people-s conceptions new haven ct of dept of … · aa0 0 yale univ new haven ct dept of psychology...
TRANSCRIPT
AA0 0 YALE UNIV NEW HAVEN CT DEPT OF PSYCHOLOGY F/6 5/10
PEOPLE-S CONCEPTIONS OF INTELLIGENCE.(U)OCT 80 R J STERNBERG. B E CONWAY, J L KETRON N0001-78-C-0025
UNCLASSIFIED RR-780 NL
*~ EIIImIIIIIIl-IhlEEEEllllEm*IIIIIIIIIIIIIn**nInnnn*-EEIIIIEEIIEElinu*
4=4A Pople'sConceptions of 'Intelligence
Robert J. Steinberg,.Barbara E. Conway, Jerry L. Ketron, and Morty Bernstein
Department of PsychologyYale University
New Haven, Connecticut 06520
DTIC
Techniczal 1.'lort No. 28 SOctober, 1980
Approved for public release; distribution unlimited.Reproduction in whole or in part is permitted for
Vb- any purpose of the United States Government.
B This research vas sponsored by the Personnel andTraining Research Prograims, Psychological SciencesDivision, Office of Naval Reaearch, under ContractNo. N0001478C0025, Caract "uthority Identification
Number NR 150-412.
80 12 24_00 4.j- Now,
DISCLAIMER NOTICE
THIS DOCUMENT IS BEST QUALITYPRACTICABLE. THE COPY FURNISHEDTO DTIC CONTAINED A SIGNIFICANTNUMBER OF PAGES WHICH DO NOTREPRODUCE LEGIBLY.
-" . n m flm m , ; , ,.- -. ... - -T : -
..
UNCLASSIFIED.gCu'*11T, "LSSIFIChIOW OF THISS 0AG6r MIfwDo* fla. me..E)_________________
REPORT DOCUMENTATION PAGE BF01O4F COIPLETINGIo URPRNs tGOVT ACCESSION NO, 3-09 SsI
Technical Report No. 28 - ~ 5~4. TITLE (Bad Swbtile) TP rR 1V etc vo~
People's Conceptions of Intelligence, J~
Research Raport No. 7-8O0-I. CONTRACT OR GRANT NUMOERk.j
Department of Psychology 61 i'N;' UE
Yale nivesityRR 042-04; RR 042-04-01;
- ~. .IL ... 22'0111..) Is. SECUIRITY CLASS. (Of 9t90fole
164. OCCLASSIrICATION DW 4DN
SC N C UL
16 DISTRIBUTiON STATEMENT (of WAWp
Approved for pubj .c release; distribution unlimited
17 AMNTLEI hA .h rct .nfood i a -0
1SIJPP MEN TARY NOT ES
Journal of Personality '-d Social Psychology: Attitudes and Social Cnti;_:-in press
Is. KEY WORDS (CoBdaIwo on rorfIoe old* it ROe...wY and Identify by block ntmbot)
Intelligence, implicit theory, explicit theory
20 AbSTRACT (C*Mfift.@W O~ne -00Old* It POCOOSS'Y and 1de"fity by block nmbor)
Three experiments are reported Investigating experts' anid laypersons' con-ceptions of intelligence. In the first experiment, persons studying in acollege library, entering a supermarket, and waiting for trains in a rail-road station were asked to list behaviors characteristic of either "Inteli-tence," "academic intelligence," "everyday Intelligence," oir "unintellipence.and to rate themselves on each of the three kinds of Intelligence. In thesecond experiment, experts and laypersons (excluding students) were asked tn
DD IF ':'7, 1473 COITION OF I'NOV 65 OSS00OL91 UNCLASSMFED544@I?.P41..@lseCIuIry CLAS5IspleATION aI ir 14G% Fagg 10. Doe -_0
----------.
UI IASS1FlED
SrCUVTYY CLASS.FICATION Of it1b PAcjr t'ho.. tbje. Fanftedi,
rate various properties of the behaviors listed in 1:,periment 3; -laypersons also rated themselves on the three kinds of Intell;*u.-.n. r.:--took an IQ test. In the third experimeut, laypersons received %.rittvndescriptions of behaviors characterizing fictitious people, and %,-ere a.k.'to rate these people's intelligence. We found that people have v:l1-fc--prototypes corresponding to the various kinds of intelligence, tLt the.sCprototypes are quite similar for experts and laypersons, that the protot:"are closely related to certain psychological theories of intelli.erce, a.!that the prototypes are used in the evaluation of one's oim and r.ers'intelligence. Moreover, proximity of one's behavioral self-ch.'rJ'ac1r. zj*.:.-
to an ideal prototype is quite strongly related to Intelligence a- n.casujr:.by an IQ test.
I.
II
I.
I*
UNCLASSmn']ri ________
5fCCUMeI CLASIWFICA1I014 C- *a-T".* .. i
C A
People's Conceptions of Intelligence
Robert J. Sternberg, Barbara E. Conway, Jerry L. Ketron, and Morty Ecrnstein
Yale University
Running head: Conceptions of Intelligence
Send proofs to Robert J. SternbergDepartrcnt of Psychology
Acesulio Por Yale University
-xGRA&I 1OX A Yale StationNTIS R&DTic TAB El N~ew Haven, Connecticut 06520Unannouncedjust if icatiol DTIC
D~strb~1t1fl/~ELECTEDi DIi tc / _ _ .C 29 MOAvanI11It Codes
IAVU and/or i
Dist special. .. ii -/er
B I+?,+I 75cz
Conceptions of Intelligerce
2
Abstract
Three experiments are reported Investigating experts' and laypersons'conceptions of intelligence. In the first experiment, persons studying in
a college library, entering a supermarket, and waiting for trains in a rail-
road station were asked to list behaviors characteristic of either "intelli-
gence," "academic intelligence," "everyday intelligence," or "unintellienrce,"
and to rate themselves on each of the three kinds of intelligence. In the
second experiment, experts and laypersons (excluding students) were asked to
rate various properties of the behaviors listed in Experiment 1; tle
laypersons also rated themselves on the three kinds of intellig:nce and tock
an IQ test. In the third experiment, laypersons received written eescripticns
of behaviors characterizing fictitious people, and were asked to rate these
people's intelligence. We found that people have well-formed prctt, tyes
corresponding to the various kinds of intelligence, that these prctot.;pes
are quite similar for experts and laypersons, that the prototy':2s are clos.,y
related to certain psychological theories of intelligence, and that the
prototypes are used in the evaluation of one's own and others' intelligence.
Moreover, proximity of one's behavioral self-characterizations to an ideal
prototype is quite strongly related to intelligence as measured by an IQ test.
7.
Conceptions of Intelligcce
3People's Conceptions of Intelligence
Because of its importance in the everyday world as well as in psycho-
logical theorizing and measureTent, intelligence has been a heavily researc',
psychological construct during most of the present century. Research on
intelligence could be broadly classified as being of two types, dcpen.in;g F
upon the nature of the theory motivating the research. V
M:ost research on intelligence has been devoted to the construction e
testing of what might be referred to as "explicit theories" of intelliner.:E:I:i
Explicit theories are constructions of psychologists or other scientists
that are based or at least tested on data collected from people p:rfc..r--ir-
tasks presumed to measure intelligent functioning. For exarrple, a battery
of mental ability tests might be administercd to a large group of pccple
and the d.ta from these tests analyzed in order to isolate the prc.pDsE
sources of intelligent behavior in test perfcrmance. Although investicatcrs
working with explicit theories of intelligence might disagree as to the
nature of these sources of intelligence-- :hich might be proposed to be
factors, cc7ponents, schemata, or some other kind of psycr.,ocical c:r.str- ::--
they would agree that the data base from which the proposed constructs sh:jlt
be isolated should consist (directly or indirectly) of perfor-ance on tasl:s
requiring intelligent functioning.
A less sizable research effort has been devoted to the disco'.cry of "
might be referred to as "implicit theories" of intelligence: Iplicit the-
ories are constructions of people (psychologists or laypersons) that reside
In the minds of these individuals. Such theories need to be "disccvered"
rather than "invented" because they already exist, in sc-le for:-i, in peoplc's
heads. The goal in research on implicit theories of intelligence is to fid
Out the form and content of people's infonnal theorieb. Thus, one attenizts
Conceptions of lntelli Ence
4
to reconstruct already existing theories, rather than to construct new thL:ries.
The data of interest are people's connunications (in whatever form) regareir.
their notions as to the nature of intelligence. For example, a survey of
questions regarding the nature of intelligence might be administered to
large group of people and the data from this survey analyzed in order to
reconstruct people's belief systems. Although investigators working with
implicit theories of intelligence might disagree as to the nature of people's
beliefs, they would agree that the data base from which the prcpcsed corstrcts
should be isolated should consist of people's stated or exercised beliefs
regarding intelligent functioning.
We believe both explicit and implicit theories of intelligence shol
be of interest to psychologists. Explicit theories are interesting becase
the inportance of intelligence to psychological theory and m asurcmcnt, c:
well as to society, make it worthwhile to know insofar as we are able wEt
IntelliSence is; because these theories can serve as the basis for the
systematic and rational assessment, and eventually; training of intellicer2:e;
and because these theories can suggest where people's conceptions are adc;.ate
and where they are inadequate, znd thereby help shape these conceptions.
Implicit theories are interesting because the importance of intelligence
in our society makes it worth-.:hile to know what people mean by intelliee'~:;
because these theories do in fact serve as the basis of informal, everyday
assessment (as in college or job interviews) and training (as in parent-cil
interactions) of intelligence; and because these theories may suggest aspezts
of intelligent behavior that need to be understood but are overlooked in
available explicit theories of intelligence.
We believe the importance of implicit theories of intelligcnce has bcen
• .-:" .. . ..- Zd' . : i... - / , - - n .9
--
Conceptions of lntelliSnce
underplayed in psychological research. Most of the assessment and training
of intelligence that transpire in the real world are based upon implicit i
rather than explicit theories of intelligence. For example, many more
assessments of other people's intellectual abilities are made in the
course of interviews and even everyday social interactions (such as cock-
tail parties, conversations at coffee breaks, and the like) tlcn are rade in
the:evaluations of scores from intelligence tests. Moreover, people
(even psychologists!) seem ultimately to trust measurements made on thc
basis of their implicit theories more than they trust measurements ra' on
the basis of explicit theories. Psychologists conduct intervic.:s all of
the time, despite the notorious low validity and reliability of intervie-il
assessments; and psychologists as well as others seem to believe in the
outcomes of these interviews. We have much more often seen p:zPle cxpr -
sing astonishment at mental test scores that are inconsistent with the
people's informal assessments of the interviewee's intellectual capabilities
than we have people expressing astonishment at their own poor jud;ient after
finding out that the mental test scores were inconsistent with their perscnal
assessments.
.The remainder of this article will deal with people's conceptions, or
Implicit theories, of intelligence, although attempts will be made to inter-
relate these implicit theories to explicit ones, and to compare people's
subjective judgments with measurements from "objective" tests. We shall !I
concerned not only with what people's conceptions are, but with how people
use these conceptions in assessing the Intelligence of others on the basis
of descriptions of these others' behavioral tendencies.
P --. ,
Conceptiunt: of ntelli L.r,
6
One question that needs to be answered in the proposed approach to
intelligence is that of whose notions are to serve as the data base of
interest. The main groups that have been studied so far are experts
in the field of intelligence, adult laypersons, children, and individuals
(usually adults) from other cultures.
Most often, the "subjects" in this approach have been "expc-rts" on
intelligence. The most well-known example of the approach is probably
a symposium that appeared almost 60 years ago in the Journal of Edjcatic.--:
Psychology ("Intelligence and its Measurement," 1921). Fourteen experts
gave their views on the nature of intelligence, with definitions such as
the power of good responses from the point of view of truth or fact (E. L.
ThorridiLe), the ability to carry on abstract thinking (L. M. Terr.n), ar,
the ability to adapt oneself adequately to relatively new situations irt lif2
(R. Pintner).
Viewed narrowly, there seem to be as many definitions of ii.telliyer:z
as thcre were experts asked to define intelligence. Viewed broLIly', ho.,cr,
two themes seem to run through at least several of the definitior.s in thce
corplete set: the capacity to learn from experience and tho capacity to
adapt to one's environment. These themes run through definitions of more
recent experts as well. Ferguson (1954) has viewed intelligence prir-arily
in terms of the ability to transfer training, and Piaget (1972) has defir.c-
intelligence largely In terms of one's adaptation to the environ.ent in t.,ic',
one finds oneself.
Some psychologists have argued that laypersons should forw. at least
one population to be studied in research on people's conceptions of intelli-
gence. A leading proponent of this point of view is Neisser (1979), %%ho
is largely responsible for reawakening modern interest in people's concrtiCs
-
Conceptions of Intellir'ci,ce
7
of intelligence. According to Neisser (1979),
"intelligent person" is a prototype-organizcd Roschian conccpt.
Our confidence that a person deserves to be called "intclligcnt"
depends on that person's overall similarity to an imagined prc-
totype, just as our confidence that some object is to be callcd
"chair" depends on its similarity to prototypical chairs. There
are no definitive criteria of intelligence, just as there rc
none for chairness; it is a fuzzy-edged concept to which many
features are relevant. Two people may both be quite intelli-
gent and yet have very few traits in common--they resemble
the prototype along different dimensions. Thus, there is n,,
such quality as intellinence, any more than there is such a
thing as chairness--rese-7blance is an external fact and nct
an internal essence. ThcrE can be no process-based defiri-
tion of intelligence, because it is not a unitary quality. It
is a resc-mblance betwacr. t.,o individuals, one real ard t-.
other prototypical. (p. I65)
Neisser has noted that he is not the first to express such a vic,,, t.:hic, t
has traced back at least to E. L. Thorndike (1924):
For a first approximation, let intellect be defi-ed as th: Q -
ity of mind (or brain or behavior if one prefers) in respe:t t:
which Aristotle, Plato, Th..cydides, and the like, differe -:-st
from Athenian idiots of their day, or in respect to which the
lawyers, physicians, scientists, scholars, and editors of rc,-.te-
greatest ability at constant age, say a dozen of each, di'fer
most from idiots of that age in asylums. (p. 126)
Neisser has suggested that tests such as the Stanford-Binet have been re .-c-
ably successful because they consist of large numbers of items that assess
resemblance to different aspects of the prototype. Individual itcms fun:tiCr
like individual dimensions of a chair in the construction of a prototype.
