P-1B-2 文章理解におけるワーキングメモリのはたらき : 検索と保持の個人差(2001年度 日本基礎心理学会第20回大会優秀発表賞)

The Japanese Psychonomic Society

NII-Electronic Library Service

The JapanesePsychonomicSociety

7VieYkopanese

.lburnai

of1ts.,chonomicScience

ZUeZ,l'o].21,Ne.1,4!

-4Z

Prizewinner's

SummaryP-IB-21

The

role

Indivi

of

working

memory

in

text

dual

differences

in

storage

comprehension

:

and

retrieval

Yukiko

NismzAKiand

Mariko

OsAKA

OsakaUitiversitv

_ofthreign

Studies*

The

present study addressed the relationship between an individualdifferenceinthe reading

span test

(RST)

and the _perfornianceintext comprehension with respect tostorage and retrieval

systems.

In

Experiment

1a

eomparison was made of an effect of the serial recall task on

performance

in

textcomprehension on high and low RST scorc greups. Experiment 2was an

investigationof an effect of the word fiuencytask on _performance intext comprehension of two groups. The results of

both

experiments showed thattheperformance

in

textcemprehension of the

low

RST

score group was

decreased

by

the$erial recall task,

In

contrast, the _performance of the

high

RST

score _group was influenceclby theword fluencytask.The results suggested thatthe

high

RST score _group comprehended text not only

by

using a temporary storage system

but

alse

by

using a retrieval system.

Key words:working memory, reading span test,text comprehension, storage and retrieval

Individual

differences

have

been

reported

in

work-ingmemery, which

is

rneasured

by

thereading span

test

(RST),

ina temporary storage system, which is

reflected

in

thememory span task

(e.

_g.,

Daneman

&

Carpenter,1980),ancl ina retrieval system represent-ed by a word fluencytask

(Rosen

& Engle,l997).

This

study

focused

on the individual

differences

in

two aspects of text comprehension :the temporary

storage and the retrieval systems. The

dual-task

method was used and two types

(easy

and

diMcult)

of

secondary task were set.

Experiment1

Individualdifferencesin the temporary storage system duringtext comprehension of thehigh-RST

score _group

(H-group)

and the row-RST score group

(L-group)

were examined,

Methods

Participants

The experiment involved two

groups, the

H-group

and L-group,each of 15

partici-'

_Dcpariment

of

Ilsychology,

Osaka

University

of

Foreigh Studies, Amadani-Higashi, Minao, Osaka 562-8558

(e-mail

:

[email protected])

pants.

Design A 2×3 factor design, involving the 2 greups of _participants

(H-group

vs. L-group) and 3 conditions

(easy

vs. diMcult vs.control) was used.

Materials

Primary

task

(the

listening

comprehension)

The

participantswere required to

listen

tothetext

and to answer a comprehension test

{three

trials

(text$)

for

each condition), The textswere taken

from

newspapers or textbooks on

Japanese

forsenior

highschool students

(mean

number of lettersinthe

texts

(M),

356; mean

listening

time

(M),

76.91

sec-onds). For each trialthecomprehension testsscored

eight _points,

Secondary

task

(the

serial recall task)

The

diMculty

level

of thesecondary task was manipulated

by

the number of words which were _presented. For

the dithcultcondition seven words were _presentedand

for the easy condition four words were _presented

visually

<e.

_g.,ant,

flower,

_persimmon, sea).

Procedure

In

the control condition the

partici-pantswere required enly to

listen

tothetext.Inthe

dual-task

condition the _participants were ulso

required to retain the words which were presented

The Japanese Psychonomic Society

NII-Electronic Library Service

The JapanesePsychonomic Society

42 e.eoza 1.00: e.eog s.oo2 4,ooM 3.ooX 2,eo8 i.oo [.[D {e)resltEefExperiment1 The e.ooza T.OOU e,ooS 5.oo2 4.oogi ;.ooa 2.ooi1,Oo o.oe

Japanese

Journal

of Psychonomic ScienceVol.21,

:'Li'//:glg]

CONTmoL EASY DtFPCULT OOHTROL EASY DMFICULT

Figure

1.

(a)

The

means ofthecernprehension test

with theserial recall task

in

Experiment

1.

(b)

The means of the comprehension test with the

word

fluency

task inExperiment 2.

priortotheir

listening.

Results and Discussion

Results of _primary task The _performance of

participants

in

the comprehension testisshown in

Figurel(a).

The

L-group

was affected by the

dificulty

oftheserialrecall task

but

the H-group was

not. Itcan

be

suggested that

for

the

L-group

therole

of the storage system isimportant fortext

cempre-hension.

Results

of secondary task

(the

serial recall task)

The

performance

of the H-group was betterthan

thatof the

L-group

irrespective

of thetask

dithculty.

The individualdifferencesinstorage capacity

during

text comprehension

between

the two participant

groups are shown,

Experiment 2

The

individual

differences

in

theretrieval systems

from long temn memory

(LTM)

of the

H-and

L-groups were examined

during

textcomprehension.

Methods

Thirty

undergraduate students _participated

(15

subjects

in

each of the

H-and

L-groups).

The

experi-mental designand theprimary taskwere thesame as

those

used inExperiment l,

Secondary task

(the

word

fiuency

task)

The

purticipantswere required to _generatethe names for as many aspossible of the

items

which were

present-ed on a

CRT

display,

All

of the

items

belonged

toa

speeific category.

For

the easy condition a category

had a wide specification

(e.g.,

sports).

For

the

dificult

condition each category had a narrow

specification with one additional character

(e.g.,

insects

thatcan

fly).

The

diMculty

level

ofthe

secon-No.1

dary

task was manipulated

by

theperiod vf time

for

generatingthe names.

Procedures In the control condition the partici-pantswere required tolistento the text and to

simul-taneouslywrite theirname as many times as _possible,

In

the

dual-task

condition theparticipantswere asked

tolistento thetextand simultaneously towrite the generated words,

Results

and Discussion

Resultsof _primary task The _performance of the

panicipants inthe comprehension testisshown in

Figure

1

(b),

In

the

diMcult

condition the _performance

of the

H-greup

fell

toalmost thesame levelasthatof

the

L-group.

This

result suggested that the word

fluency

task

had

a _greater

infiuence

on thetext

com-prehensionof the H-group.

Results

of secondary task

(the

word

fluency

task)

The

H-group

performed

better

than the L-group

irrespective

of the

diMculty

of the word fluencytask.

Moreover,

although theprimary task performances

of thetwQ groups were not significantly

different

in

the

diMcult

condition, the secondary task

perfor-rnance of the H-group was

better

than that of the

L-group.

This indicateda differenceinthe retrieval

system of thetwe groups

during

textcumprehension.

Conclusion

The

L-group

was influenced

by

the serial recall

task. Incontrast,

the

H-group was influencedby the

word

fiuency

task, which needed access to LTM

cluring

textcomprehension,

The

textcomprehension

processof theL-group was

found

to

depend

on

tempo-rary storage and theirretrieval system

from

LTM

did

not work as effectively as that of the H-group,

The

H-group however, could use aretrieval system as well as a temporary system.

References

Daneman,

M.,

&

Carpenter,

P.A. 1980 Individual

differences

in

working memory and reading.

nal

of

VigrkalLearning a"d V2Mbal

Behavior,

19,

450-466.

Rosen,

V.

M.,

&

Engle,R.VL'.1997The role of

ingmemory capacity

in

retrieval.

foumaag

of

Empe"