Subliminal mere exposure effect and pupil responses(Summary of Awarded Presentation at the 30th Annual Meeting)

The Japanese Psychonomic Society

NII-Electronic Library Service

TheJapanesePsychonomicSociety

TheJapaneseJotarnalo.t'k.vcJionomicScienrc

20t2,

VoL

3I,

No.

],75-76

Summary

of

Awarded

PresentationIG22

Subliminal

mere

exposure

effect

and

_pupil

responses

Sanae

YOsHiMoTe*'*2,

Hisato

IMAi*3,

and

Tatsuto

TAKEucHi*

JLipan

IVbmen's

University*,

_JSTCREST*!,

Gakushuin

University*3

The

subliminal mere exposure effect

is

a

_{psychelogical}

_phenomenon

uthere

_people

tend

to

_prefer

stimuli

that

they

have

been

subliminally exposed

te

even

if

t]iey

cannot recognize

the

observed stimuli.

One

explanation

ft]r

this

effect

is

that

the

resulting

_perceptual

fluency

is

misattributed toa

feeling

ofprcference,

Thus,

an

increased

perceptu-al

fluency

should correspond

to

decreased

invested

menta] effort.

Because

the

_pupil

constricts as mentai effort

de-creases, we

_predict

that

if

_perceptual

fluency

does

induce

a

_preference

for

the exposed stimulus,

then

_participants

showing

_pupil

[onstriction

during

subliminal

exposure

will exhibit

the

mere exposure effect

later.

fo

examine our

hypothesis,

we measured

the

_pupil

diameter

while

_participants

were visually exposed

to

subliminal

stimuli,

After

ex-posure,

participants

judged

their

_preferences

to

the

stimuli.

"ie

found

that

_pupil

diameter

during

sub]iminal

expo-sure was significantly smal!er

for

_participants

who

later

exhibited

the

mere exposure effect,suggesting

that

perceptu-al

fluency

may

be

the

underl}ring mechanism ofthe subliminal mere exposure effect.

Key

words: subliminal mere exposure etfect;

_pupil

diameterl

_perceptual

fluency;

mental effort

Kunst-Wilson

and

Zajonc

(1980)

showed

that

people

tend

to

_prefer

stimuli

to

which

they

have

been

repeatedly exposed

subliminally even

if

they

cannot recognize

the

observed

stim-uli.

This

_preference

is

called

the

''subliminal

mere cxposure

ef-fect

_(SMEE)''.

One

of

the

most widely supported explanations

for

the

SMEE

is

that

_perceptual

fluen[y,

which

is

typicaLly

en-hanced

by

exposure

to

astimulus,

induces

a

feeling

of

prefer-ence.

Because

_perceptual

fluency

refers

to

the

ease

ofprocess-ing

astimulus,

it

is

reasonab]e

to

assume

that

an

increase

in

perceptual

fluency

corresponds

to

a

decrease

in

invested

men-tal

effort, which can

be

_{quantitatively}

estimated

by

measufing

autonomic

nervous system responses.

For

example,

conduct-ing

a

dirn[ult

task requiring a

rot

of mental effort causes

the

pupils

to

dilate,

whereas an easy

task

requiring

littie

effort

causes

the

_pupils

to

constrict

{Kahneman

&

Beatty

1966),

Thus,

we

hypothesize

that

if

_perceptual

fluency

is

the

underly-ing

mechanism,

then

_pupil

diarneter

should

be

constricted toa

greater

degree

for

participants

exhibiting the

SMEE.

In

thc

present

study we examine whether

_pupil

diatneter

during

sub-liminal

exposure

is

related

to

the

strength ofthe

SrvlEE,

Method

Fourteen

adults

(seven

femaies

and seven males) with

nor-mal or corrected-to-nurmal vision

_participated

in

the

study.

As

affect-neutral and novel visuaL stimuli,

two

t)rpes

of

$timu-lus

rnaterial,

Bengali

characters and

line

drawings,

were

adopt-ed.

Twenty

Bengali

characters or

line

drawings

were

divided

equal]y

into

set

A

and set

B,

Figure

1

schematical]y

illustrates

an experimental session

(Bengali

character condition).

The

experlmentaE

design

in-volved

two

_phases:

asubliminal exposure

_phase

and a

subse-quent

preference

judgment

phase,

The

two

types

ef

stimuli

(Bengali

characters and

line

drawings)

were always

_presented

in

separate sessiens.

In

the subliminal exposure

_phase,

the

fix-*Graduate

_Schoel

of

Integrated

Arts

and

Japan

NN'omen's

University,

1-1-1

Nishi-Kawasaki,

Kanagawa

214-8565,

Japan

E-mail:s.yoshimoto3141@gmaiLcom

Social

Sciences,

Ikuta,

Ta:na-ku,

copY

tw

Figure

1,

Schematic

il}ustraLion

of

the

experirnent.

ter

_presenting

a

fixation

cross

for

1

s,an

8-ins

stiinulus

wa$

_presented

followed

by,

a

492-ms

mask stimulus.