Neisser (1979) has collected informal data from Cornell unJcrarad.c.
regarding their conceptions of what intelligence is. More formal studios ''.
.1 '
Conceptions of Intelligence
8
been conducted by Cantor (Note 1), who asked adult subjects to list attri-
butes of a "bright" person, and by Bruner, Shapiro, and Tagiuri (195C), O.o
asked people how often "intelligent" people also display other personality
traits. These authors found, for example, that intelligent pccple are li!c.'y ,
to be characterized as clever, deliberate, efficient, and ncr etic, bit ro t
as apathetic, unreliable, dishonest, and dependent.
Siegler and Richards (in press) asked adult subjects to characterize
intelligence as it applies to children of different ages. They found a trcr.
toward people conceiving of intelligence as less perceptual-,:otcr ar. as e
cognitive with increasing age. Yussen and Kane (in press) asked childre. in
thk. first, third, and sixth grades what their conceptions of irtelliger-c- zre.
They found that older children's conceptions were more differentiatedi t ,6r %.are
younger children's; that with increasing age, children incrcasi bly ch.r-ct-r-
ized intelligence as an internalized quality, that older children were less
likely than younger ones to think that overt signs signal intelli;Lnce, ar.d
that older children were less global in the qualities they associated wit ,
intelligence than were younger children.
Wober (1974) investigated conceptions of intelligence arc!-. r-.bers cf
different tribes in Uganda, as well as within different subgroups of the tribes.
Wober found differences in conceptions of intelligence both bet','een an.- ,.irIin
tribes. The Baganda, for example, tended to associate intelligence with
mental order, whereas the Batoro associated it with some degree of r~ent3l tur-c-'1..
In terms of semantic-differential scales, Baganda tribespeople thought of intelli-
gence as persistent, hard, and obdurate, whereas Batoro thought of it as soft,
obedient, and yielding.
Conceptions of Intolliicenc,
9
We performed three experiments investigating American adults' conceptioi1s of
intelligence. In the first, people in a train station, entering a super-
market, and studying in a college library were asked to list behavicrs
characteristics of either "intelligence," "academic intelligence," "everyday
intelligence," or "unintelligence," and to rate their own intelligence,
academic intelligence, and everyday intelligence. In the second experiMEnt,
both laypersons answering a newspaper advertisement and experts answering
a mail survey were asked to provide various kinds of ratings of the behaviors
obtained in the first experiment; the laypersons also rated their ow.;n intelli-
gence, acadcmic intelligence, and everyday intelligence. In the third
experiment, laypersons selected at random from a New Haven area phone book
were asked to rate the intelligence of various fictitious people who were
characterized in terms of differ-t mixes of the behaviors listed by ut - cts
in the first experiment.
EXPIRIM.ErNT I
In this experiment, we set out to compile a master list of intelligent
and unintelligent behaviors, and to ascertain various characteristics of
these behaviors and their relptions to the people who supplied them.
Method
Subjects
This experiment involved 186 subjects in all, including 61 people stud-,in
in a college library at Yale, 63 people waiting for trains in the Netr Haven
train station during morning and afternoon rush hours, and 62 people entering
a local supermarket.
Materials
Subjects received a blank page on which to list behaviors characteristic
of "intelligence," "academic intelligence," "everyday intelligence," or "unin-
4j.
Conceptions of Intulligcncce
10
telligence," and a page on which to rate themselves (on a l=low to 9:liih scalL',
on intelligence, academ'c intelligence, and everyday intelligence.
Design
Subjects listed behaviors characteristic of just one of the four in'es-
tigated attributes, but rated themselves on each of three attributes.
Procedure
People were approached by one of four experimenters (t:o nales, t .,o
females) in each of the locales and wiere asked to give five minutes of thEir
time to the experiment. They listed characteristic behaviors first, E-J-
then rated themselyes on the three scales.
Results
Comnlation of r4aster List of Behaviors
Behaviors listed by the subjects were compiled into a master list of
behaviors. Behaviors .,ere included if they were listed even just o lc, -
though obvious redundancies .:ere eliminated. The final list consistL2 cf
250 behaviors, of which 170 were for the various kinds of intelligerce
("intelligence," "academic intelligence," "everyday intelligence") ar
were for unintelligence.
Correlations of Frequencies of LisLed Behaviors
Table 1 shows the correlations between the frequencies with which ca:h
of the 170 intelligent behaviors was listed by subjects in each setting for
each type of intelligence. Since responses were summed over subjects, these
data can be interpreted for the three subgroups, but not at the level ef ir.i-
vidual respondents. In the library setting, frequencies of listed tc ;aviors
were significantly correlated for intelligence and academic intelligcnce
but not for intelligence and everyday intelligence; in the railroad a,!
supermarket settings, the opposite pattern of results obtained: correla-
tions between frequencies were significant for intelligence and ever.-,'.y
intelligence, but not for intelligence and acaidemic intelligence. J.j-ir,3
t.-
Conceptions of Irtlli;:..
11V
by the frequencies of behaviors listed for each type of intelligcrncc, tk.en,
we conclude that the denizens of the college library (mostly Yale uder-ra.;-
ates) perceived "intelligence" as being substantially similar to "acadc-ic
intelligence" but not to "everyday intelliqence." People in the railrcd.
station (mostly commuters) and supermarket (mostly housei..ives and schonl
teachers) perceived "intelligence" as being substantially similar to
"everyday intelligence" but not to "academ-ric intelligence."
Insert Table 1 about hcre
Ccrrelaticrs of Self-Ratings
Table 2 shows the correlations between the self-ratings of subje.cts ir.
each setting for each type of intelligence. In the library settirv, slif-
ratings of irtelligence wcre ver, highly correlated with self-ratir . cf
acadc--i: ir, tcliigence, buit only r:oecrately correlated -with self-ratings cf
everyda intelligence, with the first correlation significantly higher tiarn
the secor-d. Self-ratings of academic and everyday intelligence yore weakly,
althogL; si ,ificantly, correlated. In the railroad setting , the corrcla-
tions bet i,-.cn intelligence on the one hand, and both acrdc,,ic and E.cy:t.y
Intelligence on the other, were high and practically identical. The cor-
relation between intelligence and everyday intelligence was signific.anrtl'
higher in this group than it was in the library group. In the super,-.arket
setting, the pattern of results was intermediate between those of tf.".• ctEr
two group,: Self-ratings of intelligence provided by the subjects .: re
significantly more highly correlated with self-ratings of academic intelli- '
gence than with self-ratings of everyday intelligence, but the cnrrc.!t4zr.
of intelligence with everyday intelligence, and of academic intellic:"n.ce
with everyday intelligence, was intermediate between those of the ot.,or
two group.. Note that the overall pattern of results rather c'.osoly rci "'%
Conceptions of Intrlligence
12
those of the supermarket group, but does not reflect the variation in the
groups constituting the sample as a whole.
Insert Table 2 about here
Multiple correlations were computed bet-,een ratings of intelliGen:e
on the one hand, and of academic and everydcy intelligcnce on the other.
The multiple correlations were .83 in the library group, .81 in the railroed
group, and .87 in the supermarl:et group. The beta coefficients closcly
resembled the simple correlations in pattern: Academic intelliccr,: re- J
ceived higher weights than everyday intelligence in the library and supemar;.et '1
groups; everyday intelligence rc:-ived a slightly higher weight tl,..n cc.d-ic
intelligence in the railroad group. The overall multiple correl.-ich fcr
the groups combined %,as .83.
Conclusions
People appear to have organized conceptions of intelligcnt b:Lticr, iut
if intelligence is to be understood in terms of prototypes (Neisscr, I:7?),
then the results of this experiment suggest that there may be more th-n one
prototype. In particular, people seem to have at least somcv;hat different
conceptions of the meanings of "intelligence," "academic intelligence," and
"everyday intelligence," and these conceptions may differ across p:z.lztio.s
of subjects. The students and the railroad commuters seemed to recresent
two extremes, with the supermarket patrons falling in-bet;,'een. 7he urdcr-
lying continuum seemed to be one in which people differ with respect to their
relative weights on academic versus everyday aspects of intelligent bchavior
In understanding overall Intelligence. The second experiment was intencd tc
elucidate the structure and content of people's conceptions.
p
,-~ ~ L -.*•. • I
Conceptions of Intelligcnce
13
EXPERIME14T 2
In this experiment, we sought to ascertain what experts and laypersons
think intelligence is, and to compare their respective views. le also
wished to discover the degrees of correspondence an-ong reasured IQ, self-
rated intelligence, and self-rated descriptions of one's own belavior.
Method
subjects
There i:ere two principal groups of subjects in this experi,(;nt. The
first group comprised 122 laypersons from the NEw Haven area an: . ri, cre
of several advertisements in local newspapers. Because the results of the
first expcriiient suggested that students' conceptions of intelliL nce can
differ substantially from nonstud ,,.;' conceptions, and t-ecause our pri-
mary interest was in the general populatiun, students wVre exclJ:: frc-
participation. The second group conprised 140 experts in the field Of
intelligence. All experts were psychologists with do:tcral degrces doir
research on intelligence in major university and research centers arcnd
the country. They answered a questionnaire sent to ther, by mail. Tt c
return rate on the questionnaire was 48%.
Materials
Materials for the experimient consisted of a list of 250 bevicrs cc--
piled frc-i Experiment 1, with different questions asked about these tcavicrs
of different groups (described below); a page on which lay,-orscI.- c:2Il rz:c
themselves using a percentile scale on intelligence, academic intelligence,
and everyday intelligence; and, agaivi for laypersons only, the adilt-level
scale of the Hennon-Nelson Test of Mental Abilities, an ci'jn irtilsi~n
test of demonstrated high validity and reliability.
- . Es' .~ . .
ConL-ptions of lntclic.:.
Des ign 1
There were four different questionnaire groups. All four tc-:.tionnir -
were distributed to laypersons; only the first two questionnairc, were dis-
tributed to experts. No single person received more than one questionnaire.
All items required ratings on a 1 (low) to 9 (nigh) rating scale.
1. h,;portance ratings: Ideal person, In Questionnaire 1, suh,.cts
(75 experts and 30 laypersons) were asked to rate how inportant e-cf, of the
170 behaviors associated with intelligent (as opposed to unintelligcrt) f;r.:-
tioning was in defining their concEption of an "ideally" (a) intellic-rt per-
son, (b) academically intellincrt person, and (c) everyday intellicc.rt .prsc .
The "ideal" was described as the best possible on a given diner, Fic ,bt ro
further infornatioi was given.
2. Chracteristicness ratic_: Ideal p erson. In Questior,,airc- 2, suP.-
jects (65 experts and 2E laypc-rs ,ns) were aslEd to rate ho..; charp.ct ristic
each of 25S behaviors %.as of ar. "ideally" (a) intelliscrt persor, (: -
demically intelligent person, and (c) everydey intellicrnt perscn." Sutjc:ts
were again told to forr their o.;n i7-age of the "ideal."
3. C,.racteristicness r~tir.jS: Trait. In Questionnaire 3, sjt .ects
(28 la),pcr..cns) were asked to rate how characteristic each of ?fl tc,icrs
was of their "ideal" concept of (a) intelligence, (b) academic intelligence, e.-,
(c) everyday intelligence.
4. Characteristicness ratings: Self 5nd other. In Qucstio-",i ,-
(35 laypersons) were asked to rate how characteristic each of 2D caviors
was of (a) themselves and (b) that other adult whom they knevw cst.
To summarize, Questionnaire 1 dealt with ratings of the importance of
behaviors in defining an ideally intelligent person; Questionnaire 2 dealt ,.ith
ratings of the characteristicness of behaviors in an ideally intcllir 'nt rcrsc,
Questionnaire 3 dealt with ratings of the characteristicness of th., behaviors
in an ideal of each term as a trait; Questionnaire 4 dealt w.ith rnti'-l of
characteristicness of behaviors in oneself and in the other adtlt ui.,' k.: i.,.
Conceptions of Intelligencc15
The order in which ratings were made on each of the Ilow to 9Ohigh
questionnaire scales was counterbalanced across subjects. For examrple,
one subject might rate each behavior first for intelligence, thcn for acadc-ic
intelligence, and then for everyday intelligence, whereas another subject
might rate each behavior first for academic intelligence, then for everyday
intelligence, and then for intclligence.
Procedure
Laypersons filled out their questionnaire, self-rating, and IQ-test
materials in experinntal testing rooms at Yale. Experts filled out their
questionnaire materials at their own institutions and sent back: their forrs
by return rail. For the laypersons, the questionnaires were always ad-iis-
tered first, follow:ed by self-ratings, and then the IQ test.
kesults
Relations within and between Pctiros of Experts and Lavnersns
lable 3 shows correlatioi, %,itin and bet'een ratin;s of exp;rts ant
layperscs . Correlations are beti;Een subject, reans oi; ea:, quc ic' re itE-.
For exa-.ple, Correlation 1-4 is between experts' and layp.2rsons' , cstionni'e
response patterns for the attribu te, intelliccnce. These correlations addrc: _
several questions of interest. Since the reliabilities of the date upo-. t.r i
they arc LsEd are generally in the high .90s, it is possible to tale the
correlatiors at face value without concerns about atteniation due to unreli,.ilit .
