(PDF) Social Intelligence--A Review and Critical Discussion of Measurement Concepts.

Emotiona I

I

ntel

Iigence

An

lnternational

Handbook

Edited by

Ralf Schulze

Richard

D.

Roberts

HOGREFE

Library

of Congress

Cataloging-in-Publication

Data

is available

via

the

Library

of Congress

Marc

Database

under

the

LC

Control Number

.2004114391

Library and

more on understanding the irri-

portance of certain social

milieu. Moreover, as

Orlik

(7978)

points out, several

validation

studies show

that variance in performance data may be

explained,

to a large extent, by verbal measures of

academic

intelligence. \Atrhereas

per-

formance in the

GWSIT

has been

shown

to correlate up

to

.70 with

academic

(verbal)

intelligence, correlations with other social intelligence indicators

show

no evidence for convergent validity.

A

second

major approach to the assessment

of socia-l intelligence was in-

troduced by Guilford and colleagues/ under the framework

provided

by the

previously elucidated Structural

Model of Human Intellect

(Guilford,

7967).

The following list briefly describes some examples of tasks

out

of the Sü and

the Four Factor Test

of Social

lntelligence

(O'Sullivan

& Guilford, 7966,1976)

and their classification in that model:

Expression

Grouping

(Classes):

Participants find

one

facial

expression, out of

four alternatives, which best fit a group of three

other

facial

expressions.

Missing Pictures

(Systems):

Participants are

presented with a sequence of

events,

pictured

in

photographs,

and have to

complete the sequence by

choosing the correct

last

photograph.

Missing Cartoons

(Systems):

Participants

are

required to fill-in a blank, in a

sequence

of

cartoons,

by

selecting

the

correct cartoon out of four choice

alternatives.

Picture Exchange

(Transformations):

A

sequence of photographs is

presented

that tells a

story.

Participants are required to replace

one

marked

photo-

graph of

this

sequence,

with

one of four

alternatives,

in

order

to

give

the

story a different meaning.

208

Social lntelligence

Social

Tianslations

(Transformations):

This

test

is

the

only

verbal

social intelligence

in

this

battery.

Participants

ut"

drr"^

u.

o

toon series.

As O'sullivan

et

al.

(1965)

recognized,

these

tests

were

measuröi

nitive rather

than

behavioral

skills.

The authors

themselves

reporterJ-;

stantial

correlations

with

general

intellectual

abilities

(O'sullivan

&

1966).

More

recent

studies

have

focused on

the

construct

validity

of

batteries.

For

example,

Probst

(1982)

applied the

Six

Factor

tesi

in

sive

study

of social

intelligence,

finding

empirical

support

for

an

ind

ability

construct.

However,

factor analysis

did

not

yield

a commonj

telligence

factor

comprising

different

types

of assessment

methods.

Ir

study,

Riggio,

Messamer,

and

Throckmorton

(1991)

neither

found

evii

convergent

nor

for

discriminant

construct

validity.

They

applied

thö'i

tor

Test of Social

Intelligence,

along

with a measure

of

academic

ihi

that is,

the

Wechsler

Adult

Intelligence

Scale-Revised

Edition

(W

cabulary

Subscale

(Wechsler,

1981). Furthermore,

the

Social

skills In

(SSI;

Riggio

,1989)

was

administered

as a measure

of

self-reported soü

ln an exploratory

factor

analysis,

the

subscales

of

the

Four Factor Te

on

one

fäctor

wlih the

WAIS-R,

showing

near to

zero

correlations wii

Thus,

neither

convergent

nor

discrirninant

construct

validity were

in this

investigation.

1:0.3.2

Individual

Tests

of

Social

Intelligence

Empirical

approaches

that occurred

after

Walker

and Foley's

(797,3)

tik's

(1978)

summarizing

works, somewhat

surprisingly,

appear

lesl

guided

than

those

discussed

in the

aforementioned

passages.

More

üut seemingly

like social

skills,

nonverbal

decodin-g

nonverbal

comrnunication

skills

have subsequently

been

operationalii

dicators of

social

intelligence

(Barnes

& Sternberg,7989; Feldman,

Tol

Coats,

1999;

Riggio

,7986;

Sternberg

& Smith,

1985).

With

every

cons

of

the contributive

value of

these

investigations,

it

seems

somehow

ment made

between

a pair of people,

in a

defined

Jocial

rela

have

to choose

one pair

of people out of

three

altematives,

for:w

given

statement

has a

different meaning.

(i.e.,

performance

vs" self-report

data).

tn the

following

subsections,

vide a critical

analysis

of

tests represented

by this

classification

Cartoon

prediction

(Implications):

participants

are

required

to

seleciu

toon, out

bf

three

alternatives,

that most appropriately

completr

cult

to

find appropriate

indicators

of

these concepts, as

task

requirer

more

explicit

and

less complex.

It

appears

necessary

to classify

these

t

.o*

upprouches

in order

to facilitate

the

interpretation of

research

r'

this

puipose,

operationalizations

can be

cross-classified

along

two

dj

One

is defined

by

the content

under examination

(i.e., cognitive

vs'

social

skills),

while

the other

dimension

describes

the

method

of

THE

ASSESSMENT

OF

SOCIAL

INTELLIGENCE

that the

treatment

of self-report

measures

wiil

include

attempts

to assess

both

the

cognitive

and

behavioral

dimensions

of social

intelligence.

Cognitiae

performance

measures.

Keating

(1978)

employed

three

verbal

in-

dicators of social

iltelligence performance

designed

previously

(e.g.,

Chapin

Social

Insight

Test,

Chapin,7967;

Gough,

1968), as well as three measures

of

academic intelligence

(both

verbal

and

nonverbal

material). Neither correla-

tional

nor

factor analytic results supported

construct validity. Within-domain

coruelations

did not

exceed

across-domain correlations, and no

coherent fac-

tor

structure was

observed.

Furthermore,

the

social intelligence performance

measures

did

not predict

effective social functioning

(assessed

by

peer-reports)

to a larger extent than academic intelligence. Sternberg and colleagues

(Barnes

& Sternberg,7989; Sternberg

& Smith, 1,985) operationalized the

concept of

nonverbal decoding skills

as an indicator

of social intelligence. Th"y devel-

oped

two

tasks

relying

on similar principles.

One was the

so-called

"Couples"

Test, which

contained

photographs of

heterosexual

couples

that were

either in

a close relationship

or

were strangers. Participants had to

judge

each

photb-

graph

for

the kind of

relation depicted

(i.e.,

close

relationship or strangers).

The

second

task

consisted

of

photographs

of a

supervisor

and his

or

her

su-

pervisee. In this instance, participants

had to

judge

who the supervisor was.

Barnes and

Sternberg

(1989)

used

self-report

inventories

of social

competence,

as well as

performance

measures

of

academic intelligence,

to

ascertain

con-

struct

validity.

Correlational

analyses showed an unequivocal

pattern with

only

significant

convergent and non-significant discriminant validity

coeffi-

cients.

Along

with studies

by Riggio

et

al.

(1991)

as well as

O'Sullivan

and

GuilJord

(1966),

these

results

alone only allow ambiguous conclusions about the valid-

ity of

social

intelligence based on performance

measurement.

At first blush, it

seems

that applying verbal

performance

measures results in

substantial over-

lap between

social

intelligence and

academic

(especially

verbal)

abilities

(Keat-

tng, 1978). Thus, investigations using nonverbal measures as indicators

of so-

cially intelligent

performance

succeed somewhat

better in identi{ying

a con-

ceptually

coherent

domain of social intelligence

(O'Sullivan

&

Guilford,

1966;

Barnes

& Sternberg,7989). However,

this result

is

not

always demonstrated

(Sternberg

& Smith, 1985;

Riggio

et

al., 1991).

The difficulties

of

both verbal and nonverbal cognitive

performance mea-

sures in defining an unequivocal social

intelligence

construct could be attri-

buted to a methodological

problem.

According to

Schneider,

Ackerman, and

Kanfer

(1996),

certain

characteristics of social cognitive tasks increase the

over-

lap with

academic intellectual abilities

by matching

their typical measurement

features.

These characteristics

include: when

participants encounter

social

stimuli that are inconsistent with their expectancies, when

participants

are

faced with novel stimuli, and

when

participants

are faced with highly

struc-

tured

tasks

(Schneider

et

aI.,7996,

p.

a69). Among

cognitive performance mea-

sures

of social

intelligence, above all, those relying on verbal material

seem

to

meet

all three criteria.

The

sequential

type

of presentation inherent to writ-

209

210 SociallntelliSence

tenlanguagedoesnotSeemtobeanadequateoperationalizationofmoreor

less

coäplä* social

stimuli'

Instead,

writfen

language

liPeals

to be

distinct

from

socially

relevant

stimuii

found

in

real-life

seitings.

Plausibly,

this

type of

presentation

conJronts

the

participants

with novel

(or thus

far

unexperienced)

ätimull

that implicate

cognitive

functions

that parallel

those

necessary

for the

accomplishme.tt

of

academic

intelligence

tests'

B'ehaaioral

performance

measures'

Ford

and

Tisak

(1983)

app-lied.a.perfor-

mance

measure

of

socially

intelligent

behavior

as

an

indicator

of

social

intelli-

sence.

Participants'behavior

in in

interview

situation

was

rated

along

certain

E

ir"Ctl.g.,

tie

ability

to speak

effectively,

to be

appropriately

responsive

to

the

interview"r',

qn"riio.rs,

to

displuy

appropriate

nonverbal,behaviors)'

Ad-

Jitionally,

the

authärs

assessed

s"ü-

ut abiftei-reported

socialüehavioral

skills

and

academic

inteliigence.

