movement . In J. B. L o ng and A . D. Badde ley (Eds.), A t ten t ion and Performance IX (pp. 187-203). H i l l sda le , NJ : E r l baum. K l e i n , R. M , K ingstone, A . & Pontefract, A . (1992). Or i en t ing of v i sua l attention. In K. Rayner (Ed.), Eye Movements and V i sua l Cogn i t ion: Scene Percept ion and Reading. (pp. 46-63). No r th -Ho l l and : Elsevier Science Publ ishers B. V . K l e i n , R. M , & Shore D. I. (2000). Relat ions among modes of v i sua l or ient ing (commentary), In S. Monse l l and J. D r i ve r (Eds.) A t tent ion and Performance XVIII (pp. 195-209). Cambr idge , M A : M I T Press Mu l l e r , H . J, & Rabbitt, P. M . A . (1989). Ref lex ive and vo lun ta ry or ient ing of v i sua l attention: t ime course of act ivat ion and resistance to in ter rupt ion. Tournal of Exper imenta l Psychology: H u m a n Percept ion and Performance, 15, 315-330. Posner, M . I. (1980). Or ien t ing of attention. Quar te r l y Tournal of Exper imenta l Psycho logy, 32, 3-25. Posner, M . I, & Rothbart, M . K. (2000). Deve lop ing mechan isms of self-regulation. Deve lopment and Psychopatho logy, 12,427-441. Posner, M . I. (2001). The deve lop ing h u m a n bra in . Deve lopmenta l Science, 4, 253-387. Rothbart, M . K., Posner, M . I, & Rosicky, J. (1994). Or i en t ing i n no rma l and pathologica l development. Deve lopment and Psychopatho logy, 6, 635-652. Rensink, R. A. , (1995). VScope 1.2.7, M i c r oPsy ch Software, Vancouver , Br i t i sh Co l umb i a , Canada , copyr ight 1991 - 1995 . Ristic, J, Fr iesen, C. K, & K ingstone, A . (2002). A r e eyes special? It depends on h o w you look at it. Psychonomic Bu l le t in and Rev iew, 9,507-513. Va lenza , E, S im ion , F, & Umi l t a , C. (1994). Inh ib i t ion of return i n newbo rn infants. Infant Behav ior & Deve lopment , 17, 293-302. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0091108/manifest

Comment

Related Items