Insert Table 3 about here
First, it is apparent that experts viev. intelligerce as ver) clusely
related behaviorally to both acadwiic and everyday intelligence; layporsons
view them as less closely related, especially in terms of the importance
of the behaviors to defining ideal persons. Experts see acadenic cnd
everyday intelligence as less closely related than is Intelligence to each
of academic and everyday intelligence, but again, the layperson, scp an
Conceptions of Intelliger,-ce
16
even i.eaker relationship. Clearly, both experts and laypersons distingui'h
between behaviors associated with academic intelligence and ones associated
with everyday intelligence.
Second, ratings of importance and of characteristicness show gererally,
similar trends, and were, in fact, highly correlatcd. mong the experts,
correlations between the two kinds of ratings were .96 for intelligence,
.95 for academic intelligence, and .93 for everyday intelligence. A-ong
the laypersons, the comparable respective correlations were .EO, .86, and .72.
Especially for the experts, then, there wes a very high degree of relatiorn:.i;
between rated importance of behaviors in defining an ideally intelligent
person and rated characteristicness of behaviors in such a persot. Such a
correlation is not a foregone conclusion: M4ost people would ac.ree, for ex:-le.
that "eating" is highly characteristic of intelligent people, Lt f - pecpi
would see eating as central to defining their conceptions of such pecple.
Third, the ratings of experts and laypersons for comparablc, kinds of
intelligence are quite highly correlated, with all. but one of six ccrrelatic s
ranging in the .80s. In each case (importcnce and characteristicress rati.;,
Lne correlation is highest for academic intelligence and lowest for everyd:y
intelligence, but given the srnall range of correlations, probably the nost
prudent conclusion would be that experts and laypersons see things very ruch
the same way, although not idcntically.
Structure and Contcnt of Peonlo's Conceptions of Intellirence
Laypersons. Table 4 shows the results of a factor analysis of layperscn's
ratings of characteristicness of behaviors in an "ideally intellicrt" perscr.
(Questionnaire 2). The factor analysis was done using correlation coeffici: t ':
as input into a principal components analysis followed by varirrax rotation of
the factorial axes. Because of the unwieldiness of the original set of 1%.
Conceptions of lntcellirfrce
17
intelligent behaviors as input to the final analysis, prelimir..-r,' factor
analyses were done in order to reduce the original set of 170 Lc!.,vicors to
a more tractable set of 98 behaviors that seemed to be the more central ones
to people's conceptions of intelligence.
Insert Table 4 about here
Three strong and interpretable factors emerged from the anz.1ysis of
ratings of the "ideally intelligent" person, accounting for 29's, lC'J, and 7%
of the variance in the data, for a total of 46',. These factors -.--- c l~ele-i
Upractical problem-solving atility," "vertbrl ability," and "social. ccetcrc.'"
Behaviors with loadings of .60 and above on these factors are listed in the table.
The first factor includes behaicrs such as "reasons logically ar-d :l"
Nidentifics connections a7-onQ ideas," and "sees all zspects of a prctIE-;" VbC
second f :tcr includzs behB~orz, such as 'sczks clearly ard artli.Al-10y ,"
"is ver~ily flueric," and "cc'wevrses well;" the thir d factor in~lu--Cs -%havi.-rs
such as "accepts others for %\hat they are," "admits r'istakes," aJ'ily
interest in the world at larqc." Althoug!; not every item loa~ir7 h i:;-ly c!"
every factor fits precisely with the assigned label, the sP"Se Of LZ.Ck factcr
does seE: consisteit with its assigned lablel; and the behaviors w;itlh the
strongest loadings were generally those most compatible with the la~'els.
Factor analyses were also. conducted on the ratings of acadc-ic and everi-
day intclli~ence. For acade7.ic intelligence, three strong and irter,'rEtatlc
factors c7-crged, accounting for 20*,, 81, and 7. of thE variance in. tr~e data
respectively, for a total of 35%'. The factors were labeled "verbal ability,'
'problem--solving ability," and "social cofn'oetence." Although *,'O t "- -dicrs
loading over .60 on these factors were of course not icotical to t:-.ce cn
the intelligence factors, they were highly similar. The problem-solving fzctcr
Conceptions of ]ntelliCcnze
18
seemed to have less of a practical orientation than did the analogous factor
in the analysis of intelligence, so that the "practical" prefix was not
placed in the label. For everyday intelligence, four strong and interpret:ble
factors emerged, accounting for 26.%, 10%, 8%, and 6% of the variance in the
data respectively, for a total of 50%. These factors were labled "practi:dl
problem-solving ability," "social competcnce," "character," and "interest
in learning and culture."
Several points are worthy of note in these data. First, the factors fcr
the three kinds of intelligence are highly overlapping, as wovld be ex .e:d
from the simple correlations of the responses, but are not idcntical. Each
set of factors shows a "slant" r-nsistent with the kind of intelligqcce it
underlies. Second, two kinds of factors cross-cut all three kin.s cf in-
telligence--problcm-solving and social ccmpttence. The first $hird of f
probably will come as no surprise to anyone, since proble,-solving ability.
would seem to be part of almost anyone's notions about the nature of intelli-
gence. The second kind of factor was something of a surprise, because soci.l
competence has played a relatively minor role in most theories uf intellir.:e.
Thorndike (1920) was among the first to propose that some kind of "sccial"
intelligence could be separated from what he referred to as "abstract" and
"concrete" intelligences, and social intelligence has played a role in the
theorizing of Guilford (1967; Guilford & Hoepfner, 1971) and of Wechsler
(1958); but most theorists of intelligence have ignored it, and indeed, a
review of the literature by Keating (1978) concluded that factor-analytic
studies had failed to demonstrate the existence of a social-intelliserce
factor. But, even if such a factor is largely missing frcm explicit thcori:v°;,
it Is obviously a salient element of laypersons' implicit theorizing. Third,
the first two (cognitive) factors constituting people's belief systc-I for
Conceptions of 1ntElli~c::.c
19
intelligence seem closely to resemble the two principal factors in Cattell
and Horn's theory of fluid and crystallized intelligence (Cattail, 1971;
Horn, 19GB). Fluid ability consists in large part of various kinds of
problem-solving skills, whereas crystallized ability consists in large part
of various kinds of verbal skills. Thus, the cognitive factors in people's
implicit theories seem quite closely to correspond to the coriLive factors
in one major explicit theory, that of Cattell and Horn.
Experts. Table 5 presents a factor analysis of experts' ratings of
the characteristicness of behaviors in an "ideally intelligent" perscn.
Because the behaviors that served as the input to the analysis ,cre provid:d'
by laypersons rather than experts, only those behaviors were rctained for
the analysis that were rated by the experts as being iipcrtant tc their
definiticns of the ideally intelliSent persun, wherc an "ipcKrl ,avicr
vias defined as one in the top third of the 1 - 9 importance scali (r-tir; of
6.33 and above).
Insert Table 5 about here
Three sizable and interpretable factors emerged in the exF-r'. ratir.;:
of characteristicness of behaviors. The factors accounted for 2> , 19", and
9% of the variance in the data respectively, for a total of 51.. The factcrs
were labeled "verbal intelligence," "problem-solving ability," and "practical
intelligence." The first factor included behaviors such as "displavs a ccz
vocabulary," "reads with high comprehension," and "displays curiosity;" tfI.
second factor included behaviors such as "able to apply knowled;c to problc-s
at hand," "makes good decisions," and "poses problems in an opti-,Ol w y;"
and the third factor included behaviors such as "sizes up situations well,"
"determines how to achieve goals," and "displays awareness to world around
him or her."
. ..... ' .# ' t
Conceptions of lntelli €;,ce
20
Comparable factor analyses were performed for academic and everyday
intelligence. For academic intelligence, three factors accountir; for
26W, 12%, and 9% of the variance in the data were labeled "problem-solving
ability," "verbal ability," and "motivation." This last factor was of
particular interest, since it has not appeared in previously discussed
analyses. Behaviors loading over .60 on this factor included "displays
dedication and motivation in chosen pursuits" (.78), "gets involved in
what he or she is doing" (.73), "studies hard" (.68), and "is persistent"
(.64). ror everyday intelligence, three factors accounting for 2C , 13 , ard
16% of the variance in the data were labeled "practical problem-solvin
ability," "practical adaptive bhavior," and "social competEeCC."
Several points need to be made about the factors that cmerged frc7m
these analy'ses. First, as w.as the case fcr laypersons, problc:;-sclvinc
ability is perceived as playing a major role in all three kinds of intel-
ligencc. Second, practical intelligence of some kind emerged in the factcrs
for Intelligence and everyday intelligence. Although these factors did
not have as clear a "social" orientation as was seen in the factors of the
laypersons, the experts, like the laypersons, perceived intelligence as
comprising quite a bit more than is presumably measured by IQ tests. Third,
a motivational factor emerged in the analysis of data for ratings regardirnZ
Intelligence. Although behaviors indicating high motivation appeared as
salient items in the data of the laypersons, they were distributed thrcu"ou.at
the factors and in no case formed a factor of their own. Finally, the first
two cognitive factors in the experts' conceptions of intelligence, like those
In the laypersons' conceptions, seemed to correspond closely to fluid and
crystallized abilities, whereas the third factor again seemed to represent sc-e
kind of practical or social adaptation. Thus, although there . diff-)-r:i.,
.. . , ' ,. " i
Conceptions of Intelli:rn:e
21
between the exact factor structures obtained for laypersons and experts, the
structures faithfully mirrored the high correlations between the two sets of
ratings in indicating remarkable similarities in perceptions between people
who male at least part of their living studying intelligence and people v;!-o
for th most part have no foiTmal training in psychology, much less the
specific field of intelligence.
Intercorrelations of Person Ratings and 1( Scores
As in the first experiment, we were interested in interreltions be-
tween people's ratings of thcmselves on intelligence, academic intcllisc-.:,
and ev ryday intelligence. This experiment had three features that enabled
us to go beyond the correlations in Experiment 1, however. The first was
that since ratings were on a percentile scale rather than a 1-9 scale, there
was at lezst a pissibility of greater precision in the rating:. Sc:ond,
since each layperson took an IQ tcst as yecll as making a self-r;.-in;, it
was possible to co..pare self-ratings with "objective" measurE-r.ns. Third,
subjects in one group (Questionnaire 4) were asked, to rate the otr ers :.
they knew best as well as themselves, so that it was possible to cc.pare one's
self-ratings to one's ratings :f another.
People's mean self-ratings on the percentile scale were 74 for intellierCe,
71 for academic intelligence, and 74 for everyday intelligence. A pcrcentilc
of 74 (for intelligence) corresponds to an IQ of 110 in the general po;.l,--cn.
In fact, the mean IQ of the subjects was 116, with a standard ci.viation cf
18 and a range froi 72 to 148. Thus, people's ratings representcd under-
estimates of their true abilities, relative to the general population agairst
which they were asked to compare themselves. The mean ratings of others#I
on the percentile scale were 76, 74, and 74 for intelligence, a: .- 'ic ir:li> .
and everyday intelligence respectively.
,w-.4,7 =77 7 I I
Conceptions of Intllig:r.:L
22
Intercorrelations of person ratings and IQ scores are prescnted in
Table 6 for those subjects who received Questionnaire 4 (the questionnaire
asking for behavioral charactcrizations of oneself and an other). Several
aspects of these correlations are worthy of ention.
Insert Table 6 about here
First, correlations of self-rated intelligence and academic intclligerce,
on the one hand, and everyday intelligence, on the other, were aot
equal, as was the case for the railroad-station sample in Experi.%rt 1.
Also as was the case in Experiment I , the correlation betvween acad;ic
an,' everyday intelligence %.as lower than either of the other t,,:o c'rrela-
tions (between intelligence, on the one hand, and acadz ic and everyday in-
telligcr.ce, or, the other). Sccoiid, the three kinds (, slf-r.t:1 intic,-:e
were also significantly correlated with IQ: People's conceptiorS of thE.-selves
were related to their objective test performance. The highest ccrrelatic" .it'
IQ was that for rated academic intelligence. Third, intercorreltions bet..c,
ratings of the other (for suLjects receiving Questionnaire 4) v%,rc Ic',.er
than intercorrelations between ratings of the self (for these sa-e s:iectK.
Apparently, subjects were less able to separate the three kinds of intellic-:ce
in themselves than in others, suggesting, perhaps, a halo effect in self-
perception of various kinds of intelligence.
As in Experiment 1, a multiple correlation was corputed between self-
rated intelligence, on the one hand, and academic and everyday intelligence, cn
the other. The multiple correlation was .69 (p < .001), with regression ccffi-
cients for academic and everyday intelligence of .34 and .38. In this exFcri-cr, i
I;- -- .
Conceptions of Irtellig.ce
23
was also possible to regress IQ on the self-ratings of academic and everyd:y
intelligence, as well as to regress one's rating of the other's irtelligerce
on one's ratings of the other's academic and everyday intelligcnce. The
multiple correlation of IQ with the two self-ratings was .38 (p e .001),
with regression coefficients of .28 and .15 for acadcmic and evcr ,y
intelligence. The miultiple correlation of the other's rated irtelligence
with the two other-ratings was .48 (p< .001), with regression wci!,ts of
.08 and .40 for academic and everyday intelligence. It is of sc intercst
to note that the weights of academic and everyday intelligence in ;redictir.;
self-rating were about equal, whereas these weights were very urL:lanced
(with the everyday wcight much highei than the academic ce) in .-Edictir.
other-rating.