Correlational,

as

well

as

factor

analytic,

results

sug-

gested

a

distinct

,oJiul

ittt"lligence

construct.

Within-domain

correlations

ex-

ceeded

across-domain

correläons,

and

social

intelligence

measures

loaded

on

a separate,

interPretable

factor'

A

comparable

study

was

conducted

by

Frederiksen,

Carlson,

and

Ward

tfq84l

Again,

perforÄut

ce

in

an

interview

setting

lerv-ed

as

an indicator

of

,o.iui

it

titig"n"".

Participants

had

to

take

the

role of

a doctor

who

w.as,inter-

viewing

his/her

patient.

Additlonaily,

Frederiksen

et

al.

(1984)

applied

vari-

ous

measures

of

academic

intelligence

and

problem-solving

abilities'

Results

st

o*"d

only

a

few

substantiai

ärrelationJ

between

interview

performance

and

academic

intelligence

measures.

These

correlations

were

partly

negative

in

sign,

suggesting

ttat

high

acadernic

intelligence

was

accompanied

by

low

social

behavioral

skills.

Finally,

stricker

and

Rock

(1990) applied

a

techniqrre

sirnilar

to

the

inter-

view

settings

d.escribed

thus

fär. Striätcer

developed

his

own

measure

of so-

.iuffy

it

t"ffigent

behavior,

the

Interpersonal

Competence

Inventory

(ICI)'

The

ICI

was

based

on

video

scenes

"ot

tuit

ing

an interview

situation

between

a

subordinate

and

his

superior.

In

the

Replies

section,

participants

had

to

re-

spo.,a

orally

to

the subärdinate

in

place'bf

a superior.

Answers

were

judged

i^

t"r^,

of

äffectiveness

and

originätity.

In

the

|udgment

section,

participants

had

to

write

down

their

descripüon

of

ihe situation

and

its important

features'

The

performance

criterion

was

accuracy.

Conceptually,

the

Replies

section

op-

erationalized

socially

intelligent

behavior,

whereas

the

]udgment

section

as-

sessed,

for

the

mosi

part,

cögnitive

skills.

Along

with

the-ICI,

Stricker

and

Rock

(1990)

ussess"d'nor,-rr"öd

social

skills,

acaäemic

intelligence,

and.self-

,"po.t"d

social

intelligence.

Results

from

correlational

and

multidimensional

r.är*g

analyses

show"ed

no

coherent

structure

either

within

the

domain

of

so-

cial

in"teliigence

or

concerning

the

relation

of

social

intelLigence

measules

to

academic

ilt"llig"r,.".

Social

ät"llig"t.u

performance

measures

correlated

in-

consistently

witl

each

other

(betwe€il.

/

:

-.08

and

.37)

and

the

judgment

section

of

the

ICi

correlated

substantially

with

the

verbal

measule

of

academic

intelligence

(r

:

.30).

THE

ASSESSMENT

OF SOCIAL

INTELLIGENCE 211

The interpretation

of results in

the aforementioned

studies

allows us to

draw

some

tentative

conclusions

concernjng

the validity

of

social

intelligence.

Both

Ford and

Tisak

(1983)

as well

as

Frederiksen et al.

(1984)

succeeded

in

sepa-

rating

social

from

academic

intelligence

thus

proving discriminant

construct

validity.

However,

Frederiksen

et al.'s

(1984)

findings-partly

negative

corre-

lations

with

academic

performance-raise doubts about

the nature of the

per-

formance construct of social intelligence. It should be

expected that a

so-called

"intelligence"

construct would

at least be slightly

positively correlated with

traditional measures

of academic intelligence. Furthermore,

a

strict

account

would

note that the

generalization of

the findings

is restricted to

a rather

spe-

cific

(albeit

practically

meaningful)

instantiation

of

the

social context in

which

humans interact:

interview settings. Finally,

it must be

stated

that

the conver-

gent construct validity was not

convincingly

proven

in

these

studies,

neither

for the restricted interview

settings

as indicators

of social inteiligence

nor for

a

possibly more

general social

intelligence

construct.

Self-reported

social intelligence. Numerous

studies have

applied self-repört

inventories

as

measrues of social intelligence. In

several

of

these investiga-

tions

self-reported

social skills serve as

psychologically meaningful validation

criteria

(Barnes

&

Sternberg,I9B9;

Ford

&

Tisak,

1983; Frederiksen

et a1.,79,84;

Riggo

et al., 1991). However,

there are a large number

of

studies

that

rely

only

on self-reported social

skills

as indicators

of

social

intelligence

(Brown

&

Anthony, 1990; Marlow

e, 7986; Riggio, 1986).

We

have already described

Riggio et al.'s

(1991)

study in the context

of

the

O'Sullivan and Guilford

(1976)

test battery. ln

this investigation,

the subscales

of the Four Factor

Test of Social Intelligence loaded

on one factor together

with

academic

intellectual abilities, whereas

the subscales

of the Social

Skills

In-

ventory

(SSI;

Riggio, 1.989) loaded

on

a

separate factor.

Self-reported social

skills and

performance measures of social intelligence

did not

correlate sub-

stantially and

only one correlation

(viz.,

the

Social Translation

Subtest with

the

SSi) reached

significance.

FIowever,

other

studies employing both

self-

reported

social skills and social intelligence

performance tests report

evidence

of convergent validity

(Barnes

& Sternberg,l9B9;

Ford & Tisak, 1983).

Riggio

(7986)

validated the

SSI

using the

traditional

personality

scales of

the

16 Personality

Factor

Questionnaire

(16PF).

The

SSI contained

six sub-

facets

that resulted from a

cross-classification

of contents

(viz.,

social

vs.

emo-

tional

contents) and

postulated ski11s

(viz.,

sensitivity,

expressivity, and

con-

trol).

Sumrnarizing

these results, the

SSI subfacets correlated

substaltially

with various

personality

traits

(e.g.,

social expressivity:

outgoing,

happy-go-

lucky, venturesome,

group dependent;

social sensitivity:

affected by feelings,

shy,

astute,

apprehensive,

conservative, tense, undisciplined).

Moreover,

par-

ticipants

scoring high

on

the different

SSI

subfacets could be described

by a dif-

fering

personality

structure.

According

to Riggio

(7986),

these results

proved

the convergent validity

of SSI as a measure

of

nonverbal

social skills. His con-

clusion

was also

supported

by further validity

evidence: high

scorers on the

SSI tended

to report more

socially effective

behavior

and

richer

social contacts

212 Sociallntelligence

(Riggio,

1986).

The

aforementioned

findings

of

Riggio

et

al.

(1991)

put

some

ambiguity

into

this

interpretation.

In terms

of Schneider

et

al.'s

(1996)

crit-

icismi, the

subscales

of

the

Four

Factor Test

of

Social

Intelligence embodied

operationalizations

that

were

conceptually

too close

to

academic

intelligence

performance

measures.

Thus,

they represent

no valid operationalization

of the

iocial

intelligence

ability

construct.

From

another

viewpoint,

it is also possible

that

these

results

could

be

attributed

to

common

method-related

variance

in

'self-report

and perfotmance

data.

To examile

these propositions

more closely,

two other

studies

are worth

noting.

Marlowe

(1986)

operationalized

social

intelligence

via a

self-report

in-

strument.

He intended

to demonstrate

that social

intelligence

would show

independence

from academic

intelligence.

Secondly,

Marlowe postulated

the

multidimensionality

of

social

intelligence.

He extracted

four dimensions

from

the empirical

literature.

Along

these

dimensions,

social

intelligence

includes

social

interest, social self-efficacy,

empathy skills,

and

social performance

skills.

Factor

analytic results of

the social

intelligence

measures yielded

five separpte

factors

labeled pro-social

attitudes,

social skills,

empathy

skills, emotionality,

and

social

anxiety.

The

postulated

dimensions

could

thus not be instantiated,

though

there

was clear evidence

for

the multidimensionality

of social intel-

ligence. Correlational

analyses

suggested

construct

independence,

showing

near

to zero correlations

with

acadernic

intellectual

abilities assessed

by

per-

formance

data. Anyway,

evidence

for the convergent

constluct

validity

was

again

missing.

Subsequently,

Brown &

Anthony

(1990)

found similar

results.

They as-

sessed

self-

and peer

ratings

of

both social

behavior

and

personality

traits,

along

with general

intellectual

performance.

A factor

analysis

resulted in

a

clearly

defined

factor structure.

The

three factors

were identified

as:

(a)

acad-

emic

intelligence,

(b)

peer

ratings

of

both social

behavior

and

personality,

and

(c)

self-reported

social

behavior

and personality.

However, it seems plausi-

ble

that these

findings

point

to meaningful

method-related

variance,

which is

idrerent

to diJferent measurement

approaches.

10.3.3 Recent

MTMM Studies

Most of

the

aforementioned

approaches

did

not clari{y

the role

of

the intended

measurement

constructs

in a putative

higher-order

framework of social

intel-

ligence.

During

the

past

decade,

however,

attempts have been

made to

ap-

ply

multitrait-multimethod

(MTMM) designs for a better

understanding of

the

structure

and construct

validity of social

intelligence

(Jones

&

Day, 1997;Lee,

Day, Meara, &

Maxwell,

2002;

Lee, Wong,

Day, Maxwell, &

Thorpe,

2000;

Wong,

Day,

Maxwell,

&

Meara,7995).

All these investigations

have assessed

verbal

and nonverbal performance

measures,

as

well

as self-

and sometimes

other-report

data,

of

the

respective

trait-facets.

Furthermore, the

use

of

con-

firmatory

factor

analysis in

these studies

allowed the separation of

trait-

and

method-related

variance to

derive an empirically

defensible

structural

model

of

social

intelligence.

THE ASSESSMENT

OF

SOCIAL

INTELLIGENCE

In

the

first

of

these studies,

Wong

et al.

