1P098 文脈手がかり効果の時空間的特性(2003年度 日本基礎心理学会第22回大会優秀発表賞)

The Japanese Psychonomic Society

NII-Electronic Library Service

The JapanesePsychonomic Society

TheJitPaneseJbuniatofPsvchonemic Sc'tence

2ee4,VoL 23,}"o.1,105-106

Summary

ofAwarded

PresentationIP098

Spatiotemporal

properties

of

effect

in

visual

the

contextual

cueing

search

Hirokazu

OGAwA,

Yuji

TAKEDA,

and

Takatsune

KuMADA

IVtitional

Jnstitute

ofAdvanced

industriai

Science

and Technolog),*

The repetition of the spatial

layouts

of search

items

implicitly

facilitates

visual search

(contextual

cueing effect; Chun &

Jiang,

1998). Wc investigated_the spatietempora] _propertiesof

the mechanism of contextual cueing by using a hybrid paradigm of visua] search and probe-dot

detection.The results showed thatfortherepeated layoutsdetectionof a _probc dot was faci]itated at asearch targetlocationand was inhibitedat distractorlocations,The effects were obtained at

a short

display-probe

SOA

(100

ms). These results suggest that in an early stage of vjsual

processing contextual cueing modulates attentional _processing

by

facilitation

tothe locationof

stimuli which are to be attencled to,and

by

inhibition'tothe

locations

of stimuli which are tobe ignored.

Key

wordsi contextuar cueing, visual search, probe

detection

Recently,Chun and

Jiang

(1998)

developed a new

methodology toexamine how an observer's

knowl-edge about invariantinformation ina visual scene

affects thecontrol mechanisrn of visual attention. In

theirexperiments, the participants perforrnedan

in-eMcient visua] search task. Inhalfof the trials,the

Iocations

of targetsand distractorsrcmaincd

con-stant throughout the entire experiment

{repeated

layouts),Intheremaining trials,thelocationsof the

targetswere unchanged,

but

thoseof thedistractors

were re-arrangcd

for

each trial

(random

layouts).

They

demonstrated

thattheparticipantscould finda

target more _quickly

in

the repeated layouts than in

the

random

layouts,

This

facjlitation

was referred to

asthe contextua] cueing effect.

Chun

and

Jiang

con-cluded thatrepeated layouts were

implicitly

learned

as contextual information, _guiding spatial attention to_{the potentialtarget}

_locations

_in

_the scene.

Although several studies have explored mecha-nisms related tocontextua] cueing,

it

is

stillunclear

how contextual information _guides visual attention

invisual search, To enforce attentional _{guidance to}

a target

location

using learned contextual

inforrna-tion,thereare two possible mechanisms tomodulate

*

_Visual

_Cognition

_Greup,

_Institute

_for

_Human

Science

and

Biomedical

Engineering, National

Instituteof Advanced Industrial Science and

Technology

(AIST),

Tsukuba Central 6, 1-1-1

Higashi,Tsukuba, Ibaraki

305-8566

activation inthcactivation map. One isafacilitation

mechanism that enhances activation at target

loca-tionsby applying contextual information. The other

isan inhibitorymechanism thatsuppresses

activa-tionat implicitlylearneddistractor]ocations. Inthepresent paper, we cxamined whether

contex-tual cueing involves an inhibitory mechani$m of

learned

distractor

locations.

We also rnanipulated

the stimulus- onset- asynchrony

{SOA)

between

onset of the scarch d'isplaysand that of the _probe

dots

so that we could examine the tirnecourse of

attentinnal modulation by contextual cueing,

Method

Thirty-two paid volunteers

(15

ma]es and 17

fe-males,

18

to

28

_years

of age}

_{participated,}

and

_gave

their

informed

consent.

The

display

elements were eight white squares

thatcontained a white, rotated-T or rotated-L.

The

targetwas a

T

rotated

900

totheright or tothe

left.