CorrelE'icrs of Factcr Scorcs with IQ ard S f-R.tings
Factor scores were computed for those subjects who received (jestionraire
4, the qjestionnaire asking for behavioral characterizations of cr.eself an
an other. The factor scores represented subjects''characterizaticns of their
own behaviors on those items loading highly on each of the factors of each
kind of intelligence, where the factors were those defined earlier. Factor
Scores were computed using an approximation technique whereby each item
(behavior) loading .60 or over on a given factor was unit-w.eig'.ted in the
computation of the score; all other items (behaviors) were weighted zero.
Table 7 shows simple and multiple correlations between subjects' factor scores
on each of the three kinds of intelligence, on the one hand, and IQ and self-
ratings, on the other.
Insert Table 7 about here
Prediction of IQ and of self-ratings of everyday intelliecnce werc not
Conceptions of Intellire.:
24
very successful. Only one (of ten) factor scores provided sisjnificant prodt
tion in each case. Predictions of self-ratings of intelligence and of ac6.t-1:
Intelligence were successful, however. There were six (of ten) significant
simple correlations in the prediction of self-rated intelligence, and fc.r f -
significant correlations in the prediction of self-rated acadc7ic intelli:: .
All but one of these ten (six plus four) significant correltions wEre for
ucognitive" factors such as practical problem-solving ability arid verbal
ability. Thus, people's ratings of the extent to which cognitive types of
behaviors associated with intelligence and academic intelligcr~e c .2rc:cr;:
themselves were related to these people's ratings of their o:r, irtclliccnce
and acadcmic intelligence. In -;me cases, it was possible to cc- tine tnesc
clusters of self-ratings, as represented by factor scores, to yicld r2t -ct-
good predciction of self-ratirgs via multiple regression. For e-a;Ae, te
multiple correlation of self-rated intelligence with the three factor scorcs
obtained for intelligence was .55 (as shown in the table).
Formrtion and Properties of "Prototypicclitv" Peasures
Consider for Questionnaire 2--the questionnaire asking subjects te
rate the characteristicness of each of a set of 250 behaviors in an "idLall,
intelligent, academically intelligent, and everyday intelligent person--
the meaning of a mean response pattern averaged over subjects, -iothcr
these subjects be experts or laypersons. One might view a mcan response
pattern as representing an approximation to the population's prctoty;e for
what constitutes an ideally intelligent, academically intelligent, or
everyday intelligent person. On the basis of the data collected in this
study, it would be possible to form three such prototypes for e\;-rts (cn'.
for each type of intelligence) and three such prototypes for lii)'ersons.
Conceptions uf Intellinc-nce25
We did, in fact, form such prototypical response patterns, and, as we su:w
earlier, they were highly correlated between experts and laylicirsUns.
Suppose one were to take each individual subject's response pattern
for Questionnaire 4--the questionnaire asking subjects to rate the charac-
teristicness of each of the 2SO behaviors for his or her own behavior (as
well as that of an other)--and correlate this individual respor,se pattern
with thQ prototypical response pattern (as obtained from different subjcc:s
filling out a different questionnaire, na-,ely, Questionnaire 2). One mi;.,t
view the correlation between the individual's response pattern and the pro-
totypical response pattern as measuring the degree to which a giver, subjczt
resc.blcs thE prctetype of an intelligei1 person. In effect, we L,,ve a
"rese-;blance" measure of intelligence, based upon a comparison bctv..ecn in-
dividudls' self-descriptions and others' descriptions of an ideal: Hig,..r
scores represent closer resc.:,blance betecii the individjal and tc pro:
We co> T-tc" tK cerrelatie -, bctween the self-ratings and protc:..;cs, b si:-
the correlations only on thc 170 behaviors that were intelliocrt (as
to unintelligent). In this %ay, we obtained measures of protcty;icalit.., f,
each suLjcct receiving Questionnaire 4 on intelligence, acade.ic i.telli:-,
and everyday intclligence. In each case, %.e correlated the pcrtir.-rt i-.i-
vidual response pattern (for intelligence, academic intelligence, or ev.rL.
intelligence) with the corresponding prototypical response pattern.
Properties of the proto.y,icality measures are reported in Table 6. I!is
particular set of correlations used the experts' prototypes as thc bcsis f. :.
comparison. As w.,:ld be cxp:ctc. fro- the high correlations b-.. c'; t'.
data of the experts and the laypersons, practically idenitical results %.ere c -
tained using correlations with the laypersons' prototypes. At the top of
Insert Table 8 about here
Conceptions of Intellig.-rce
26
the table are"the mean, standard deviation, and range of each of the trce
measures, where each measure is a correlation coefficient (c.-puted for each
Individual subject) with a potential range from -1 to 1. These statistics
show that the average correlation bet%.:een the response pattcrn of each
individual subject and of the prototypically intelligent p'.rson was si'..ifi-
cantly different from zero for each of the three kinds of irtclligence. On
the average, people saw theomsclves as having a moderate degree cf rese-blance
to each of the prototypes. The range in degrees of resemblrce was quite
large, although, as is sho',.:n in the last column at the top, there were no
individuals with nontrivially negative rescmblances to the prototype. At
the bottom of the table are correlat:ons between the prototpiczlity rre:sures
and each of IQ and self-rated intelligcnce, academic intelligc-:.:c, and very-
day intelligznce. The corrcltions betv;ezn the prot typicalit' _aSure ai,
IQ were both statistically significant and substantial for each of the three
kinds of intelligence, with academic intelligence sho;.;ing the crctest rcla-
tionship and everyday intelligence the smallest relationship. These ccrrcla-
tions, which are as high as or nigher than correlations typically obtain..
between cognitive measures ard psychometric tests (see, e.g., hunt, Frcst,
& Lunneborg, 1973; Hunt, Lunneborg, & Lewis, 1975; Sternberg, 1977, l92.)
indicate that it is possible to obtain a rather good estimate of IQ on thc
basis of the correspondence between a person's self-perceived pattern of
behaviors and the pattern of behaviors in an "ideal" person. Five of to
six correlations with self-ratings of intelligence and academic intelliterce
were also statistically significant, although lower than the correlatior:s with
IQ, meaning that subject's self-descripticns better predicted their obj-c~i.cl,
measured intelligence than their subjectively rated intelligence. None cf the
correlations with everyday intelligence were significant. Here, as previct:sly,
40 1.
Concepticns of Intellic.c'..
27
self-ratings of everyday intelligence prove to be harder to predict then
self-ratings of academic and everyday intelligence.
Conclusions
To conclude, people do appear to have prototypes corrcspo,,ding to
different kinds of intelligence. These prototypes are very si-ilar, bit
not Identical, between experts and laypersons. The prototy;,.s are orga-
nized into sensible factors of behavior, such as "practical prc'lEm solvin.
ability," "verbdl ability," and "social competence." One's self-perceived
standing on the "cognitive" factors is predictive of one's scl -ratin c
intelligence and academic intelligence. Moreover, one's self-porceived
correspondence to the prototype!s for each of the three kinds of intelligence
is predictive of one's IQ and of one's self-ratings of intelligence ar.d
academic intelligence. Self-ratings of academic and everyday intellic 7.:e
are highly predictive of sclf-ratings of intelligence (overall), tut rtir7;s
of an othcr's academic and everyday intelligence are less 1icl'ly (alt':
still significantly) related to ratings of that other's overall intelli; :o.
It also appears that one is less able to dissociate various Linds cf ir.tcili-
gence in oneself than in otl~rs. In su., people do appear to h:ve pro't
for an "ideally" intelligent person, and their self-perceiveJ corresp:zdcc:es
to these prototypes are associated with their estimates of their intelli:csce,
as well as their measured intelligence. One question that still needs to
be answered is whether people use these prototypes in infor-al, everydZay
evaluations of the intelligence of others. The third experi:cnt was intur>c
to address this question.
. . .. . . . . . .
Conceptions of IntcloiH:-ce
28
EXPERIIIE14T 3
In this experiment, we sought to ascertain the extent to which people
actually use the behaviors associated with intelligence ard unintelli~ence
In their evaluations of other people's intelligence, in particular, wier,
they are presented with written behzvioral descriptions of otters.
Method
Subjects
A questionnaire was sent to 168 persons selected at rand-" from a Nc ,.
Haven area phone book. Of these persons, 65 responded in ti-e for their data
to be. used in the study. Twelve persons responded too late for their data
to b2 included in the study. All data were sent to us by return -ail.
Materials
The principal expericntal material was a SO-item questiv!:irc. Each
itci consisted of a verbal description of behaviors charactcri-irg sc-,
particular person. People were told that they would "find bricf descriptions
of different people, listing various characteristirs they have. Assu':.e
that the list for each person is made of characteristics that teachers h!ve
•'.-pplied to describe that pcrson as accurately as possible." Th e subject's
task was to "read the characteristics for each person and then to rate each
person on how intelligent" the subject considered the person to be. Ratin;s
were made on a I to 9 scale, where 1 was labeled on the scale os "not at
all intelligent," 5 was labeled as "average intelligent," and 9 was latelcd
"extremely intelligent." Half of the items on the questionmaire presented
unquantified behavioral descriptions (e.g., "She converses well") and half
presented a mixture of quantified and unqjantified descriptions (e.g., "ShE
often converses well"). In fact, the ite.ns in the two halves of the qucs-
tionnaire were identical except for the presence of quantification in one
half of the items and its absence in the other half. Half of tle doscri~tic'.s
29
were paired with male names and half with female names. A given description
was paired half the time with a male name and half the tir:je with a fL: l E
name (across subjects). Typical descriptions of people invrcntcd for E.i;;:rircnt
3 are shown in Table 9. Each item was chosen for the questionnaire so Ls to be
representative of items loading highly (.60 )r over) on one of the thrc.j
factors of "intelligence" identified in Experiment 2, or of behaviors i-r, i-
fied in Experiment I as unintelligent. (Academic and everyJy intelliyc-:.:c ,.ere.
not dealt with in this experirt, nt.) For cxample, one of the chosen bc.L. .Krs
"keeps an open mind," was chosen because of its high loadir (.73) on FtcTcr I,
Practical Proble.m-Solving Ability. Thirty-six different behaviors were ';-
including 24 positive ones (f for each factor) and 12 negative or~S. E&-:.
behavioral description could consist of from 4 to 8 state.ents. tost be' -.' , ioral
descriptions contained (randomly ordered) mixtures of intelli,,r t arn "r.i.e-li
gent behaviors, although somc descriptions contained only one of theLc
of behaviors. The fictitious persons thus covered a ran e-cf le.c 's c.-'i '-'i
Insert Table 9 about here
Des ign
All subjects received the same questionnaire items. The various Q..ier-
naires differed froi. one another only in that (a) half of the subjclts rc:civcd
quantified ite-s presented before unquantified ones; the other sut-jects rcceivd
reverse ordering; (b) different pairings of names with descriptions of r:cple werc
used for different subjects, with the constraint that a givcn dnscriptic, L.--.
paired half of the time with a male name and half of the tine .-ith a fc:le r.-e; e
(c) different random, orders of items .,ere given to each su'je.t, ith tL co.sr,'
that all quantified items and all unquantified items be blocked together.
Procedure
Subjects viere told in a letter sent to their homes that they would receive
$5 by return mail if they sent in the accompanying questionnaire anon)--ously and
sent in a separate verification fomn indicating that they had, in fact, riturr,'.
their questionnaire.
* '.1 . ' - m.
Concepti oos of IItcli .nrce
30
Results
Basic Statistics
The mean rating of intelligence over the 45 unquantified descriptions
was 5.09 (on the 1-9 scale); the mean rating over the 45 quantified dc:crip-
tions was 4.49. The difference between means was highly significant, t%-)
12.77, 2( .001, indicating that quantification generally lo:ered ratirs
of intelligence. Such a result would be expected, since quLntificatio,
amounted to qualification of the statements that were made. The correlaticn
between the unquantified staterients and their paired quantified versicrs
was .87, indicating that although quantification lowered ratings, it char;Ad
their pattern only slightly.
It made no difference in means whether a given descrirtien was paired
with a male iire or with a female name. The means in the unc.untifiCd
con dition were 4.49 for both male and female descriptions; the means i:
the quantified condition were 5.09 for male descriptions and 5.C3 fcr fcL-le
descriptions. Moreover, the correlation between identical descripticns
paired for male versus female names was .99 in the unquantificd conditics
and .93 in the quantified cornition.
The reliabilities of the data were very high: Coefficient alpha for
all random split halves of subjects was .91 for the ratings of unquantificd
descriptions and .92 for the ratings of quantified descriptions.
Modeling of Data
The main data of interest in this experiment were those deriving frc1
linear modeling of the ratings of intelligence. Modeling was done both for
unquantified and for quantified descriptions.
Unquantified descriptions. Two basic kinds of modeling were done. In
the first, we took means and sums of characteristicness ratings fron cxpcrts
..-, . I -- '
Conceptions of Int, liS:-:e
31
and laypersons answering Questionnaire 2 from Expcriment 2 (the que-stior.,.ire
asking for ratings of how characteristic each bVhavior is of an idcally intel-
llgcnt person), computing these means and sums on the basis of those Le-
haviors listed in each description given in the present exFri'.nt. .s ,re
rescaled to have a theoretical mean of 0 (by subtracting t frc-i each . .).
Thus, we obtained for each description a mean nd rescalcu . of ch:.t:r-
isticness ratings for the behaviors listed in ttit descripticn, both fc, ev;--
and 1,ypersons. The correlation bet .'e&n ratings of intelli;cnce and - -e- :, C -
acteristicncss ratings for each fictitic:: s pcrsor ,ere .£C .for e r- a--
for layperscns; the correlations between ratings of intellicErt-, ar.d rc:c.
su--ed characteristicness ratings for each fictitious per:c,. , .9" fr (A-
perts and .9C for layperscrs. Hence, rc;:rdliss of wFc:rJ,. .L.;crt or .
dat. were usc.!, tne means and su-s of the char2, . risticr,:,
vided excellent prediction of pecple's overall evaluations of the i
of the fictitious persons v:ho were described in the brief r.Lrrti'e ; .