(1995,

Study

1)

set out to

measure

academic

intelligence,

social perception as

a

cognitive facet

of social

intelli-

gence/

and

socially intelligent behavior

(operationalized

as effective

hetero-

sexual

interaction).

The

latter

include.d ratings

of both

verbal

and

nonverbal

behavior

in

a first

encounter between

a male

and a female

(recorded

on

video-

tape).

verbal

social

perception

was operationalized

by

a

subtest of the

George

Washington

Social

Intelligence Test

(i.e.,

recognition

of the mental

state

behind

words,

see above).

The

Expression

Grouping

subtest of the

Four Factor

Test

of social

Intelligence (o'sullivan

&

Guilford, 1976)

was also

used

as a

mea-

sure

of nonverbal

social

perception.

Results

yielded a

model with

four

uncor-

(viz.,

verbal,

nonverbal,

self-report,

and

other-report)

and three

correlated

trait-factors

(viz.,

academic

intelligence,

social

perception,

and

effective

heterosexual

interaction).

However,

both

zero-order

correlätions

as well

as trait-factor

intercorrelations

pointed to

substantial

overlap between

social

perception

and academic

intellectual

abilities

(r

:

.67),

a value that

ex-

ceeded

the intercorrelation

between

social

perception and

effective

heterosex-

ual

interaction

(r

:

.54).

I

In

the

second

of these

studies, Wong

et al.

(1995,

Study

2)

postulated three

facets

of social

perception,

social insight,

and

social knowledge.

In the

verbal

measures

of social

knowledge,

participants

had to identify

the

best

solution

for

a

social

problem. The

nonverbal

measure

demanded

the identification

of

etiquette mistakes,

pictured in drawings.

verbal

social

perception was

opera-

tionalized

by the

Social Translation

Test

of the Four

Factor

Test

of Social lntelli-

gence

(o'sullivan

&

Guilford, 7976),

while

the nonverbal

measure

of this facet

was

again

the Expression

Grouping

subtest

of the

same

test

battery.

The verbal

measure

of social insight

was the

Judgment

in

Social

Situations

subtest

of

the

GWSir

(Moss

et al., 1955).

The

nonverbal

measure was

the

Cartoon

prediction

subtest

of o'Sullivan

and

Guilford

(1976).

The

authors

successfully

identified

the

cognitive facets

of social insight

and

social knowledge

as trait-factors

sep-

arable

from,

but

positively

related to,

academic

intelligence.

social

perception

could not

be

separated

from

social insight.

In

yet

another

study,

Jones

and

Day

(7997)

applied

Cattell's distinction

of

fluid versus

crystallized

intelligence

on the

social intelligence

construct ald

thus

operationalized

verbal

and

nonverbal

social

cognitive

flexibility

(fluid

intelLigence)

and verbal

and

nonverbal

social knowledge

(crystallized

intel-

ligence).

The nonverbal

measure

of

social cognitive

flexibility

contained

short

video

clips of ambiguous

social situations.

Participants

had

to list all

possible

interpretations

of

each scene.

The verbal

task

of this facet

included

written

descriptions

of

ambiguous

social situations.

Participants

had again

to list

all

possible interpretations.

The Expression

Grouping

subtest

of o'sullivan and

Guilford

(1976)

represented

the

nonverbal

measure

of social

knowledge.

The

social rranslation

subtest

(o'sullivan

& Guilford,

r976)

was

used as the verbal

measure

of social

knowledge.

|ones

and Day

(7997)

could

show

a

trait-factor

of social cognitive

flexibility

again

separable

from, but

positively related

to,

academic

problem

solving, whereas

social knowledge

could not

be

separated

from

academic

problem

solving.

21,4

Sociallntelligence

Extending

on these findings, Lee

et al.

(2000)

operationalized

both

fluid

and crystallized

social and'academic intelligence.

The authors

specified

fluid

and

crystallized

social intelligence

as

social inJerence

and

social knowledge,

respectively.

Results

showed

that all four

postulated trait-factors

were

dis-

criminable from each

other. Lee

et ai.

(2002)

diverged

from the

just

described

approaches

by using tasks

with

open-ended

questions

to

operationalize

social

knowledge

and the flexible

application

of it. Thus,

they

rather

represented

the

ideas

of Cantor and Kihlstrorn (7987),

who claimed

that

open-ended

questions

would

be

more indicative

of real-life

social

problems

than tasks

with

just

one

correct answer.

The verbal

measure

of social knowledge

was the Role

iategory

Questionnaire

(see

Lee

et

aL.,2002).

Participants

had

to write

detailed

desirip-

tions

of

persons fitting into

a

certain kind

of

social role

(e.g.,

liked

rurn"

,ä"

friend).

ln the nonverbal

measure

of social knowledge,

participants

had to

de-

scribe as fully

as they

could well-known

target

persons

(e.g.,

oprah winfrey),

whose

photos were

presented

on

screens. Answers

were

scored in terms

of

the number

of different

personality

and behavioral

characteristics

identified.

The

verbal

and nonverbal

measure

of social

cognitive

flexibility

representdd

the

same as applied

in the

study

of

Jones

and

Day

(1997).

Results

of this

study

showed

separable

social intelligence

trait-factors

distinct

from,

but

positively

correlated

with,

(general)

creativity.

ln

summary, these

MTMM-studies provide

clear

evidence

for the

murti-

dimensionality

of

social intelligence.

Ho*ever,

although

the

method-related

variance

of

self- and

other-report data

was

controlled

by

the introduction

of

method-factors

or correlations

among

the respective

measures,

trait-factor

Ioadings

vary

strikingly

between

performance

measures

and

self- and

other-

report

data.

Moreover,

the different

measurement

procedures

exhibit no

co-

herent

loading

pattern on one trait-factor.

Consequently,

it remains

uncertain

what

influence

the inclusion

of self- and

other-report

data has

on the

identi-

fied

trait-structure.

Particularly,

no further

(convergent)

validity

evidence

was

available

since self-report

data were

already

included

in

the

social intelligence

models.

1.0.3.4

Summary

In

spite

of the

early,

extensive work

of Guilford

and

his

colleagues

on

so-

cial intelligence

and

their attempts

to

establish

a theoretical

framework,

not

many

comparable

systematic approaches

may be

found

in the

literature.

Most

empirical

studies

focus

on

a

single,

very

specificcognitiae

aspect

of social in-

telligence.

These

operational definitions

seldom

clarify the

role

of measure-

ment

constructs

within the

context

of a higher-order

framework.

hr

addition,

MTMM

approaches

do not

conceptually

integrate

lower-order

facets

of

social

intelligence

(and

their

concomitant

cognitive

determinants)

into

a comprehen-

sive model

of social intelligence.

Since

there is

clear

evidence

for

the

multi-

dimensionality

of social intelligence (Lee

et aL.,2000,2002;

wong

et al. 1995),

it

seems

important

for

future

studies

to locate

constructs

within

a

coherent,

taxonomic

model

of

social intelligence.

The

same kind

of critique

might

be

FACETS

OF SOCIAL INTELLIGENCE 21.5

addressed

to

approaches

focusing on

the

measurement

of social

behaaioral

ef-

fectiueness.

Neither

the role of

effective social

behavior in a framework

of

social

intelligence

nor

an

internal

classification of relevant

social settings appears

to

underlie

extant approaches.

Empirical

evidence for

the construct validity

of social intelligence varies

strikingly across

the measurement procedures that have been adopted.

For

example,

self-report inventories

and behavioral

effectiveness

criteria

suggest

a distinct domain

of social

intelligence.

Approaches relying

on

verbal

(and

sometimes also

nonverbal)

tasks

fail to provide incontrovertible

evidence of a

discriminable

performance construct. However,

the problems associated with

the various

types

of

measurement

procedures remajns an

empirical

issue.

It is

relatively

self-evident that self-report data

better serve as measures

of

typical

social

intellectual

performance in comparison to measures

of

a

performance

construct that is

based

on

the idea

of

maximal

performance. Approaches re-

lying

on pure

performance

measurement

should carefully

consider

the nature

of the task material,

both with respect to the

selection of convergent validity

criteria

and

to real-life

congruence

(Schneider

et a1.,7996).

10.4 FACETS OF SOCIAL INTELLIGENCE

ln this section, we will

attempt

to

integrate

past

theoretical and

empirical

work

into a

performance model of

social

intelligence.

This model is based

on

the

idea

of

a faceted

intelligence model as a framework

for the description and

classification

of

variables

or

tests

(Süß

& Beauducel, 2005), the main focus

of

which will

be various cognitiae facets. It

does

not

lay claim

on

completeness

or

conclusiveness

and will need to be

supported

by

empirical data. In any

event, considering the diversity

of

past

empirical approaches,

it seems

neces-

sary

to

classify the theoretical and

operational definitions

of social

intelligence

in

a unified framework.

The model is based on five

cognitive facets: social ul-

derstanding,

social memory,

social perception, social creativity

(or

flexibility),

and

social knowledge. After a description

of this model, we will

provide some

preliminary results

of

a

study

based

on

this

performance model.

10.4.1 A Taxonomy

of Cognitive Facets of

Social

Intelligence

The facet

of

social understanding

(or

insight) was included

in a large number of

theoretical

and operational definitions,

given different labels

but comprising

similar requirements.

It can be

perceived

as the

pivotal facet

of social

intelli-

gence

in

past

investigations.

Thus,

several

definitions

of social intelligence re-

ferred to in

Section 70.2, narnely, the

ability to understand

people

(Thorndike,

1920), the ability

to define a

given situation in terms

of

the behavior imputed

to others

present

(Chapin,

1942), and to

judge

correctly the feelings, moods,

and motivations

of individuals

(Wedeck,

7947)

cowld

all

be subsumed

under

the facet

of social understanding

or social

insight.