Then

participates

were asked

to

_judge

their

_preference

using a

likeabilit>r

rating and a

twe-alternative

choice

task,

The Japanese Psychonomic Society

NII-Electronic Library Service

TheJapanesePsychonomic Society

76

The

Japanese

_Journal

of

Psychonomic

Science

Vol.

31,

ation mark was a white cross

_presented

in

the

center

for

1

s.

Each

visual stimulus was

_presented

for

8

ms

fo11owed

bv

a

492-ms

random-nolse mask stimulus.

In

a

_preliminary

experi-ment, we verified

that

every

backward-masked

8-ms

stimulus

presentation

was

below

the

discrimination

threshold.

While

the

_participants

were exposed to the

stimuli,

their

pupil

diam-eters were measnred.

Halfofthe

_participants

were

initially

ex-posed

to

set

A,

while

the

ether

half

were

initially

exposed to

set

B,

In

the

_preference

_judgment

_phase

after

the

exposure

phase,

participants

were asked

te

rate

their

_preferences

to

the

stimuli using

both

a

likeability

rating and a

two-alternative

forced-choice

task

_(Fig.

1).

Participants

initially

expesed to set

A

were

tested

for

their

_preference

to

new stTmuli

("NEW")

tak-en

from

set

B

and old stimuli

("OLD")

from

set

A,

while

those

initially

expesed to set

B

were tested

for

their

_preference

te

"NEW"

frem

set

A

and "OLD"

from

set

B,

Results

Similar

to

the

findings

ef

Kunst-Wilson

and

Zajonc

(1980),

participants

in

this

study significantly

_preferred

the

OI.D

stirnuli

to

the

NEW

stimuli

in

both

the

forced-choice

task

and

likeability

rating.

Although

the

data,

which was

the

average

from

the

Bengali

characters

and

line

drawings.

confirmed

the

existence of

the

SMEE,

the

effect was not observed

for

all

par-ticipants.

About

twe-thirds

of

the

_participants

exhibited

the

SMEE,

Thus,

the

data

was

divided

into

two

sets:

the

"SMEE

observed"

_group

and

''SMEE

NOT

ebserved"

_group.

Figure

2

plots

the

normalized

_pupi]

diameter

during

the

ex-posure

phase

for

the

two

_groups

_("SMEE

observed'' and

"SMEE

NOT

observed") as a

function

of

the

number of expo-sures.

The

result

demonstrated

that

the

_pupil

diameter

gradu-allyconstricted as

the

number of exposures

increased,

indicat-ing

typical

responses

for

habituation

of autonomic responses.

AdditionallM

the

''SMEE

observed"

_group

exhibited a

larger

decrease

in

_pupil

diameter,

A

two-way

ANOV:A

indicated

that

the

main effects of

_group

(p

<.OOO1)

and number efexposures

(p<.Ol)

were significant,

but

the

interaction

of

these

factors

was

insignificant.

We

also

discoyered

that a

few

_participants

1.02-O 1.oo-eEre･--O.9Sva=cr o.se8N=. o.scroE2 o,s2

No.1

o,sc 2 4 6 e le 12 14 A6 te 2e

Number

of

exposures

Figure

2,

Normalized

pupil

diameter

during

subliminal

niere exposure as a

function

of

the

number of

sures,

Filled

[ircles

denote

the average

data

of the

"SMEE

observed''

_group,

while

bLank

squares

denote

the

average

data

of

the

"SMEE

NOT

observed"

_group.

exhibited the opposite tendencv with regard

to

the

stimulus

type.

However,

a similar retationship

between

_pupil

diameter

and

SMEE

was observed

for

each

individual

throughout

the

experiment.

Discussion

In

the

present

study;

the

_pupil

constricts rnore

for

partici-pants

who exhibit

the

SMEE

than

for

those

who

do

not,

Our

results are consistent with

the

idea

that

the

SMEE

is

induced

by

a

_perceptual

fluency,

which corresponds

to

areduced

de-gree

of mental effbrt

invested

in

_processing

the

exposed stim-uli.

In

addition,

the

_pupil

modulation

_precedes

_preference

judgment.

Thus,

comparing

_pupit

diameter

before

and

during

exposure may

be

able

te

_predict

the

occurrence of amere

ex-posure

effect,

References

Kahneman,

D.,

_&

Beatty,

_J.

_(1966).

_Pupil

_diameter

and

Ioad

on

memory.

Science,

154,

1S83-158S.

Kunst-Wilson,

MC

R.,

&

Zajenc,

R.

B,

(1980).

Affective

crimination of stimuli

that

cannet

be

recognized.

Science,