Length of description, incidentally, correlated o, ,y trivilly. Vith t' ,"
In the sccor-d kind of .cdeling, multiple rcgresion wzs use tc ,-i::
the ovcrall rting of the intelligence cf the fictitious p~r -c;, fr- cc.ts
of the number of behaviors in each of the factors of intelligrce (art "'c
behaviors characterizing unintelligence) that w.rc found in each pass-:.
For example, if a given passage had one behavior listed fro- "practicl
proble--solving ability," two behaviors listed fr, - "verbal altlitv," c"-'
behavior listed from "social competence," and t%..o behaviors listed frc-
the unintelligent behaviors, then the independent variables entered in'c t':.,
regression would have been 1, 2, 1, and 2 respectively. The nultiple czrr.-
tion between the ratings of the intelligence of the fictitiois person, c' t.
one hand, and the aspects of perceived intellic-nce and unintcllicrc
the other, was .97. Regression weights were .32 for "practical pro!,-
Conceptioiis of Intelligcnce
32
solving ability," .33 for "verbal ability," .19 for "social cc:ipetence," arid
-.48 for unintelligence. All weights were sigificant and all signs were in theA
directions, with only the unintelligent behaviors showing a negative
weight. The unintelligent behaviors had the highest regression weight,
as might be expected, given that there was only one independent variable
for such behaviors, as opposed to three for intelligent behaviors; morc-
over, as anyone who has read letters of recom-.endation kno;;s, even one
negative comment can carry quite a bit of weight. Of the three kinds of
intelligent behaviors, the two cognitive kinds (practical prole.-,-solving
ability and verbal ability) carried about equal weight, and the noncornitive
kind (social competence) carried less ,eight. These relative weichtirngs
were consistent with those obtained in Experiment 2 in the prediction coF cne's
ratings of one's own intelliOEhce: The cognitive factors w.re ri htCJ rore
heavily than the noncognitive ones.
Quantified descriptions. These descriptions were modified by the
adverbs "always," "often," "sometimes," and "never," or by no adverb at all.
When a given description was of an intelligent behavior, a priori weig'ts
of 3. 2.5, 2, .5, and -2 were assigned to "always," no adveiv, "often,"
"sometimes," and "never respectively. When a given description was of an
unintelligent behavior, a priori weights of -3, -2.5, -2, and -1 were
assigned to "always," no adverb, "often," and "sometimes" respectively; t'.e
adverb "never" was not used to modify an), unintelligent behaviors, because
of the confusion that might be engendered in the case of double negatives
(e.g., "He never fails to ask questions" seemed a bit confusing). The weights
were chosen on the basis of an informal survey of colleagues rather th:r ,
parancter estimation in order to keep the number of parameters that nccJ- to
be estimated to a reasonable size.
L1 __Wr -"
Conceptions of Intelligcnce
33
The same two kinds of analyses were performed on the data for quanti-
fied descriptions as were performed on the data for unquantified descrip-
tions. The unweighted independent variables described earlier for the
unquantified descriptions were multiplied by the weights ap;-ropriatc, to
them to yield new, weighted independent variables for the quantified
descriptions. The correlation between ratings of intelligence and tte rean
characteristicness ratings for each fictitious person were .97 for experts
and .96 for laypersons; the correlations between ratings of irtelligence and
rescaled summed characteristicness ratings for each fictitio, s perscn v-re
.9G for both experts and laypersons. Hence, prediction for the quantified
descriplions was comparable to that for the unquantified deS:ripticns. A-ain,
length of description was only trivially corrclated with evaluations.
The multiple correlation between the evaluations,on the one h 'ad, an t
three factors plus unintelligence, on the other, was .95. Rec-ression .i-hts
were .37 for "practical problem-solving ability," .48 for "verbal abilti:y,"
.20 for "social competence," and -.32 for unintelligence. A.gain, all ci;,ts
were significant and in the expected direction. Also again, t!.e wei,,s fer
the cognitive factors were grcdter than the weight for the noncogniti~e cr_,
although in this data set, the weight on the unintelligence variable ._s rela-
tively lower than in the previous data set.
Conclusions
People use their implicit theories of intelligence in evaluating the
intelligence of others as well as of themselves. Their evaluations of
others, based on relatively brief behavioral descriptions of these others,
can be predicted at a high level on the basis of their inplicit thcories.
As in the self-ratings, people seem to weigh cognitive factors more h, 'il,
than noncognitive ones, and to take into account negative as well as pDsiti. e
-- "
Conceptions of Intelligcn:e
34
information. The implicit theories of experts and of laypersons are similar
enough so that it makes little difference which is used in predictions: Re-
sults are almost identical for each. In sum, knowledge of a person's
implicit theory can be used to predict that person's evaluztions of both
him or herself and others.
General Discussion
People have well-developed implicit theories of intelliqcnce that tftey
use both in self-evaluation and in the evaluation of others. Althcj( it .re
are differences in these theories across groups, there seems to be a cc--Dn
core that is found in the belief systems of individuals in all of the groups
we studied. The corrnon core includes some kind of problem-solvin factsr,
some kind of verbal-ability factor, and some kind of social-cc'7petencc fcc: 'r.
A recent review of litcrztures covering different appro:chZ to .-
standing intelligence, including the present one as well as the psych:--:tric,
Information-processing, and mental-retardation approaches, con:ludes tl-
these three aspects of intelligence, plus a motivational one (v.hich did,
In fact, appear as a factor in the experts' ratings of acadc-ic intellii-zz,)
bee- to emerge from a variety of approaches to intelligence (Stermmerg, in ; -c'.
* Thus, the results of the present research seem to converge with resear:h
of other kinds in suggesting that intelligence is found to corise certain
kinds of behaviors almost without regard to the way in which it is stu.lid.
These behaviors include (am:ong possible others) problem solvin;, veral f:ili:..,
social competence, and possibly motivation.
In particular, problem solving (or fluid ability) and verbal facility
(or crystallized ability) see:m to be integral asrects of intelligcn: fr.c-
tioning. These abilities can be identified by both correlational rreins
(Cattell, 1971; Horn, 1968) and experimental means (Sternhern, 19 Ga .
L ,..
- \.,: "',.. , -, ..... .. - ;' - . ..r
Conceptior of Iitcll. -e
35
191Oc ). In psychometric investigations, fluid ability is best reasuc . by
tests of abstract reasoning and problem solving, such as abstract aralcies,
classifications, series completions, and the like. Verbal itcms are also
useful if their vocabulary level is kept low. Crystallized zbility is tc2st
measured by tests that require for their performance the products of accjl-
turation: vocabulary, reading ccmprehonsion, general informztion, and " e
like. In information-processing terms, crystallized ability seems test tc
separate the products of acquisition, retention, and transfer of verLal
materials. These tests measure primarily outcomes of previously execU
cognitive processes, rather than of current execution of these processes.
The vocabulary that is measured by a vocabulary test, for exp-ple, rav h,,e
been acquired years ago. Fluid ability tests, on the other hand, se:7. t-ast
to separate the execution of component processes of reasonin- and prc~l-
solv-ir,. They n1easure pri:; rily current rather than past Fc rr :r -.
irproving the functioning of rildly or moderately retarde " irividuels,
it seems necessary to conduct training in both the acquisition, retu.ior.,
and transfer skills that lead to thE development of cryst-lli: abi1i'
(e.g., Belmont & Butterfield, 1971; Carpione & Bro',,'n, 1977) ad in t.c
reasoning and problem-solving skills that constitute fluid ability (e.9.
Feuerstein, 1979). Motivational intervention may be needed, too CZi.-:', ,
Implicit and explicit theories of intelligence are actually theories
of different things. Implicit theories tell us about people's vie- s of
what intelligence is. They are theories of word usage, and in the case of
"intelligence," the word is one of interest to a large number ard variet.'
of people. Explicit theories tell us (we hope) .:hat intellicence is; in
real life, it is more likely they tell us what some aspect of intelli-,,ce is.
None of the currently available explicit theories seem to do justice to t'e
Conceptions of Intelligenc e
36
full scope of intelligence, broadly defined. Perhaps no one theory o.cr
could, whether the theory is implicit or explicit. But thcory-constrjction
has to start somewhere, and in the course of scientific evolution, it se.ms
that implicit theories of experts give rise to the explicit theories of
these experts, which are in turn tested on objective behavicrJL data. Le-
cause of this developmental relationship betwcen implicit and explicit
theories, there is almost certainly going to he considerable overlap :-te_2r
them. We believe that a study of this overlap, as well of the overlap e::, g
theories of each of the tao kinds, can inform and strenqth..r, b&th .ir:' cf
theories and research. The kind of "prototypical analysis" pcrfc--e h-rc
see,.s tu be a useful complement to the kinds of "co-pon-nti;.1 ar.lsi .
other forms of analysis that have been conducted in laboratory znalysc.: cf
intellectual functioning.
Conceptions of Intclli;cnce
37
Reference Note
I. Cantor, N. Prototypicality and personality jud nents. Unpublished doctorz"
dissertation, Department of Psychology, Stanford University, 197'.
A1
I
Conceptions of Intelligence
38
References
Belmont, J. M., & Butterfield, E. C. Learning strategies as determinants of
mejioory deficiencies. Cognitive Psvcholool, 1971, 2, 411-420.
Bruner, J. S., Shapiro, D., & Tagiuri, R. The meaning of traits in
isolation and in combination. In R. Tagiuri & L. Petrollo (Eds.),
Person pcirception and interpersonal behavior. Stanford, Californi::
Stanford University Press, 1958.
Campione, J. C., & Bro%:n, A. L. Memory and mcteremory development in Cevzable
retarded children. In R. V. Kail, Jr., & J. W. Hagen (Eds.), Pers...ciives
on the development of me--or, ard cc-)ritior. HillsdalE, N.J. : ErlR:-.r, 1977.
Carroll, J. B. Psychometric tests as cognitive tasks: A ne,.: "structure, of
intellect." In L. B. Resnick (Ed.) The nature of intMeli- crce. !:i11 lde,
N.J. : Erltaum, 197G.
Cattell, P. B. Abilities: their structure, aro,.th, and actir.. Bostc -:
,, g hgtnr,-N'ifflin, 1971.
Ferguson, G. A. On learning and human ability. Canadian Jcijr- of P. C-,
1954, 8, 95-112.
Feuerstein, R. Instrumental enrich-ent: An intervention pr!ro7. for cc-itive
m. difiailitv. Baltimore: University Park Press, 1979.
Guilford, J. P. The nature of intelligence. New York: McGra,:.-Hill, Il' 7.
Guilford, J. P., & Hoepfner, R. The analysis of intelligence. Ne.. Yor":
McGraw-Hill, 1971.
Horn, J. L. Organization of abilities and the development of intelli~cr.ce.
Psychological Review, 1968, 75, 242-259.
Hunt, E. B., Frost, N., & Lunncborg, C. Individual differences in co Ti'ien:
A new approach to intelligence. In G. Cower (Ed.), The psychol-.v of
learning and motivation (Vol. 7). New Yor.: Academic Press, 197".
Conceptiuns of IntclliS..:e
39
Hunt, E., Lunneborg, C., & Lewis, J. What does it mean to k' high u L
Cognitive Psychology, 1975, 7, 194-227.
Intelligence and its measurement. Journal of Educational P,.choloiy, i
12, 123-147, 195-216, 271-275.
Keating, D. P. A search for social intelligence. Journal of Fducati-r, I
Psycholof'y, 1978, 70, 21-223.
Neisser, U. The concept of intelliScnce. In R. J. Sternbcr, & D. Y. . -r
(Eds.), Human intcllioence: Perspectives on its theory ard r.e:su-co-:.
Norv'ood, N.J.: Ablex, 1979.
Pellegrino, J. W., & Glaser, R. Components of inductive reasoning. I r P..
Snow, P.-A. Federico, & 14. Montague (Eds.), Aptitude le rnc, _ .... .. ..
Cognitive process analysis (Vol. 1). Hillsdale, N.J.: Erlb., L I
Pi.get, J. The psychol c 'v of intellic-rce. Totom..,a New Jers .e': LitflJ>"
Adams, 1972.
Resnick, L. B. (Ed.) The nature of irtelligence. Hillsdale, n.J.: Er, , t
Siegler, R. S., & Richards, D. D. Tlh development of intelli-,:rce. I 2.
berg (Ed.), handboc" cf hm-n irtcl ier.ce. e.,. York: C, ri :c '.
Press, in press.
Sro,.,, R. E. Theory and method for research on aptitude processes. Ir.'....
1978, 2, 225-278.
Sternberg, R. J. Intelli ence, infor:tation p rocessin3, and a rlocic -I :-.
ThCo-Porcrtial analvsis of huan abilities. Hillsdale, N,.".: Erl -- ,
Stenberg, R. J. The nature of mental abilities. f.-erican,,hL." i", '";:,
34, 214-230.
Sternberg, R. J. Factor theories of intelligence are all righ, alrc. st. .. i
tional Researcher, 19.0, 9, 6-13, IS. (a)
Sternberg, R. J. Representdtion and process in linear syllogistic reas.>
Conceptiowu of Intilli5Ct'i1,(-
40
Journal of Experimental Psychologv: General, 1980, 109, I19-150. (b)
Sternberg, R. J. Sketch of a componential suhthccry of hurar, intcllisCr.-:e.
Behavioral and Brain Sciences, 1930, 3, 573-584. (c)
Sternberg, R. J. The nature of intelligence. Now York Universitc., Edcs1ic
_Quarterly, in press.
Thorndike, E. L. Intelligence and its uses. Iarper's K'._irc, 1920, l,
227-235.
Thorndike, E. L. The measure;v.ent of intelligence: Present status. ..,'-c
Review;, 1924, 31, 219-252.