Additionally, both broad

and

specific

operationalizations

of socially

inteliigent

cognition may be classi-

21.6

Sociailntelligence

fied

under this facet:

the

GWSIT

(Moss et

al., 1955),

the Chapin

Social

lnsight

Test

(Chapin,1967;

Gough,

7968),

the

broad

test batteries

of

O'Sullivan

and

Guilford

itOOO,7976),

no--nverbal

decoding

skills

(Barnes

& Sternberg,

T9B9),

and so

forth. Social

understanding

abilities

thus

require

individuals

to inter-

pret

or

understand

given

social

stimuli,

which

may

vary

according

to their

iomplexity,

in termsäf

the

implications

for

the situation

and

their

underlying

features.

the

point

is

well

ifustrated

by

a

sample

test requirement:

ulder-

ständ

correctty

wtiat

a

person

wants

to express

via verbal

or

nonverbal

means

of

communication.

The

facet

of

social

memory

(i.e., behavioral

memory)

was included

in

the

structural

Model

of

Human

Intellect

(Guitford,

1967).

Kosmitzki

and

]ohn

(1993)

also

discovered

a social

memoly

factor

in laypersons'

implic-it

theories

aboui social

intelligence,

that

is,

memory

for

names

and

!,'9s.

One

docu-

mented.

operational*ization

is

that provided

by

Mosset

al.

(1955) in

the GWSIT

(see

also

Probst,

7982).

They

operationalized

social

memoly

as

memory

for

,ru^"s,

and

faces.

The

facet

of social

memory

requires

the

intentional

storing

and

recall of

both

episodic

and semantic

memory

contents.

Correspondingly,

social

memo.y

p"tiot*ance

is

determined

by

the conscious

recall

of objec-

tively

and. expliiitly

given

social

injormation

that can

vary

along

a continuum

of

complexity.

So

far,

the facet

of

social

perception

has

not

been

reflected

in

theoretical

ac-

counts

of social

intelligence.

Nevertheless,

according

to

our

view,

the

ability

to perceive

socially

relävant

information

should

play

a role

in

a performance

modet

of

social

intäligence.

The

ability

to

(quickly)

perceive

social

information

in

a

given

situation

"o=.tta

deterrnine

further

inJormation

processing

that

isrel-

evan"t

for the exhibition

of socially

intelligent

behavior.

Only

Wong

et

al'

(1995)

attempted

to

operationalize

sociäl

perception.

However,

they

did

not succeed

in sepärating

social

perception

from social

understanding

abilities.

These

re-

sults

could

üe attriblted

io

the requirements

of

the selected

tasks.

The

tasks

also included

interpretational

demänds

that,

in our

view,

cannot

be

subsumed

under

the

facet of

pure

perceptual

abilities.

To meet

these conceptual

require-

ments,

social

perception

can6e

specified

as social

perceptual

speed,

analogous

to the

idea oi

peräptual

speed

in models

of

academic

intelligence

(Carroll,

1993;

Thurstone,

1938).

Social

creatiaity

(or flexibil-ity)

was

conceptualized

in

Guilford's

Structural

Model of

Human

Intellect

(Guilford,

7967)

as

divergent production

of

behav-

ioral

contents.

Recent

empirical

work

(Jones

& Day,

7997;

Lee

et

al.,

2002)

operationalizes

social

cognitive

flexibility

as the

flu9nt production

of possi-

bie

interpretations

of,

orlolutions

for,

a given

social

situation.

Importantly,

participäts'performalce

is

not

based

on one

corlect

answer

but on

the num-

L"t u.,ä

diveisity

of

ideas.

The

measures

used

by

both

Jones

and

Day

(7997)

and

Lee et

al.

(Zö02) to define

this

construct

were partly

in line

with Guilford's

early propositions.

Note

that

these

authors

were

capable of

successfully

distin-

guishing^the

domain

of

social

cognitive

flexibility

from

academic

intellectual

abilities.

l'

FACETS

OF SOCIAL

INTELLIGENCE

217

Social

knowledgehasbeen

given

credence

in

the

definitions of

Vernon

(1933,

viz.,

knowledge

of social

matters)

and

Cantor and Kihlstrorn

(1987,

viz., in-

dividual's

fund

of knowledge

about the

social

world). The

concept of social

knowledge

also

plays

a

substantial

role in recent

conceptualizations of practi-

cal

intelligence and

the

related concept of wisdom

(Baltes,

Staudinger, Maer-

cker,

& Smith, 1995; Staudinger,

Lopez, & Baltes, 1997;Sternberg,7998;

Stern-

berg et al.,

2000).

So

far,

social

knowledge has mostly been

operationalized by

measures relying on

knowledge of good etiquette

(Lee

et a1., 2000; Wong

et

al.

1995). Contrary

to these

operationalizations,

Kjhlstrom and

Cantor

(2000)

differentiated

between procedural

(or

so-called

tacit knowledge)

and declara-

tive

social

knowledge.

They

postulated

that

procedural

knowledge could

not

be

taught

or

recalled explicitly,

in

contrast

to declarative

knowledge and the

corresponding

memory components

of episodic and

semantic

memory. With

respect

to these

considerations,

social

knowledge

can be specified as contents

stored in the

procedural

memory component that cannot be taught

or

recalled

explicitly.

6

Given

these constraints, social

knowledge becomes

conceptually

distinct

from

social

memory. However, social

knowledge, as

specified in

these

consid-

erations,

is dependent on

the influence of the cultural

environment

in

general

or

the

specialty of

the

situation

(Weber

& Westmeyer,2001). The assessment of

social

knowledge,

thus,

would require a comprehensive

classification of pos-

sible social situations. Still,

any assessment

would be

subject to respective bul-

tural values and

standards.

From ttris description,

it is not

possible

to conclude how these facets inter-

act to enable people

to exhibit socially intelligent behavior, in

general.

These

cognitive

determinants need

not necessarily

stand on one and

the

same level

and, thus, contribute to

higher-order

performance

to the

same extent.

Figure

10.2 portrays

the

proposed

model of social intelligence including the cognitive

facets, their

possible

interactions

with

each other,

and

with social

behavior as

the

outcome of social

cognitive

intelligence.

IrL this illustration, the

just

described facets constitute the

social

(cognitive)

intelligence construct.

F{owever, the facet of social knowledge, as depicted in

Figure 10.2, does not

play

the

same role

as the other

four facets.

It

seems rea-

sonable to assume

that

social

knowledge

(as

a kind

of

meta-concept) might

also irL{luence the

performance,

for example, in social understanding

or social

perception

abilities. Furthermore,

it is

questionable

whether

social knowledge

only

contains cognitive

requirements, following the aforementioned

consider-

ations

of

Weber and Westmeyer

(2001).

Altogether, the

social cognitive facets

surely

determine social

behavior

performance

to an important

degree. How-

ever,

the

extent of this determination

and, hence, the final

exhibition of socially

intelligent behavior is also influenced by some other, at this

point

indetermi-

nate, array

of person

and environmental

variables

(i.e.,

situational demands,

moods, personality, aims, etc.),

as indicated in the figure.

21,8

Sociallntelligence

o

Situationai

demands

o

Values

/

Norms

o

Personality

o

Moods

r

fnterests

o

Aims

o

Experience

*fillill;t.,

A

possible

performance

moder

of

social

intelligence,

incruding

five

cog-

70.4.2

A

Preliminary

Test

of

the

Model

The

focus

of

the

p-reselt.rlvestigation

(weis

&

snß,2004)was

to

assess

three

cognitive

facets

of

social

intellilence:

social

"r,a"r.tu.,äir,g,'socrat

memory,

and

social

knowredge,

based

ori

performance

measures.

Tö

contror

possibre

effects

of

task

mater-ial,

we

used.

ir"Jui

iurt

",

pictures,

and

videos.

The

ver_

bal

measures

of

social

äg::rtqdTg

yere.t1"

Chapin

Social

L_rsight

Test (SIT;

chapin,

7967;

Goush,

196g)

and

the"social

tansla^tio*;;t*

(o,Surivan

&

Guilford,

1976).

T]^e

pictoriar

measure

of

sociar

understanding

was

the

Faces

Te-st (Mayer,

salovey,

Caruso,

and

sitarenios,

2002),

while

tire

video-based

measure

was

the

Interpersonal

perception

Task-15

liortu"ro

äAr.h"r,

rg93).

The

tasks

for

the

sociär

memory

facät

were

art

newly

constructed.

The

Tacit

9:1t"0g,"

tnventory

for

Managers

(TKIM;

wug^e,

d

S,;;Jä

,

rsgt)served

as

the

verbal

measure.

of

sociaf

knowredge.

Ä

confirmatory

iactor

analysis

supported

the

post'lated

trait

structure

#itni"

th"

;"1il;ämg"^."

perfor_

mance

measures/

when

variance

due

to

verbal

content

*u,

lo.,irolled.

The

fit

indices

for

moder

with

best

data

fit

were

as

fonows:

cpi

r

.öo+

;

v21267

:

30'277,

p

:

'256;

RMSEA

:

.037

with

a

90%

confidence

interval

of

[.000,.0g5];

SRMR

:

.056.

The

moder

is

depicted

in

Figure

10.3.

It

postulated

three

corre_

lated

trait-factors

corresponding

to

the

a"rYg.,

of

the

strlJy (rir.,

,o.iur

under-

standing,

memory,

and

knowleäge)

and

a värbal

facto,

with

buäirrg,

from

all

measures

based

on

verbal

materiäI.