The

distractors

were

Ls

which were rotated OO,900,

lsoO,or

27oe,

At

the

beginning

of an experiment,

20

layouts

of

search items were generated as repeated Iayout for

each _participant. These layouts were _presented

re-peatedly

throughout

the entire experirnental

ses-sions inthe search and _probe trials. In halfof the trials,therepeated layoutswere _presented and inthe rernaining halfof the trialsrarpdom layouts,which

The Japanese Psychonomic Society

NII-Electronic Library Service

TheJapanesePsychonomic Society 106 T,Ev-E=:=ogkBRtu 6SO 6eo 550 soo

The

_Japanese

_Journal

of

Psych

(A)On-targetprobe

onemic

Science

Vol.

23,

No,

1

･eEvvs･.-aexesspt

650 600 550 soe (B)On-distractorprobe

leo 2oo 3oo loo

SOACms)

Figure 1. The rnean correct reaction times to_probes thatappeared at

distractorswith

three

SOAs, The leftpaneL

(A)

shows data

from

tria]s

inthe target location:The right _panel

(B}

shows data from trialsin the

distractor

location,

were newly generated

in

each triaLwere presented.

Each trialstarted with a white

fixation

cross ap-pearing at the center of a computer screen for500

ms,

followed

by

an array of stimuli. Inthe search

trials,the participants

had

tosearch

fer

theT target

amongst the L distractors,and to discriminate the

orientation of the target as

fast

and accurately as

possible.The target was _presentinall of thetrials.

Inthe probe trials,the array of stimuli was

accom-panied

by

a

1000-Hz

tone

of

50-msec

duration. This tone indicatedthat the_participantswcre required to

detecta sma]] _probe dot thatxv'as _presented on one of

the

search

items,

The

SOA

bctween thesearch items and the _probe dot was randornly chosen tobe 100 ms, 200 ms or 300 ms. The _participantswere asked

to_press a rcsponse key as soon as the probe was

detected

_(gofno-go

task). The probe dot was

prc-sented in two-thirds of the _probe tria]$,and

it

ap-peared equally often at the locationof a target

(tar-get-probe trial}or one of distractoritems

{djstractor-probe tria]).

The

participantswere givcn a

learning

session

in

which they performed only scarch trials,

fo]lowed

by

a testsession

in

which they performed

both

scarch

and probc trials,

Result

and

Discussion

The

results of the search

task

indicated

that

the

participantsresponded significantly fastertothe

tar-gets in the repeated layouts than to those in the

random ]ayouts. The result consequently replicated

200 300

SOA{ms)

the locationof a targetor one of

in

which the probe

dot

appeared

which the

probc

dot appeared

in

the contextual cueing effect reported by Chun and

Jiang

(1998).

Of the most irnportance was _performance in the

probe-detectiontrials,For thetarget-probe trials,the

results showed that _probe detection was gencrally

fasterin the repeated layout condition than

in

the random layoutcondition, Incontrast,

for

the

distrac-tor-probetrials,the RTs were slower intherepeated

layouts than jnthe random

layouts.

These

findings

are consi$tent with the operation of contextual

cue-ing in a facilitationof a search for learned search

target]ocations,and

in

an

inhibitjon

of the

learned

locations

of

distractors.

Moreover,

thesecontextual

modulations of probe-detection performance were

consistently observed across all of the SOA

condi-tions,and even at

100-ms

SOA.

This

suggests that

contextual

information

is

available

during

an early

time coursc of visual search _processing.

Our

study

has

clearly

demonstrated

that

contex-tualcueing

includes

an

inhibitory

proccs$, Previous

studies have reperted directevidence only fora

fa-cilitation ofa

target

Iocation

in

the

contextual cueing

effect

_(Chun

&

Jiang,

1998).Our findingof an

inhibi-torv effect on distractorlocationsshould contribute

to

an understanding of the mechanism of contextual cueing.

Reference

Chun,

M.M.

&

Jiang,

Y. 1998 Contextual cueing:

Implicit learning and memory Qf visual context

guides spatial attention.

Cognitive

Ps.vcholog:y,,36/

28-71.