Wechsler, D. The neasuremert and epprzisal of adult intelli:E,-rc (" iti
Baltimore: Williams E, Wilkins, 1958.
Wober, M. Towards an understanding of the Kicanda concept of intelli:r,-.
In J. V!. berry & P. R. Dzsen (Eds.), Culture and cc 'niti'r-: R - ft
cross-cultLral psycholc-,. London: ,eth-ien, 1974.
Yussen, S. R., & Kane, P. Children's concept of intelligence. In S. R.
Yussen (Ed.), The grov:th of insi.ht in children. Net., Yor;,: Acad ic
Press, in press.
Lgler, E. The retarded child as a wvhole person. In H. E. Ad;-.s 5 ,. L.
Boardman (Eds.), Advances in experimental clinical psvch,lcyv (Vol. 1).
New York: Pergamon Press, 1971.
Conception- of Intclli'.n:t
Footnotes
Portions of this article were presented at the annual mectirgs of the
Psychonomic Society, Phoenix, 1979, and of the American Psychological Assccia-
tion, Nontrcal, 1980. We are grateful to Ulric tNeisser, %.hozc thin0ir. tas
helped shape ours, and to Elizabcth Charles, for valuable assistarce il. all
phases of this project. The research reported in this article, was fur,'cd
by a grant from the Department of Psychology, Yale University, and by Office
of N(aval Research Contract IN0001l47C0021 to Rc.ort J. Stern.corj. kocosts
for reprints shDuld be sent to Robert J. Sternberg, Departrent of Psy:Klo:,
Yale University, Box 11A Yale Station, New Haven, Connecticut 05520. Tre
master list of 250 behaviors will also be sent upon request.
IWe were chagrined by the fact that our scr,,ple, which excluded s
and :as based upon responses of people %-;ho ans,,crcd nesp q.r adveri: ts,
had an averagc. IQ one standard de.,'iation above the general po xjlatic- .
We, like others, have found it much easier to obtain specialized Fc{,,:icns
than to obtain the elusive "average" one.2Various informaticn-processin vie,.,'s on and approaches to fluid a-.
crystallized abilities can be found in Carroll (1976), Pellegrino and GIAser
( .1900 ), Snow (1978), and Sternberg (1979). See also Resnick (197) for
a variety of contemporary views on the nature of intelligence.
Conceptions of 1njelli9.ne
42
Table 1
Correlations between Frequencies of Listed Behaviors
Academic Everyday
Intelliognce Intcllir.:e,Li brary '
Intelligence .24* .12
Acadc.ic Intelligence ... .240,,
Everyday Intelligence
RailroadIntellience .10
Acade ,ic Intelligence ---
Everyday Intelli.nce .
Superr:_arkct
Intel 1 icence --- .05
Academic Intelligence .1l"
Everyday Intelligence
Overall
Intel Ii gence- .38**.
Academic Intelligence --- .3L
Everyday Intelligence
Note: Correlations are based on frequencies for the 170 intelli;cnt b
(** <.001
Conceptions of Intclligence
43
Table 2
Correlations between Self-Ratings
Academic EverydayIntelligence
Intelligence IntelligenceLibrary
Intelligence .80" * .42***
Acadcmic Intelligence --- .28*
Everyday Intelligence
Railroad
Intelligence .73t** .74**
Academic Intelligence
Everyday Intelligence ---
Supermnarket
Intelligerze .83**.
Acade-ic Intelligence --- .41***
Everyday Intelligence
overal l
Intelligence .80*** .60**
Academic Intelligence --- .44**
Everyday Intelligence ---
Note: Correlations are based on self-ratings for 61 collep library sbjccts.
63 train station subjects, and 62 supermarket subjects.
< .05
**R'Z.0"*4*£ .. 001
- ' '~~.< €
Conceptions of 1ise. :
44
Table 3
Correlations within and between Ratings of Experts and Laypcrsoni
Experts . Laypersons
1 2 3 4 5 6Int. Ac. Ev. Int. Ac. Ev.
mport"ncc Ratin-+s
I Intelligence 1.00 .90 .90 .80 .75
Experts 2 Acadc:ic Int. 1.00 .67 .69 .84 .28
3 Everyday Int. 1.00 .73 .52 .72
4 Intelligence 1.00 .81 .76Lay-
5 Academic Int. 1.00 .35persons
6 Everyday Int. 1.00
Ch.racteristicness P:tinrrs
I Intelligence 1.00 .83 .84 .82 .68 .9
Experts 2 Academic Int. 1.00 .46 .72 .89 .43
3 Everyday Int. 1.00 .69 .31 .81
4 Intelligence 1.00 .75 .86Lay-
5 Academic Int. 1.00 .45persons
6 Everyday Int. 1.00
Note: Correlations are based upon Questionnaire I (ir-'ortance of t -.,icrs
In defining conception of ideally intelligent person) and Questicn~zire 2
(characteristicness of behaviors in repertoire of ideally intellicct ;:-'
Correlations are between subject means on each questionnaire itc rc-
example, Correlation 1-4 is between experts' and laypersons' qu, :iov:
respnnse patterns for the attribute, intr linence.
L~
Conceptions of Intcig:-,.
45
Table 4
Factors Underlying People's Conceptions of Intelligence:
Laypersons Rating Characteristicness of Behaviors in "Ideal" P..-rsc- ,
I. Practical Problem-solving Ability
1. Reasons logically and well .77
2. Identifies connections among ideas .77
3. Sees all aspects of a prcblcm .76
4. Keeps an open mind .73
S. Responds thoughtfully to others' ideas .70
6. Sizes up situations well .69
7. Gets to the heart of prob1cs .69
8. Interprets inforr;ation accuirately .6
9. Makes good decisions .65
10. Goes to original sources fCr b:sic illfe7ation
11. Poses ,roble-,s in an optimal way .62
12. Is a good source of ideas ,E2
13. Perceives implied assumpticns and conclusions .62
14. Listens to all sides of an argument .61
15. Deals with problems resourcefully .61
I. Verbal Ability
1. Speaks clearly and articulately .83
2. Is verbally fluent .2
3. Converses well .76
4. Is knoqledgeable about a particular field of knowledge .74
5. Studies hard .70
6. Reads with high comprehension .70
7. Reads widely .69
S. Deals effectively with people .68
9. Writes without difficulty .65
10. Sets aside tine for re3dinl J4
11. Displays a good vocabulary .61
12. Accepts social norms .1
13. Tries new things
Conceptions of Irt,1igcc
Table 4 (Contd.)
Factor Lcf- ir-
Ill. Social Competence
1. Accepts others for what they are C8
2. Admits mistakes .74
3. Displays interest in the world at large .72
4. Is on tire for appointrents .71
5. Has social conscience .70
6. Thinks bcfore speaking and doing .70
7. Displays curiosity .68
8. Does npt make snap jud-ents 8
9. l',akes fair judgments .65
10. Assesses well the relevance of informtion to a problc-
at hand .66
11. Is sensitive to other people's needs and desires E5
12. Is frank and honest with self and othcrs El
13. Displys interest in the ir.-iciate environment
Conceptions of Intelligence
Table 5 41
Factors Underlying People's Conceptions of Intellig-nce:
Experts Rating Characteristicness of "Important" Behaviors in "Ideal" Person
Factor Loadirg
I. Verbal Intelligence
1. Displays a good vocaulIary .74
2. Reads with high comprehension .74
3. Displays curiosity .68
4. Is intellectually curious .66
5. Sees all aspects of a problem .66
6. Learns rapidly .65
7. Appreciates kno,:led;- for it, own sake .65
8. Is verbally fluent .65
9. Lister.s to all sides of an arginent before deciding .64
10. Displk s alertness .G4
11. Thinks dceply .64
12. Sho,.'. creativity .64
13. Convcrses easily on a variety of subjects .64
14. Reads widely .63
15. Likes to read .62
16. Identifies connections among ideas .60
If. Problem-solving Ability
1. Able to apply knowledge to problems at hand .74
2. Makes go~d decisions .73
3. Poses problems in an optimal way .73
4. Displays common sense .66
5. Displays objectivity .66
6. Solves problc:ns well .66
7. Plans ahead .64
8. Has good intuitions .6?
9. Gets to the heart of problems .62
10. Appreciates truth .61
11. Considers the end result of actions .61
12. ApproachQs prohlems thou-htft-Ily ED
Conceptions of Intellrc,-Ice
Table 5 (Cortd.)
Factor Lo-'Zin;
IIl. Practical Intelligence
1. Sizes up situations wcll 6"
2. Detenmines how to achieve goals .3
3. Displays awareness to world around him or her .69
4. Displays interest in the worldc zt large .63
- ..- n..-~- 4* - *-
Conceptions of Intellicnze
49
Table 6
Intercorrelations of Person Ratings and IQ Scores
Self
Int. Ac. Int. Ev. Irt. IQ
Rated Intelligence 1.00 .60"** .624* .23*
Rated Academic Intelligence 1.00 .54'** .36+* *
Rated Everyday Intelligence 1.00 .30*
IQ 1.0
Other
Int. Ac. Int. Ev. Int.
Rated Intelligence 1.00 .25* .48 * *
Rated /-:ae:ic Irtelligence 1.00 .39**
Rated Everyday Intelligence 1.00
*p_ < .0
** <. .01
***2i)- .0ol
Conceptions of Intelli rce
50
Table 7
Correlations of ractor Scores with IQ and Self-Ratinr;
Self-Ratings
IQ Intelligence Ac. Intel. Ev. Intel.
Intelligence
I. Practical problem-solving
ability .16 .44** .31 .19
I. Verbal ability .23 .41* .35* .27
I1. Social competence .14 .16 .21 .07
Kiltiple Correlation .24 .55* .38 .30
Acadcmic Intelligence
I. Verbal ability .29 .38* .36* .22
I . Problem-solving ability .01 .49** .37* . *
III. Social competence .03 .09 -.06 .02
Multiple Correlation .33 .53* .50* .39
Every"dy Intelligence
I. Practical problem-solving
ability .11 .48** .34* .25
II. Social competence .03 .37* .28 .20
III. Character .10 .18 .28 .C5
IV. Interest in learning and
culture .52*** .31 .20 .30
Multiple Correlation .57* .57* .35 .43
*i < .05
"£< .001
ti
Conceptions of Intclli.r.c.:e
51Tablc 8
Properties of Prototypicality M'.,asures
Basic Statistics
Measure Mean Standard Deviation Range
Intelligance .404,** .20 -.05 - .65
Academic Intelligence .31*** .19 -.03 - .56
Everyday Intelligence .41*** .18 -.02 - .64
CorrelationsSelf-PRtings
Measure IQ Intelligence Ac. Int. Ev. Int.
Intelli9ence .36* .40*
AcadEr-ic Intelligence .5G*** .40* .42* .3',
Every-,z Intelligcnce .45** .32 .34* .17
Note: Prototy iclity measure cc- uted as correlation bet ,ecn sutject'- c-_ra:-
terization of his or her own behaviors and "prototypic2" eXq:-:'s
characterization of "ideal" person's behaviors.
*p. .05
**j_ < .01
***p .001
4
Conceptions of Intellig nce
Table 9
Typical Descriptions of People Used in Experiment 3
Unquantified
Susan:
She keeps an open mind.
She is kno'.,ledgeable about a particular field of knowle,'rc.
She converses well.
She shows a lack of independence.
She is on time for appointments.
Ad am:
He deals effectively with people.
He thinks he knows everything.
He shows a lack of independence.
He lacks interest in solving problemis.
He speaks clearly and articulately.
He fails to ask questions.
He is on tir,)e for appointments.
Quantified
Alice:
She sonetir.es shows a lack of independence.
She often reads widely.
She is never verbally f1kcnt.
She often is or time for appointr.ents.
She has a sc-ic cieoc.
She ofter rL.sors loica,; and well.
She sc-eti- lazcs a u' Jcrstdndir, of the nature of thince.
Bob:
He ofter ,' . . , ir, the world at large.
He often h3, a scc' -' ence.
He S C,-(-I, I .
He always f ,rs tr, e urf'- ' ;ar.
Technical Reports Presently in this Series
iI 150-412, ONR Contract N0001478CO025
No. Name Published Reference
I Intelligence Research at the _nterface Sternberg, R. J. Intellieence researv*between Differential and Colnitive at the interface between differenti%!
Psychology. January, 1978. and cognitive psychology. yn~ eli1e:-19781, 2, 195-222.
2 Isolating the Comoonents of Intelli- Sternberg, R. J. TIclatin; the cc=:=cnz
genc. January, 17. of intelligence. Intelliience. 1973,2, 117-128.
3 Deductive Reasoning. January, 1978. Sternberg, R. J., Cuyote, M. J., &Turner, M. E. Deductive reasonin.In R. E. Snow, P.-k. Federico, SW. Montague (Eds.), A-:itule, lea.-:-
and instruction: Coniti'.,e :rcessanalysis (Vol. -1) .ilisale, A'.J.:Erlbaum, 193.
4 Toward a Unified Connonential Theory Sternberg, R. J. Toward a unified co-z:
0FHuman Reasonin. Ail, 3 nential theory of human in:el!-.-'I. Fluid ability. in X. trie:=an,J, Das, & N. O'Connor (Eds.). Inte~il-gence and learninZ. New York: ?-e:-
5 A Transitive-Chain Theory of Svllo- UMPUBLISHED TO DATE
Aistic Re-isoning. April, 1976.
6 Components of Syllogistic Reasonin. Sternberg, R. J., & Turner, M. E.
April, T973. Components of sylloglstic reascnin;.
Acta Psycholc'~ia, in press.
7 Metaphor, Induction, and Social Poli- Sternberg, R. J., Touranceau, R., &SE: The Converence of .-acros-ooc Nigro, G. Metaphor, induction, and
and Xicroscopic Views. April, 1978. social policy: The conver;ence ofmacroscopic and microscopic vicws.