FACETS

OF SOCIAL

INTELLIGENCE 219

Figure 10.3

Structural model of social

intelligence

(standardized

solution; ML). SK

=

Social Knowledge;

SM

=

Social Memory; SU

=

Social Understanding; TKIM P1-3

=

Tacit

Knowledge Inventory for Managers Parcel 1-3;

SIT

=

Chapin

Social

Insight Test;

ST

=

Social Translation Test; Faces

=

Faces Test; IPT-15

=

Interpersonal Perception

Test-15.

The

social

knowledge

factor correlated significantly

with the

social memory

and

the

social understanding factor

(.42

and

.50,

respectively). The

social

mem-

ory and

social

understanding

factors

also

correlated significantly

(i.e.,

.45). The

factor loadings

of

the manifest variables

on

the respective trait-factors

showed

a

coherent pattern. The loadings on

the verbal

method factor

were

heteroge-

neous, but all verbal indicators loaded

positively on

this factor.

We further investigated whether

social

intelligence was

separable

from aca-

demic

intelligence, as specified by the

Berlin

Intelligence Structure Test

(BIS-

Test;

]äger,

Süß, &

Beauducel,

7997). Corcelational

and

multiple regression

analysis

showed domain-specific overlap of the social intelligence trait-factors

with

specific domains of the BIS

(Weis

& Süß,

2004).

Results from

confirmatory

factor analysis

suggested still separable

trait-factors

of social and academic in-

telligence.

Additionally,

several social

intelligence

self-report inventories and

scales of Extraversion,

Openness,

and Agreeableness were assessed.

However,

just

as in

past studies, results did not show any evidence for the

convergent

construct validity

of performance

based

social

intelligence with

self-reported

social skills. Furthermore, self-report

data

on

social

intelligence

could

be ex-

plained,

to

large measure,

by the

personality

traits

that we assessed.

'ij-

220

Sociallntelligence

1.0.5

SOCIAL

INTELLIGENCE:

CURRENT

AND FUTURE

PERSPECTIVES

Despite arguably

uncritical

acceptance

of social

or emotional

intelligence

as

relevant

individual differences

constructs,

the introduction

of new statistical

methods

(e.g.,

structural equation modeling) provides opportunities for

clari-

fying formerly unresolved

problems.

Furthermore,

recent

advances in technol-

ogy, including digital means

of

stimulus

recording,

preparation, and

presen-

tation, allow the application

of

task material

that

is closer to real-life

scenarios

than

paper-and-pencil

drawings

or

black and white copies

of photographs.

However, future

research on social intelligence

is

still faced with

overcoming

the

failures and difficulties

of past

research and, thus, the

challenge of

prov-

ing the nature,

structu{e, convergent

and

discriminant

validity,

and

predictive

value

of

the

social intelligence construct.

L0.5.1

Importance

of Resolving Conceptual Issues

b

Ford

(1994)

claimed that social

intelligence

could not be

specified

as

a

pure

ability construct. According to Ford, individual differences

in

socially

intelli-

gent performance

should

not be

specified

without considering

situational

de-

mands,

social

values, and

personal

airns. Weinstein

(1969)

also related

socially

intelligent

behavior to its underlying intentions. For Weinstein,

one aspect

of

social intelligence is

the

ability to

manipulate

the

responses

of others. As a

matter of course, the

eventual exhibition of social

behavior

cannot be

specified

without

considerilg, and

perhaps specifying

in advance, the

relevant äelirnit-

ing conditions in

which social intelligence operates. Nevertheless,

it is neces-

sary to differentiate between the fundamental

cognitive ability

structure and

the

conditions that allow

or

influence the final performance

of social behavior.

If not, this criticism

might

justifiably

be

applied

to the construct

of

academic

intelligence.

Of course, intelligent

performance

in real-Iife

situations

certainly

depends on

present

moods

or

motivation, and/or

on peer group values

(Steele,

1997).

So

do

socially relevant

personality

traits

(e.g.,

Agreeableness,

Extraver-

sion), while

social

interests

clearly

influence

socially intelligent behavior

in

everyday-life.

Even

so,

certain cognitive determinants

of socially intelligent

behavior are

necessary requirements for the accomplishment

of social tasks

and

need to be identified by

empirical

research.

Consequently,

concepts like

social

engagement,

social

interests, or Machiavellian world

views

should not

be

con-founded with a

pure social

(cognitive)

intelligence construct.

No

matter whether

future

studies

rely

on

broad

measurement approaches

or rather

focus

investigation

on specific

domains

of social intelligence, the

con-

ceptualization of the

design demands a thorough specification

of the intended

measurement

constructs and

the corresponding

task

requirements.

This ap-

proach has

proven useful in the academic intelligence

domain and

we argue

that

it is equally important

when considering social intelligence.

Thus,

even

when the

focus is

on

a

narrow constructs,

which

claim to measure

social intel-

ligence in terms

of specific

social skills,

there will be a need

to

place

these

con-

SOCIAL

INTELLIGENCE:

CURRENT

AND FUTURE

PERSPECTIVES 221

stlucts

within a higher-order

framework.

Analogous

to Carroll's

idea

to sys-

tematically integrate

the

various

specific and more

general

constructs

of

acad-

emic intelligence

in

his

hierarchical

Three-Stratum-Theory

(1993),

it might be

possible to estäblish

a framework

of social

intelligence with comparable char-

acteristics. In his late

work,

Guilford

(1981,

1985) already

recognized the pos-

sibility of several

higher-order

factors within his

Structure

of Lrtellect Model.

Anyway,

empirical

studies

and

theoretical

accounts

are

far away from solving

these questions

that are

thus still

subject

to

speculations.

10.5.2 Resolving

Measurement

Issues

Design

issues.

Besides these

fundamental conceptual concerns, past

research

has

clearly

demonstrated

that the application of

MTMM designs is inevitable

for avoiding

any effects of

task material on

the research results.

Additionally,

the construction

of

new

tests appears necessary

for all different facets of social

intelligence.

The

latest technical developments,

(i.e.,

DVD,

digital cameras,

web-based

test delivery,

and

so

forth) ailow

the development of task

materlals

that are closer

to real-life scenarios

than only verbal performance

measures.

Relying

on spoken

language

(auditory

stimuli) seems

just

one

way

to realize

the assessment

of socially

relevant attributes.

Furthermore, new

tests

should

take

into account

the topicality of

the

social

milieu

(just

as intelligence tests

need to be

modified

to take into account emerging

historical events,

technblo-

gies,

and the like).

Validqtion.

According

to the

postulated

performance

model of social

intel-

ligence,

social

behavior

appears to be an adequate criterion

to validate social

cognitive

intelligence.

However, considering

the aforementioned criticisms,

the conceptualization

of appropriate

indicators

of social

behavior seems

to be

a

difficult

obstacle

to traverse.

The exhibition of intelligent social

behavior

is

certainly

irLfluenced

by the

social

environment, present

moods, prevailing

so-

cial norms,

values, and so

forth. In this

respect, it appears difficult,

if at all

possible,

to

assess

performance

in real-life contexts

under

the

control of

all rel-

evant boundary

conditions constituting

the social

world. This

point

notwith-

standing,

it appears

important not

to

lose sight of

the need to specify

limiting

conditions

in

advance. Consequently,

future

studies

need to

establish

a more

comprehensive

classification of social

settings

and

both

universally and specif-

ically

valid criteria

for the

judgment

of socialiy

intelligent behavior.

Investigating

the construct

validity

of social

intelligence

also needs to match

the latest state-of-the-art

in

scientific

research

in terms of the selection

and

specification

of

validation criteria. Certainly,

the replication of past

findings

by applying similar or

the same

measures of academic

intellectual

skills

is

valuable.

However, in order

to

gain

further

information about construct

va-

lidity,

validating social

intelligence

performance

with what are now thought

of as

relatively obsolete

indicators of

academic intelligence

(e.g.,

simple grade

point

average) or apparently

deficient operationalizations

of

g

appears inade-

quate.

In any case,

it

should

be stated clearly,

in correspondence

with under-

222 Sociallntelligence

lying theory,

what type

of

academic

intelligence

is purportedly

assessed

(i.e.,

g, ciystalliied

inte[ifence,

reasoning

abilities,

or

some other

constellation

of

measurement

constnicts)

and

the

strata

upon

which

the construct

resides

(see

Carroll,7993).

Building

the bridge

to

the

main

topic

of

this

book,

certainly,

empirical

inves-

tigations

aie

req,rirä

that

allow

conilusions

about

the overlap

of

social intelli-

gänce

to the puiportedly

of

emotional

intelligence

(Matthews

ät

ul.,

ZOOZ;.

As

mentioned

at

the

beginning

of

this chapter,

at an

empirical

level,

few

data

is available

that

provides

evidence

for the

relation

of

the

two

constructs

(Davies et al.,

1998).

At

a theoretical

level, some

cofiunentators

see

the

constructs

as positively

interrelated

(Salovey &

Mal'eL

1990;

Sternberg

et

al.,

2000).

More

specifically,

salovey

and

Mayer

(1990) defined

"emotional in-

telligence

as

a sJbset

of social

intelligence"

(p.

189).

At

the same

time,

emo-

tionäl

intelligence

was concePtualized

as

a kind

of

metacognitive

ability

(Gole-

man,

1995)

äith

effects

or,

äU

kinds of

cognitive

tasks,

including

tasks

from

the

domain

of social

intelligence.

Anyway,

the

absence

of a

common

model

of

social

intelligence

and

tie

elusiveness

äf

emotional

inteiligence

(Zeidher,

Matthews,

&

Röberts,

2001)

inhibits

a

more

detailed

theoretical

description

of

construct

overlaP.

Consequently.

any statement

about

the

relation of

these

two

constructs.can,

at present,

only

be derived

from

a comparison

of

operationalizations.

For ex-

ample,

the

Mayer-salovey-Caruso

Emotional

Inteiligence

Test

(MSCEIT;

Mayer

et

il.,Z00Z1is

based

on

a performance

model

of emotional

intelligence

(Mayer

& Salovey,

1997)

containing

four

branches

of

different

classes

of

abilities:

per-

ception

of

emotion

(Branch

1),

emotional

facilitation

of

thought

(Branch

2),

understanding

emotions

(Branch 3),

and

managing

emotions

(Branch

4)'

A

detailed

descrlption

of

the

model

and

the

four

branches

can

be found

in Chap-

ter

2by Neubauer

and

Freudenthaler

in

this

volume.

Table

10.1

contrasts

the

operationalizations

of

the

MSCEIT

with

some

traditional

operationalizations

of

social

intelligence.

tra

a

q)

'l

ö

U

.o

..1.t

o

.F

G9

pi

Yoü

::-X

oü

I

J-d

*pa

6:=

;U^

-.!u

horbO

:l*=

s03

*E*

c..i

*'

SOCIAL

INTELLIGENCE:

CURRENT AND

FUTURE

PERSpscTryEs

va

üd

vv

m

3öF

üR

ü

ä

g..r'n;!DoE:

F"i,X EOg8; ** *

T ä e fr E":XE

*

q-8

e: ärH.ä+i5 tFE

,IüEö€aLs

üs.i

d\v

JnF'H

C)

gd

I Pq€.^

tE

5

-9t' s v

EE -E

AFeFE

E4

ä;

€E€Eb

Rs

*E

HEF"

äöo

osl

;ril?E

;.= Yo

üE'O*o

Es

ör F.EE*fr

Hor .('<

.9F'ü,e

t

.öE ,; EöFEE

€B

äE

gs3$H

aE .q.'i

'-*

(

c *

g;

us

SeqF:

eU H *o ä E i

u

a

üäi

.:qrEch*ic4E

sä$i

B€EE^!8+Q

ä

be!äbt'Fi&

f

se

a

Er ir&= *

?'EE

H

Y€EgS:öFs=."öo

EaEE:.äEUEi"cF

J E

g

Ä

ä;

E

E I8 E ;

!O

64.

xü

oo

=a

(!Oo

.Föo

"-

'-

0J

:HO

CJd>

J4E

o

G.- O

*E ä.

.ci0X

'o

o€

>\(6

d'^

-5.-

!?o

-u<

Ht

-Vd

o

c,)

(Eul

tU

qD

a'i' d

kx,^

*cB

r9g

;i0)x

F

o.qr

I

l*

8l&

cil

bol

=l

EI

qt

'-l

?l

'ul

ol

qtl

trl

ol

ft

GI

NI

qt

öl

FI

b 1.9

l3

l0r

l'o

t>,

lr

lc

IE

Itr

lö

IX

lEl

I

Frl

-t

r!l

UI

cnl

>l

c.r

I

.qt

#l

ol

trt

ol

Fl

Slc

.i

t-

dto

klcJ

aJ

t^

o-lH

-rl

I

l.

IV

lo

l.(g

IF

e.

OJ

oo

o

cn

*

o

.0,,|

-

öo

d

a

t

k

U

c.l

öl

H

,

U

vl

OJ

d

H

6

e

N

e

(g

H

Fi

Fl

F

q

+

U

-

bo

=

e.

Äa'" ö

0Jq

^.

--P

';

'.i

+

'F1

tr 9.,|

YHV

9

.qtr

'-

i^

.i

':.=

äigFs

Y AAU

!!^

P-!:

esI9g

.:':

'H

st

ü

-JVyr

P5

T

F€

ö

6HIi

>-

(üotrtr9

cJ+j(d!?o

lUdv0i

aü9w\)

eg"csF

ht I

o

ä

^.-:#!

Y'-H

tr I

'-

>.,-

rii X

C.o

'c

il !.1'! x

o'

rx(!ö0r

-äiN@r

rrln- Y

u

ä'ö.x >-k

trr6.;i.i

ä'>-h ö

tr

bo5Xt9

F 3

ö-P

H

A-!

q2

ft

!ruöJd

*

oG

orä

x; ä >-s,

v+x!H

EJ ä=J

u

o

E'A

o

H+.:(u+

F.l

c,, i

a-, o-,

,,-9la

il----

-L--

ä

"

*",

u

o;

,;

F-t

ü^

E

9.^-'6.

H

S

3&:eE s

.E :,d-i3 R

+to +

go>

E

s9EoVQ

UJTUUHTID

OG

v!

Hq

qJ

d -- 9**4

-q Ä ,'; h

o

x-

+-=

ö.9oJF

+..i';ictotrtr

g*

4öHfrp

3ä

3+gHä_P

9ö

EüEäES

H

L

|

9.9,q I )

.99

öä.EtEE

+E

Ehe;H:

-rn+ö:l

Il: orhi9?EE9J

: e i: E

I

o4

.

!J@j!iq7f(u(6

5rtr-O^Y^^n'!

:cö.rEr:1 äüo!

# SE

Ef

g#E€

F

A=

tr.g.:i

"r

s

öE

ui

-

gr

o.! e&

o

ü I

s

ü-e*

QE

oq

q

9

or

=.:.F.EsEE<lf

'jä8t;;Fn;E?

Ei El P:ilHg frä

d.: orü.I.I-CP äi

o

fiP:VHITg#H

\-./:

boF

öFr

ö

o.ö ö

RA

-

ti

oE

9 > :

_.ä5!Orrq

P ü.!

p

.q.E

;

-rrul:nl.n

I

ö

c{ * E t

.if

a,.i

)

: ä ö

F;*-90

l;;t

H.!-Vd-:/

^#iLi-LH

Y

F fi

o

6.9'J

Lly*^fe

n,-HrUUr

YV-9pd=

Eää.E

frEE

.!'jrE

c V I

P

HIYF.IQ

Gö9

-6aL.rqil

-c

c

boo-d

ö

e

öo

s;;

ö

ä

sä lg

9u.yfi;

qo.9a.

r#tF

F

r'94

8

:ä

ö

3

ö

:ä

t

ö

<r+

d

qu

g> gH

m rrl m rrr

Social

Intelligence

L a^^ n

I c',:

:.?

(d

16;

* CN.A

E-c'; Y,

a

H

-.i\Oco

c

u

F t trH d

r_rro\\O

U.X

öF-

.9 I S F*o'

:^trr.L - /'E !

>i-1

\H

j.orf-

x x tr.9=\J.E --

!

i- Y Y

u'E'=..

^*rö

frE:::äEöOf

Pi

=

ü s ü

I

!

ö05

Uü

E

-

ai EY

rs

.E

E&

ÄX Gc

or-

I8 Fy

ir.g

^LV.i

;4.

=bo

liF

e"q

5

P

o

o-

-3

3

r€

j

I

ö

"

A ä'!. Y a

I

u

gö.E:6

ög

o

BHESTN+E

ü€

E#

ä fE;

PEPqilog6;

ä€rx-i?ü:H

ElEiäs#[ü

iE;!!€..'ä8"

d;i^F!-td4

tr.Y-:.a+?t-.5(g

:.gEsPF'tnx'Q

ö:; ü

.9

{o.E ;

;

Er'igEEgilF

u;

öor6-EU'9'do

q3

-.-L

wH-

x "x

n -i c

E

pH

H !

j-,

a

E!:EE

gF

+

u:.! CJ

H.=

o.iX"tr

h:

\

! Y- ^ V

v

=H!19v!

90(strtraH

i .r i A

H

Y

v

: äü

E

*.F

c!# c F

d

*j1

g

:r'H

ö

.9 :"^ ö

c.,ötr.! äu

.q

o.ö

bog

.i-c

-

rnu E ! Y.:

@

=/= !u

g.H-!Pt

ei

dFä.9ä

pE

3 ä I fr

e-

J';

doäJ;i ar(ü

H

O.:9

F

L

5

:ö858

:ä

UI

co

,.Q a

u

.r{

gq

caj

o

-o

'.)-1

öO

h*

cou

bo

fiH

OJJ

^L

!o

cJ*

JYa'n

ouX

9oY

J

X a.

oeY

Y -:Z

F>;

'^'n

ö'i

a

(g

9--

I

rl-

ut@

c

t,q

bol

-=l

aJl

tt

trl

:l

.gl

:l

at

sl

ol

EI

GI

NI

=l

cl

ol

!l!

!ulo

öl

o.

-li

lö

tq

l:

Its

l6

lE

t-

laJ

IX

lF.l

I

.l

r!l

UI

(nl

>l

qil

rrl

^l

;l

trl

ol

5l

6l

Nl-

(Ü

t.;i

Elh

o.lt-..1

ol

I

lo

t-v

lo

t.q

t-

qJ

CJ

-

U

r/)

q

-

bo

e

a

0i

k

U

€

öl

öt

:.

o)

k

0)

L

U

v)

e

q

e

ac

N

e.

d

H

11

F

ri

0)

r':

-

SOCIAL

INTELLIGENCE: CURRENT

AND

FUTURE PERSPECTIVES

Without

going

too

much into detail

of the

single

operationalizations,

most

of

those

belonging to

the

social

intelligence

domain

were

included

in

the

test

batteries

of

O'Sullivan

and

Guilford

(7966,1976)

and have already

been

de-

scribed

in the

first

part

of this chapter.

Some aspects

of

the

overview

shown

in Table

10.1

need to be comrnented

on. Two tests

of the

MSCEIT

(Pictures

and

Sensations)

do not find any equivalent

operationalizations

in

the domain

of social intelligence.

From

our

viewpoint,

it is

not conceivable

to construct

equivalent measuräs

for the assessment

of social intelligence,

as can

be done

for

other tests

of

the

MSCEIT.