In A. Ortony (Ed.), ' .a.ho tT: :
New York: Ca=bridge Unvcrsi:'y ?ress,
1979.
8 A Proposed Resolution of Curious Con- Sternberg, R. J. A proposed resoluticm
flicts in the Lite ature on Linear curious conflicts in the literature on
Syllogis=s. June, 1=15. linear syllogisms. In R. 'lickerson (Attention and perfor-nance VIII . isdale, N.J.-£rlbau , 133%'"-
9 The Nature of Mental Abilities. Sternberg, R. J. The nature of mentl
June, 197d. abilities. knericnn Ps~cholozis:.1979, 34, 21 'Z3K.
TechnIcal Reports Presently in this Series
WR 150-412
Page 2
No. Name Published Reference
10 Psychometrics, Mathematical Psychology, and UNPUBLISHABLE.Cognition: Confessrons of a Closet Psy--c-ometrician. June, 1978.
11 Understanding and Appreciating Metaphors. UNPUBLISHED TO DATE.June, 1978.
12. Representation and Process in Transitive Sternberg, R. J. Renresen:ation a:Inference. October, 178. process in linear syllogistic
reasoning. Journal of -Psychology: General15,Y7 1,:,119-159.
13 Aptness in Metaphor. October, 1978. Tourangeau, R., L Sternberg, R. J.Aptness in metaphor. CognitivePsychology, in press.
14 Contrasting Conceotions of IntejlAlence and Sternberg, R, J. Factor theories ctheir Educational Implications. intelligence are all ri;ht al=.o3
November, 1978. Educational Researcher, in pres:.
15 An Aptitude-Strategy Interaction in Linear Sternberg, R. J., & Weil, E. M.Syllogistic Reasoning. April, 197T An aptitude-stratczy in:erac:izn
linear syllogistic reasoninz.Journal of Educatc'.l ?socn:1980, 72 9 226-234.
!- Intelligence Tests in the Year 2000: What Sternberg, R. J. Six authors in
Forms will they Take and what Purposes search of a character: A playll the erve April, 1979. about intelligence tests in te
year 2000. Intelli;ence, 1979,
31, 281-291.
17 New Views on IQs: A Silent Revolution of Sternberg, R. J. Stalkin; the 1.0.
the7T"s. April, T979. quark. Psycholocy To 3v, 1979, i42-54,
18 Unities in Inductive Reasoning. October, UNPUBLISHED TO DATE.1979,
19 Components of Human Intelligence. October, Sternberg, R. J. Sketch of a co=r:nential subtheory of h_=an in:ee-
ligence. Beh3vioral a-u 3raiSciences, Tnprs
20 The Construct Validity of Aptitude Tests: Sternberg, R. 3. The construct v2!
An Information-Processing Assessment. ty of aptitude tests: An Infor-A
October, 1979. tion-processing assessment. In
Technical Reports Presently In this Series
MR 150-412
Page 3
No. Name Published Reference
20 (Continued) A, P, Maslow, R. V. McKillu?.& M. Thatcher (Eds.), Cons:r=:validity in psycholo-1cal =e_.
ment. Princeton: Educati_:i.Testing Service, in press.
21 Evaluation of Evidence in Causal Inference. Schustack, M. W., & Sternberg.
October, 1979. R. J. Evaluation of evide=:ein causal inference. Jaurni.
of Experinental Psvcholor":-eneral, in press.
22 Componential Approaches to the Training of Sternberg, R. J., Ketron, 3. L.,
Intelligent Performance. April, 1980. & Powell, J. S. Co=panentia.approaches to the trai=ing ::intelligent performance.Intelligcnce, in press.
23 Intelligence and Nonentrenchment. April, 1980. UNPIrBLISHED TO DATE.
24 Reasoning, Problem Solving, and Intelligence. Sternberg, R. J. Reasoning,
April, 198J. problem solving, and intelli-gence. In R. J. Sternber;eHandbook of human i tel -New York: Ca=bridgo n~ve=s_:Press, in press.
25 Claims, Counterclaims. and r-mDonents: A Sternberg, R. J. Claims, ccz=:e
Countercritique of Componential Analysis. claims, and components: AJune, 196). countercritique of co7panen-:i
analysis. Behavioral an! 171:Sciences, in press.
26 Interaction and Analogy in the Comprehension UNPUBLSHED TO DATE.
and Appreciation of Metaphors. October, 1980.
27 The Nature of Intelligence. October, 1980. Sternberg, R. J. The ratur.
intelligence. New Yor; _'.__
sity Education Quarter'.,press.
28 People's Conceptions of Intelligence. October, Sternberg, R. J., Conway, B. Z..1980. Ketron, J. L., & Bernsten,
People's conceptions cf i: .:
ligence. Journal of Pr--..and Social pc)-o1j6 : ;::.-tudes and Social CUnitic-,in press.
Technical Reports Presently in this Series
NR 150-412, ONR Contract N0001478C0025
No. Name Published Reference
29 Nothing Fails Like Success: The Search for Sternberg, R. J. Nothing fails likean Intelligent Paradigm for Studying In- success: The search for an in-telligence. telligent paradigm for studying
intelligence. Journal of Educa-tional Psychology, in press.
30 Reasoning with Determinate and Indeterminate NOT YET PUBLISHED.Linear Syllogisms.
31 A Componential Interpretation of the General Sternberg, R. J., & Gardner, M. K.Factor in Human Intelligence. A componential interpretation of
the general factor in human in-telligence. In H. J. Eysenck (Ed.)A model for Intelligence. Berlin:Springer, in press.
14-
YALE/ STERdIfERG NOVeiober 4. j9f1Pije I
Navy
I Dr. Ed Aiken I Dr. P'nry Mi. H,2frN3Vy Perscnnel RPtD Center Dj'v.rtcf :y.xlyC-CSan Mesc. CA 92152 Ur zty cf C:.fcrnia -,t 3 'n Diegr-
IJcA±,), CA 920O-1 Meryl S. Baker
?JPRt 1r L7 ~ n D. 1'. rrir., 'ISC U.3"%cde P?'O C "San Diegc. Ct. 9215? ,i r Df,%v' 1c, r:* Centcr
1 Dr. Jack R. BcrntinPrcvcs*t & krile,c De-n 1 f,!- r 1c k P. I .r r rU.S. WJivrl Pcstgra-1u~te Schcil P! :c..y Cctu': .' rcctcrFcntce~y , CA 93C,110 LE VT . " ;P (7b)
1, rz F-C -' L T VE Ln;T-Fr1 Dr. Rcb,2t Preat:x 11~. ' AL AC: , ;
Orlandc, FL ?2F12 1 :. H2zCc
1 Chief cCf Nac Edu.-oticn and -r,.injng -:. scr.n)O F T) C-ntu-r
liasc-r, Cffict ~c , CA >'1Wr
Air Fcrce Humian ;-scurce Lsbcr~tcryFlylr<,- Trzini',g Pviscn I CU 1.HWILLIWIS AFE, AZ P2L rSyt.:sCcr.:-n: :
1 Dr.' Larry !- :r 17, !Z, UC; U:: rtmri'.Psyccl.3y Cepar',-cr.*...,ci C2~Navpl Sum:rir~c il Resc-rch LabNaval Subizrine Ls 1 C,' - r'S.I :~:
Cfrctcn. CT I!,.~ n - r '
I Dr. hichari Elst~r FC IDpart-xc1 cf Ser~tri c cences 1 1) '1
a4avzl PcstF*r~du--,, 2chccllcnte,-ey, CA )Y"'1r --
C I .rv, It ~r.I DR~. PAT FEIDPICO iJ ir r. tcr, 1-ip 15 (7:
tAVY PEPZY : EL RI.: 1 CE!!E R ISAN~ DIEGc1. CA 92152
I Mr. Paul Fzl'ry Pri j C,.; 4Navy P:rscnr&i FV' CL-nter k: %L. ct -
&-n~ Diec. CA 9215? !6tj r,-,. r-l Cc -
I Dr. Jchn Fcrdf~avy Prscnnrl Br)~ Center I .r . A 'r.
Son DieAc. CA 9?152 S, 1 1C 1 Lf,- v , :c tc DC? F' r,
4' ~~ DC r*
YALE/STrRIPERG Ncfcnb--r 4I, 1930 Pale 2
Navy 14-vy
1 CAPT Richard L. ?'-.rtin. USN 1 Psy--t.:1cgistPrcsp ,ctive Ccinri'ndxr. Office'r C:R Lrznz-h Offic#"USS Cirl 'Jlnscn (CV'.-7r) rlde 1 1-, S.!cticn Dflew.,pcrt llews Siipbuildin3 ond Drydcck Cc WC. Su-n-ner 5F.rcctIlewport f1!ews, VA 27fl7 Bcstcn, MA 0221(r
I Dr. ti'' sTcPri1'. 1 Psy: .lczibt113VY P,rz~cnnel PR,*) Center ONR Brznc:h CfficeS&n Dicrc. CA 9217-? 53 S. Clzr:. Streat
Chic - c. IL (/1 Dr. G ,cr,,c ?'tefllr
Head, H-j,-:n F,'clcrs rept. 1 ofricc cr r:oviA r' se--rchtNavzi1 Submarjno 1*-dical Fesearch Lab Cc do " I~7
Grctcn, C: CC 3L4C? F0, N. ouincy trtArlirngtcn. VA 22217
1 Dr IfIi~,!cnt?;uet'avy Prs-crnel F-.) Cerftcr 1 Cffice cf IFav:I Fits -rchSan ri,,c, C 21r'? CCjr LLJ1
3 r) !. Quinc~y Street1 Library /.r~ir.;tcn. VA 22217
P. 0. H-Cx :"5122 5 P-srnnvl L Trainir.:; FP-serct. P: :Fr;;s5-r. Nczc. CA 1'111 (CcJi 4~5?)
Cff!:.- cr 1:.svpl Ressarel1 e:J,;c l F.r Cc:-.:ir. jA- . .c n. VA 2??1-
1K.t2crLl !:z~va I'Ld ica C. nter 1 Pzy .~Ic CZistrr 2C'14 r~ [r jnch f f ic c
1 ClrT F.,.; ! L tI : r, CA 9 1 1 rCii:f. :-ic~2 2 rvic-? Ccrp5
P i r f r ei~: -iuc-r (j-~ C. ff>: cf 4~- C4~c Cf I- -Cp'rZ:
1 Tcr! '. 1. Ycllc-n-.c1rv2 nfr;ti ficr. Cc 1 1 r2dF. F.irk.-r
h1WVT P Yr77,'EL V.)CE:TER 0-1~ Schrcl cf rusiness ~ sr.rl* s.1 L'- n. i ~ILy c f Vici~ ;n
A,-. ;-,cr . i HI LS1~
1 Libjr -y. Ccd. P2'1L* t~~~Nvy Perscnnel rt7) C. n:r I L7 Fr~C. Pethc (.L'' m.
San Dic3c, CA ?5.C L9
6 Ccn-.ildrp, Officer fl.r..-tcr-a,. FL ~~tiavol Fesercl LaborctcryCcde 2627Wshin3tcn, DC 2l3r.0
- - - w - - -
YAsLE/ STER?" IE RC tic v -i Ur ii, 19.fl Par' 3
Navy N"iv y
1 Rcgcr V. Fcrni;tcr, Fl.I) 1 Dr. Pcbert VisherCcdc L52? Ccd- 3r?9N A 1J t Nay) rc.rscr.n)el R&O C',nlerPensaicclz&, FL San fliS.'n CA 9215?
1 Dr. Ivroiard Fi';.. (f31) 1 DR. iVI~.f!; F. i,-S 'OFFNavy P.-rfzne 1,* Cc nt' r NJAVY Pf'.. CN"IEL R& D CE;J'-EhSon M-.-c, C;. ~1~SAN D)H CS, CF 92152
1 Mr. Arvzcld P~-in1 Mr Jchn 1!. Wc-lfeF13a Pcs;&.>ppcrt - chnsic~y Ccdc F-1c113ri!.l z~ o i '~?4 U. s. ',:ivy Perss.nne1 Pesearch andiRccri1'-, 'YJPl~zi' tzvelcrent Ccntcr2221 J f fc r cn i S.,n i c c, CA 92 1
1 Dr. VcrtKl fca-.1 ,,
Ccde N-r
IJA Fs-P L 1750!,
1 Dr. Sa-i S.chifl t , FY 7?1
U.S. !:-,v,!, Air *. Ccn,'.r
I Dr. Pct)rt G. rh
Mifce cf C- i'!f sr ":av aC;?rt !