For

two

tests

(Emotion

Management

[EM]

and

Emotions

in Relationships

[EiR]),

the TKIM

(Wagner

& Sternberg,7997)

repre-

sents a test

with rather

equivalent

cognitive requirements,

orLly differing

with

respect

to

the range

of contents

of

the

predeterrnined aim

of actions

(EM:

reg-

ulate

one's own

emotionsi EiR: achieving

an

outcome involving

other

people;

TKIM:

the combination

of both for the

solution

of

a

given

problem). Compara-

bly,

the

Chapin

SIT

(Chapin,7967;

Gough, 1968)

asks the test-taker

to

identify

the

most logical

or

intelligent

solution

or explanation

for a

given social

prob-

lem

and

just

omits the

effectiveness

ratings for the

alternatives.

Furthermor"e,

for most

of

the

MSCEIT tasks

the

purported task requirements

contribute

to

the

accomplishment

of

several social intelligence

tests. For

example, the

abil-

ity to

identify

how a

person feels based

upon their

facial

expression

(Faces,

MSCEIT)

contributes

to the

performance

in the

Faces Test

of

O'sullivan et;a1.

(1965)

(choose

one

of

four

photographs

of men's faces

that

expresses the

sarne

feeling

as

that

of

a

woman's

face). Moreover,

the ability

to

perceive emotions

as

specified

in the Faces

Test

of

the

MSCEIT

surely contributes

to the

accom-

plishment

of

the

Couples

Test

(Barnes

& Sternberg,

7989) where

test-takers

have to

decide whether

a

pictured couple

represents a

real

or

a faked

couple.

Furthermore,

the knowiedge

of

how

moods interact

(Facilitation

Task)

surely

contributes

to the ability

to choose the

one of four

verbal

statements that

de-

scribes

what

precedes,

or

will follow

a cartoon

situation

(Cartoon

Implications;

O'Sullivan

et al., 1965).

The knowledge

of experiencilg

possibly

con{licting

emotions in

certain

situations and

understanding

emotional

chains

(Changes

Task)

might also

help test-takers

to accomplish

the

Chapin SIT where

solu-

tions

or explanations

for

given problems have

to be identified.

At a

scale level,

the

SIT intends

to measure

the ability to

evaluate

others, to

foretell what

may

occur

in interpersonal

and

social

situations, and the

ability to

rectify disturb-

ing

tensions

or conflicts.

Conceptually, this

definition

certairLly

contains the

requirements

of the

scale definiiion

oi Branch

3

(Understanding

Emotions),

that is, the

ability

to understand

emotional

information,

how

emotions

com-

bine

and

progress through

relationship

transitions,

and to

reason

about

such

emotional

meanings.

Obviously,

several abilities

belonging

to the emotional

intelligence

construct

form

a

subset

of those abilities

belonging to the

domain

of social intelligence.

This

supports

the

early conceptualization

of Salovey and

Mayer

(1990).

As a

matter

of course,

the last

considerations

only

represent

statements

about the

face

validity

of

the

compared tests

and

only with respect

to the

constructs

as

assessed

by the

given operationalizations.

Any further

going

conclusions must

ry

l

226 Sociallntelligence

be

subject

to

empirical investigations.

Flence, at present,

many questions con-

cerning the

overlap of social

and emotional

intelligence

remain unanswered.

For

example,

do the tasks

with no corresponding social intelligence

test

assess

a facet

of

emotional

intelligence

independent

from

social intelligence? That is,

is

emotional

intelligence

not only

a

subset

of social

intelligence,

but contains

distinct abilities? Or is

it possible

to regard social

and

emotional intelligence as

constructs

comprising

the same cognitive

requirements based

on

two dif{erent

kinds of contents

(social

vs. emotional contents)?

10.6 CONCLUDINGREMARKS

To

conclude

the chapter,

we would like to provide some surunarizing

state-

ments on

the assessment of social

intelligence, the associated

problems and

expected

future challenges.

Without repeating

in

detail future requirements

as already outlined

in this chapter, elements

that appear most

important for

successful

studies are

(a)

a theory-guided

approach to

the conceptualizaticin

of

the

construct

with respect

to higher- and lower-order facets and necessary

task requirements,

@)

the control

of method-related

variance

(".9.,

by MTMM-

designs), and

(c)

the application of

nonverbal and auditory

task material to

enhance

real{ife equivalence.

\Arhen the construct can be

conceptually

,!le-

lineated

and adequately operationalized,

examining the

construct and

finally

predictive

validity must be

the focus of research. It appears inevitable for the

conduct of

useful studies

to

provide

evidence for

the

convergent validity as

an essential step

for the establishment of

a new ability

construct

(Süß,

2001).

Moreover, it should be noted

that the

subfacets

of a new hierarchical

concept

are in need of support

by evidence of convergent

validity.

Last but not least,

a discussion

about

the position

of social

and

emotional intelligence within the

field

of

individual differences

research appears indispensable.

Lr our view, it

does not appear convincing

to generate practical relevance

only

by the lexi-

cal introduction of new ability

constructs. The apparent

gold

rush associated

with the

introduction and exploration of emotional

intelligence

might

easily

seduce

researchers to adopt

its irnportance from lalpersons'

theories without

supporting

the

relevance

by meaningful empirical evidence,

especially

for the

convergent

and incremental validity

(Sriß,2001).

In this

respect,

Schaie

(2001)

elaborates

necessary steps towards

the

establishment of the emotional intel-

ligence construct like

a

comprehensive

convergent and discriminant valida-

tion, a well-founded

selection

of

the validation

sample, and the application

of

multivariate statistics for data

analysis. Besides the already

mentioned further

duties

as elaborated

in this

chapter,

these methodological

challenges might in-

spire

researchers to

come

up

with-as a

seemingly overdue

step-a book on

social intelligence in the near

future.

REFERENCES 227

REFERENCES

Baltes, P. 8.,

Staudinger,

U.

M., Maercket

A., &

Smith,

j.

(1995).

People

nominated

as wise:

A

comparative study

of wisdom-related knowledge. Psychology

and

Aging,

10,155-166.

Barnes, M. L., &

Sternberg,

R.

I.

(1989).

Social intelLigence and decoding of nonverbal

n,,oc I u t ot I io"^""

13, 263_287.

Biown,

L. T., & Anthony, R.

G.

(1990).

Continuing

the

search

for

social

intelligence.

P er s on

ality

and India idual D if er enc es, L L, 463 47 0.

Cantor, N., & Kihlstrom,

|.

F.

(1987).

Personality

and

social intelligence. Englewood

Clif{s, NJ: Prentice Hall.

Carroll,

I.

B.

(1993).

Human cognitiae abilities: A suraey of

factor-analytic

studies. New

York:

Cambridge

University Press.

Chapin, F. S.

(1942).

Preliminary standardization of a social insight scale. American

Sociolo gical Reaieu;,

7, 21,4--225.

Chapin,

F.

S.

(1967).

The Social

Insight

Test. Palo Alio, CA: Consulting Psychologists

Press.

Costanzo,

M., & Archer, D.

(1993).

The Interpersonal PerceptionTask-l1

(IPT-1,5)

[Video-

tape]. Berkeley, CA: University of California Extension Media Center.

Davies, M.,

Stankov,

L., & Roberts, R. D.

(1998).

Emotional intelligence: Lr search of

an elusive construct.

lournal

of Personality and Soci,alPsychology,75,989-101,5.

'

Feldman,

R.

S.,

Tomasian,J.C., &

Coats,

E.

I.

(1999).

Nonverbal deception abilities and

adolescents

social competence:

Adolescents with higher

social skills

are better

liar s.

I

ournal of N ona

erb al B ehaoior, 23, 237

-249

.

Ford, M.

E.

(1994).

A new conceptualization of social intelligence.

In R.

].

Sternberg

(Ed.),

Encyclopedia of human

intelligence

(pp.

974-978). New York: Macmillan

Publishing

Company.

Ford, M. E.,

&

Tisak, M.

S.

(1983).

A further search for

social

intelligence.

lournal

of

Educational P

sychology,

7 5, 1.96-206.

Frederiksen, N., Carlson, S., & Ward, W. C.

(1984).

The

place

of

social

intelligence in a

taxonomy

of

cognitive abilities. Intelligence

,

I

,

3L5-337 .

Gardner, H.

(1983).

Frames of mind: The theory of multiple intelligences. New York: Basic

Books.

Goleman,

D.

(1995).

Emotional

intelligence:

INhy

it can matter more than

IQ. New York:

Bantam

Books.

Gough, H. G.

(1968).

Manual

for

the

Chapin Social Insight Tesf. Palo Alto, CA: Consult-

ing Psychologists Press.

Guilford.,

I.P.

(1967).

The nature of human intelligence. New York: McGraw-Hill.

Guilford,

J.

P.

(1981).

Higher-order

structure-of-intellect

abilities.

Multioari.ate Behaa-

ioral

Research, 16,

411-435.

Guilford,

I.

P.

(1985).

The structure-of-intellect model. In B. B. Wolmn

(Ed.),

Handbook

of intelligence. Theories, measurements, and applications

(pp.225-266).

New

York:

Tohn

Wilev &

Sons.

i

228

SociallntelLigence

Hendricks,

M., Guilford,

J.P.,

& Hoepfner,

R.

(1969).

Measuring creatiae

social

intel-

ligence

(Psychological

Laboratory

Reports

No. 42). Los Angeles:

University of

Southern

CaLifornia.

Jäger,

A. O.,

Süß,

H.-M.,

& Beauducel

,

A.

(1997).

Berliner Intelligenzstruktur-Tesf

[Berlin

lnteliigence

Structure

Test].

Göttingen,

Germany:

Hogrefe.

Jon-es,

K., &

Day,

I.D.

(1,997).

Discriminations

of two

aspects

of

cognitive-social

irrtelli-

gence

from

academic

htelligence.

lournal

of Educational Psychology

,

89,

486497 .

Keating,

D. P.

(1978).

A search

for social

intelligence.

lournal

of Educational

Psychology,

70,2L8-223.

Ki.hlstrom,

I.

F.,

& Cantor,

N.

(2000).

Social

intelligence.

In R.

j.

Sternberg

(Ed.),

Hand-

book of

intelligence

(pp.359179). New York: Cambridge

University

Press.

Kosmitzki,

C.,

&

John,

o. P.

(1993).

The

impJicit use of

explicit conceptions

of social

intelligence.

P

ersonality

and Indiaidual

Dffir

ences, 1' 5, 11'-23

-

Lee,

J.-E.,

Duy,

J.D.,

Meara,

N.

M., & Maxwell,

S. E.

(2002).

Discrimination

of social

knowledge

and

its flexible

application

from creativity:

A multitrait-multimethod

approach.

Personality

and Indioidual

Dffitences,

32,913-928.

4

Lee,J.-E.,

Wong, C.-M.

T.,

Duy,I. D.,

Maxwell, S.

E., & Thorpe,

P.

(2000).

Social

and

academic

intelligences:

A multitrait-multimethod

study

of

their crystallized

and

fluid characteristics.

Personality

and Indiaidual

Differences,

29, 539-553.

Marlowe,

H. A.

(1936).

Social

intelligence:

Evidence

for multidimensionality

and con-

struct

independence.

I

o ur n

aI

of

E ducatio

nal P sy cholo gy, 7 8, 52-58

Matthews, G.,

Zeidner,M.,

&

Roberts,

R. D.

(2002).

Emotional

intelligence: science

and

myth. Carnbridge,

MA:

MIT Press.

Mayer,

J.

D., DiPaoto,

M., & Salovey,

P.

(1990).

Perceiving

affective content

in ambigu-

ous

visual stimuli:

A component

of

emotional

intelligence

.

lournal

of Personality

Assessment

,

54,

772-781..

Mayer,

J.

D., & Salovey,

P.

(1997).

\Alhat is emotional

intelligence?

Lr P. Salovey

&

D.

J.

Sluyter

(Eds.),

Emotional

deoelopment

and emotional

intelligence: Educational

implications

(pp.

3-31).

New

York: Basic

Books.

Mayer,

J.

D., Salovey,

P., Caruso,

D. R., & Sitarenios,

G.

(2002).

The Mayer, Salouey,

and Caruso

Emotional

Intelligence

Test:

Technical manual

Toronto, Canada:

Multi-

Health Systems.

Moss,

F. A., & Hunt,

T.

(1'927).

Are you

socially

intelligent? Scientific

American,

137,

108-110.

Moss, F.

A., Hunt, T., Omwake,

K. T., &

Woodward,

L.

G.

(1955).

Manualfor

the George

Washington

lJnioersity

Series

Social

Intelligence Tesf

.

Washington,

DC: The Center

for Psychological

Service.

Orlik,

P.

(1978).

Soziale

Intelligenz

[Social

intelligence].

In K.].

Klauer

(Ed'),

Handbuch

d er P

ädago gis chen

D iagno s

t ik

(pp.

347154).

Düsseldorf

,

Germany:

Schwann.

O'Sullivan,M.,&Guilford,

j.P.

(1966).

SixFactorTestof Sociallntelligence:

Manualof

in-

structions

and interpretations.

BeverTy

Hills, CA: Sheridan

Psychological Services.

O'sullivan,

M., & Guilford,

J.

P.

(1976).

Four

Factor Tests of Social

Intelligence: Manual

of

instructions and

interpretations.

Orange, CA:

Sheridan Psychological

Services.

O'Sullivan,

M., Guilford

,1.P.,

&

deMille, R.

(1965).

The measurement

of social

intelligence

(Psychological Laboratory

Reports No. 34).

Los Angeles: University

of Southern

California.

REFERENCES

))q

Probst, P.

(1982).

Empirische

untersuchung zurn

Konstrukt

der "sozialen

Intelligenz"

[Empirical

investigation

on

the

'social

iltelligence'

construct]. In K.

päwhk

(Ed.),

Multioariate

Persönlichkeitsforschung

(pp.

201-226).

Bern,

Switzerland:

Hans Huber.

Riggio, R. E.

(1986).

Assessment

of basic

social skills.

lournal

of Personality

and

Social

P sy chology,

51,

649-660.

Riggio,

R. E.

(i989).

Manualfor

the

Social SkiIIs

Inaentory. PaIo

Alto,

CA: Consulting

.

Psychologists

Press.

'

Riggio,

R. E.,

Messamer,

J.,

& Throckmorton,

B.

(1,991).

Social and

academic

intelli-

gence:

Conceptually

distinct but

overlapping

constructs.

Personality

and Indiaid-

ual

Differences,

12,

695-7 02.

Salovey,

P., &

Mayer,

I.

D.

(1990).

Emotional

intelligence.

Imagination,

Cognition and

P

ersonality,

9, 185-277.

Schaie,

K. W.

(2001).

Emotional

intelligence:

Psychometric

status and

developmental

characteristics-Comment

on Roberts,

Zeidner,

& Maithews (2001).

Emotion,

1,

243-248.

Schneider, R.

j.,

Ackerman,

P. L.,

& KarLfer, R.

(i996).

To

"act

wisely

in human

relj-

tions":

Exploring

the dimensions

of social

competence.

Personality and

Indiaidual

Differences,

4, 46948L

Süß,

H.-M.

(2001).

Prädiktive

Validität

der

Intelligenz im

schulischen und

außer-

schulischen

Bereich

[The

predictive validity

of intelligence

in the

scholastic and

non-scholastic

domain].

In E.

stern

&

|.

Guthke

(Eds.),

Perspektioen

der

InteIIi-

genzforschtmg

(pp.

109-136). Lengerich,

Germany: Pabst

Science

publisher.

Süß, H.-M.,

& Beauducel,

A.

(2005).

Faceted

models

of intelligence.

In

O. Wilhelm

& R.

Engie

(Eds.),

Handbook

of understanding

and measuring

intelligence

(pp.

313-

332). Thousand

Oaks, CA: Sage.

Staudinger,

U.

M., Lopez,

D. F., & Baltes,

P. B.

(1997).

The

psychometric

location

of

wisdom-related

performance:

Intelligence,

personality or more?

Personality

and

Social P sy

chology Bulletin,

23, 1200-121.4.

Steele,

C.M.

(1997).

A threat

in the

air: How

stereotypes

shape the

intellectual

identi-

ties

and

performance

of women and African-Americans.

American

Psychologist,

52,6rH29.

sternberg,

R.

I.

(1998).

A balance theory

of wisdom. Reoiew

of

General

psychology,

2,

347-365.

Sternberg,

R.

J.,

Forsythe,

G.

8.,

Hedlund,

j.,

Horvath,

J.

A., Wagner,

R.

K., Williams,

W. M.,

et al.

(2000).

Practical

intelligence

in eoeryday

lr/e.

New

york:

Cambridge

University

Press.

Sternberg,

R.

I.,

& Smith,

C.

(1985).

Social inteiligence

and decoding

skills in nonverbal

communicatron.

S ocial Co gnition,

3, 168-792.

stricker,

L.

I.,

& Rock, D.

A.

(1990).

lnterpersonai

competence,

social intelligence,

and

general ability.

P

ersonality and

Indiuidual

D

ffirences,

71,

833-839.

Thorndike,

E.

L.

(1920).

Intelligence

and

its use. Harper's

Magazine,740,227-235.

Thurstone,

L. L.

(1938).

Primary

mental

abilities.

Chicago:

University

of

Chicago Press.

vernon, P.

E.

(1933).

some characieristics

of

the

good

judge

of

personality.

lournal

of

Social P sy chology,

4, 42-57 .

':a.E

'I

I

230

Sociallntelligence

Wagner,

R. K.,

& Sternberg, R.

l.

(1991).

Tacit knowledge

inuentory

for

managers.

San

Antonio,

TX:

The Psychological

Corporation Harcourt Brace & Company.

Walker,

R.

E.,

& Foley,

].

M.

(1973).

Social

intelligence:

Its history and measurement.

P sy

cholo

gical Rep or ts,

3 3, 839-864.

Weber,

H.,

& Westmeyer,

H.

(2001).

Die Inflation

der Intelligenzen

[The

inflation

of

intelligencesl.

ln E. Stern &

].

Guthke

(Eds.),

Perspektiaen der

Intelligenzforschung

\pp.

257-266).

Lengerich,

Germany:

Pabst

Science Publisher.

Wechsler, D.

(1981).

Wechsler Adult

Intelligence

Scale-Reaised. New York: Psychologi-

cal Corporation.

Wedeck,

I. $947).

The relationship

between

personality

and

'psychological

ability'.

British

lournal

of Psychology,

37

,

133-151.

Weinstein,

E. A.

(1969).

The development

of interpersonal

competence. tn D. A.

Goslin

(Ed.),

Handbook of socialization theory

and research

(pp.753-775).

Chicago: Rand

McNaliy.

Weis,

S., & Süß, H.-M.

(2004).

Social intelligence-Inoestigating its

structure and construct

oalidity in a

multitrait-mtLltimethod

design. Manuscript

submitted for

publicat{on.

Wong,

C.-M. T., Duy,

1.D.,

Maxwell, S. E.,

& Meara, N. M.

(1995).

A multitrait-

multimethod

study of

academic

and social intelligence

in college

students.

/our-

nal

Social Intelligence--A Review and Critical Discussion of Measurement Concepts.

Abstract