Washirnj" . C
1 Dr. 1.1 ftci F. L-c,!,
Orlandc. F'L '3 1
I W. Gary TczN3 V al C: F7,r Sy s'c- Cen terCcde 71'2San tic:Lc, CA .V
1 Dr. Rcn,2A 'it'Cc~dr 5L.Deportiert~ cf Al-i istr.iv, rSciences
tcntercy, CA 9:'<
Army Air Fcrce
Technicz! DIr-ctcr I Air Uivvrrity Libra~ryU. S. Army Ins~c"Th~itut., fci ,fi,2FUL/LSE 7f) I1
Bcha icr;. p~ :. ci~ fl ic:cc; Kxw.~2l IFP. tL 36112
Alexaniria. VA 22 7-. 1 fr* Earl t.. Alluisi
11:1. AFHFL (AF.-C)HQ USP"FUE7 & "th.I- rrccs AFL. TX 7"235
USAAP.ELE !"ictc-f Cr.F1 1 nr. CG.nevieve F'-iAddFAPO !:ew Ycri. C)IIC7 Prc!,rz-i L73ri~qcr
I.A fc :ci(flC'cL Ti rectcri.teDF. PALr'1 V2. 7KF2 "1U.S. AF~'I I F : TT !':lling AFB1, PC 20 2
ALEX.AL7' .Vf 2,- 1 ' cnqld G. Eu.,hes/F 2 L /IT P
Dr. 1-lict -. , Yzpl ,! Vi I IL-is P.F , .2 z,22LU.S. AR.1Y ?E7 < -
5001 E': F* -i? :ER V.u< 1 I r . ss L. 1!trg~n CFLL.ALEYAD Y2. * VA 2c33 -,. -P;atlcrssn AFT
Dr. t? tct? . Kn'_-Trainir:; T'J icz;l ,*- .~ cc 11Ki Fe
50l01 Ei~hizv-r ^'c:o I' ', ,, TX 725Alex ,njrac, V;, 227,
I - rty F-cch .!y
1 r .HarclL F. O'Wei) Jr. P TiN1DrtctcrAttn: PER'--)IArm~y tsc.--: Ir.zt~ n --s /AFI, A7 5"2%.500~1 Eis~ri,zwcr l,Vf-,u:-Alexzn~ri -, VA 22V: CT/7G tp~
r F- H. TX 7 6--I I
Dr. Fcb-rt f:>s'1crU. S. ,-y rATntut : ' 1 2* .*U3.
Alexanriz,, V, 22;. - AFE, rN 2 .U)-
Dr. Fredcricr Etcin'. ~rU. S. A-n-y ir!.2ztt
Dr. jcr5 .p'. -U.S. Army F~cr[P: tt
Alexandria. W. 22'-
YALE/ JERNER(J Ncvonibr.r 4. 19?n aL
Marines Cc~iLGu~ird
1 11. will!--Ir Grernorm 1 Cni.f. Psycjrc.~icj E rarZJ.Educaticr t1i'iscr CwiU. F. CCoJt C.Jrd (--//P2E'Jucatjcn (Ynt.pr, C7 W:rhinFltcA. D, ~MQuanticc. VA Z1'
1 1r. Thcrnns A. ~'v1 eadqur.,r:-., U. S. r':.rine CcrrL U. S. Cc.ost Gu-r1]d z~u
cdc ?KPI-2v: P. 0. Substcticn 1iL
1SpeCi~l tS:-iSt:]nt fcr -!rinc
Ccde 101"Office cf ! Tn*c: rc
Arlingtor. X'.. 22"117
1 DF~. A.L. SF:Y:SCIENZTIFIC ,V; (CC2E7 UX-1)
HO. U -S. c iir.
I~iA IIHN3-11 I:- 20-Z
YAL/ST7W:=1P: Nc v .i b r 4.Iile P;j,;e 6
Ot h e-r rr- Civil Ocvt
171 Thfc'ns" Technic.l 1 nfcrm,ticn Ctentcr 1 Dr. Susan Chipi&nCzmercrn taticnr, Elc%5 Le~jrninp, and Develcpint.Aexpnria. W, 27 1l4 N-aticn-31 Institute cf f~ucF't.cr,Attn: TC 12T) 19th Street I.,:
'.r.shin~ktcn, DC 202!"A1').1~v e ri (1.cl.cr
AR\':u:E: , PR''JECT5 AOE'2-Y 1 Dr. j-,s71h 1. Lipscn
APL~Y~~.VA 2;2V) !Jatcnzl Science FcuiJ..ticnV-s~int~cn. DC 2055)~
".12aty -Y r* fc r Trainr- r-.ndF r -~ ' - cy 1 i2~iJ. 1cLaurir
Cf f~ c cf t 1-I -V. :- '-cret..ry c f Zf ens 5, P. '01. Intern; Fcv2nue 5vcrvicefcr F-se-cK En Ein rinr- 21221 Jcfferscn Zavis "i,-hwoy
7. A f Jr I irn VA 22 2
1 C.Ar.irew P. 1Vc~n~rZc ':~EJucF-ticn fX.'
ai PcsenrcK.n~'ttr. Scece 2C%~n'.
1 Ptrscrnel FID Centerfffic-, cf Perscrnn2 ; '..~
1 £ E Street !..
I Pr. P. V!a11rc- in~i;,cPrc,7ri. !71rcztz:r
Smi*.thscni -n :n 'titL.*
501 Ncrth Pitt Sr.Alex :ndria* VA 2'2311
1 Dr. FrEnk LithrcwV. S. OffricC cf FdUCL.1cnL' "rylni Ave.bit S..'C20C?.
1 Dr. J::zPh L. Ycu., )ir,:.tcr
"'cry I~ Cc~nitive rrccc~ss~s
~~ ~DC ?~
YALE/"TERNBERG !Ncvm.ber 11. 1930 PiZ
t'cn Covt Mc n Gr.v t
D Tr. John P~. tmer& 1 Dr. J-3ckson .;yDepartment of Pzycho'lc-y De partrnent o-f Psy ,olcgYCornegie t'.llon UMivtrsity University of Calafcrrni-Pittsburgh. PA 15213 Los fngcles, CA 9CS24
I An dcr sc n, Tis E!. . PI, P. I Dr . Is3ac Ep .-rCentrr fcr t!'-, r'uiy cr F2,.ing Educaticnzal Testir,: ' rvicc174U Children's~ F-?Sczrci C-nter Princeton, UJ C311c"51 Ccrty Drivr -
Charmpia~n, IL f'12*1C 1 Dr. ?Vichclcs A. FcrEflpt. of P5YohCICv
I Dr. J~hr. A-,n,*.. Sacra;icnto Statc Ccilt2,-.D--p'rtnet cf Psyv4Elo,--y 600 Jziy 27trectUnivcrsity :f ~>~c:Sacramcntc . CA 9511
E 1: GL A NI 1 Dr. Lyle Bo ur ner pirtr~ent. of Psyc',.ccy
I DR. ',lCHIAEL ~'~Univc'?rsity cf Colcradic.SZIE'E r LI:Tc; I,:rTL'TF Eculdcr, CC U'03')9J
70313 £. PRTr""Er 'AVZ.L 1 Dr. Robert ErerrnE:ILE,.:CCr', CO '2'-11 A-rerican College Tcsl~ng Prcrgrzns
P. 0. E-ox IGS1 1 psymcloci. rcs~z:rcl. inl Icw City. IA 5;?21;0
IYpt. of ,~~ (A-iy crficcIC.-,,bcll F,:,. 'f f cs 1 Dr. Jc-r S. Erc -rC,-'ib~rra AC 126".0, A ust r.1i a XEROX Palo Alto Pesfczrch Center
3z3? Coyote Rczd1 D'-. Plan F&F-1 alo Alt,. CA ~C
Appli-' P-ync'~ U'nit I Dr. Pruce Euchanpn19, Cu,.-~C-P -4z Cp:rtvrent cf Corn u -2r Scienc2
E::JLA Stijnford, CA 9~
1 Dr . Pp t ic~ [i -1 1 DF. C. VIZTOP ~XE.O!r p-rtior. of Pryz-.cgcy 0:1C A 71 IC .Lniversitv% cf F-,r.'jr UJIVEF.ITY PLAZt, * UITF 7r,11ni/frrsity F ,r; 11 3 SrC. STA7E ST.cnvcr. C.' S--- ORE2 * UT CJC5'
1 ?Ir Avrcr. r-rr 1 Dr. PF . CarperterrI-p-irt~nnt of r(c puts-r Scienc" Vpz:-nen. of Psyctzlc,'
Stanfcrd. CA ~K5Pilttsburgh, PA 15?13
N~C 4.r~ Pa,,v 3
Ur-n Gov*, firn Sov t
Dr. John V. Carrcl 1 Dr. Lyn-) A. Cccrper
Psychc-lct-ic Lob LR DC
Univ. of "c. Circl itw Univer:;ity of Pittrbu~vh
Davie E1ll C013A 3939 O':Iara 5treetCh,:pel Vill.C 14C Pittsbur-h, PA 15212,
Charles !'v'rs Li )r,.?y 1 Dr. 1'crcdit!% F. Crzv;fcr,"!Livizigstci.t !iou--c Anlericafl Psyzhlc;jcl~ 5C.C
L iv in:; , , nc Rc. z 1 20? 17 th Str,?t.
SL rat f crd Washinmtcn. C 27,7rLct,,or. F15 2LjE NZ L A '7 1 Dr. Yenreth T. Crs
Anacap3 Sciences. In..
Dr. Wi11ic-i Cha:: P.'-. rra-.cr C
Depirtr~nt cf Pv1ocyS~nt.: E,,rba.ra, C.:.91?
C~rnei rclc L-vrrsiityPittslurg,., Pft. 1")217 1 Dr. Emmnanuel DcncVhin
E -,p zr tme r.'- of Psychciocy
Dr. TcTh Chi Univers-ity of Ml1incis
Learr.,n. R S D Cerntcr Cn.-rpcign, IL '3182,?
?399 " ;1Lr: StrocrL 1 Dr. Hubort DrE-yfus-
Pittzbu-;i. PA 15?1-, D- T-rtmmnt of Prilcz-cp1-yUniversity cf C.,)ifcrni:
r. i~ Cncry Pcrkely, CA ~2D c~~o Cc7-.~. .7-2,-nc
1 LCft J. C. ES~cntr-er
'tanfo-' C4 D IPECrC.ATE OF P7F.SC'!"EL PPL:EC-1:ATm:ALDEFE:ZE
Dr . F, . C2 1r C1 C' L, FL BY EIVE
CE r> ir 1 EP IC FciII tY--Ac qu. i a 7f,
Dr. .:cr'--m CliffD.pt. Cf F:cy--.clc-y 1 Dr. Ed FC1en,r-,
uni V. C f 7rc. CP-c~~Dp~rt~rn fcc . r Ic
Un~iv'-r.-it y Park vtfo1
Dr. AlI;-. 1'. Ccl:rz1 . .. LFruz
2c~Cz~f- rc ~:& ~. Prcz-r,-
50 'cul cn -. tr "
YALE/FTERIJBERG Nc v anber *4. 19-21 Pzpe~ 9
Non Govt Nion Go vt.
1 Dr. Edwin f~. Fleishinan 1 Dr. Ihrcld Hawk'insAdv~nced Res-2arch Resource.- Crjan. Dvop:rtintnt cf Psycrx.lcrySuit-e 900 Univcr~ity of Oregon41330 East IRest Hirh%.'ay Fuigcn OR 974C34-as'ingtor., DO 2C,01
I Dr. P.~rbarr. !Iays-Fcth1 Dr. Jchn F~. F-rcirri,;sen The Bin~d Crpcrzticn
Bolt &ranc'k . ?2"-,nrn 1700 'lin Street50 I'cultcn Szc~ nta licnica, CA 90L40[Cambridge, ;,.. r2133
1 Dr. Frederick P.yes-RcthI Dr. Al indp Fr ifednan The Fand Crpr-tion
D'prtlnent c-f Psychology 17CI ?lain StreetUniverzity cf Alerta Sants 1'cnica, CA 90!406F6mc-rtor, L.1Lc rta
CA'ThI G 27-9 1 Dr. James R. HcffmanDep:,rtment cf PsychzlIc~y
I Dr. R. Edw~r ! G-isclman University of rze1,;arePDcp rt-.rnt cf Psychclcgy 1.zwrk. DE 19711llnivcrs. ty cf Califerni3Lcn .;1s C". 9")2" 1 Glcnda Greenw, ld, E .
"Hfumcn intelligence Newsicttcr"1 DR. P:TEP' OILEF P. 0. Pcx 1163
L . T' Bir:,inhan, 111 42012
~G., -'-T; 1 Dr. Licyd ILunpn-cysP77TSFI9, 7-2A 15213 Depirtment of Psy-i-ccy
Uiivcrsity of Illincis1 DV. rmnrvir r. Glc.&( Ch ,ipzign, IL 61P20
2 17 F !cr I:n-I'cr!-,el Inivc(trzj,y 1 Library
itL. Y Hum FFc/Ve stern Piv isicn27F-57 Berwick Drive
'. :niel Gc--C;rncl, CA 93921~nt'ril ' 8ncg.:,eent Enmincerin,Tch~on:s~ainstitute of Tcchnc~lcgy 1 Dr. Eorl Hunt
UI if. Dpt. of Psychclc.,yI Z ; L University of Was'.inatcn
Se~ttle, WA 961 CcI r-.. JPA'ES G. GREP7n
LP- 1 Dr. Steven W.. KeeleL:!VrP.I-,Y OF PITTSRURCH Dpt. of Psychclc~y
3%~, YA~A FEETUnivcrsity of Gre~cnP:-TT-7su%,H, PI. 15213 Eu~enc. OR 97403
1 Dr. Ron Hanbletcn 1 Dr. Walter Kintschf.chCcl of Edic; I-in Dopart~nent of PsychcaopyUrnicrrity of !'-issechusetts University of ColoradoA; .erst. VA 01,T)2 Bculd(er. CO F0C302
YALE/nTER'JPERG ?Vvmb'"r '1. 1 ) j,
Ncn Govt Ik~n Gcvi
1 Dr. Divid Kir~s D r. tDzi,;Ij.~, %rr:,Dcp:rtncnt Cf Pnyclr,,-y INpt- cf r jt~c~ -University cf Arizcr... Univ. cf C,jlifcrnia. AF-r Die~cTusccn. AZ ?5721 Lo Jcoli.2. CA~ 92093
I Dr. Stcphon !Ycrslyn 1 Dr. I'tlvn R. I:CvickHa~rvard University 356 L :,Jcuist Cerster fcr Vesurs..-.-De-particrit cf psycl..cc-y Univrrznity C~f IcW333 Kirklznd Strc ,t Jcv,i Ci ty. 11, 522i2Cobrid-*e, XA C21,!$
1 Dr. J,-jsse OrlzinskyI Mr. Mitrlin 1.rcgcr Insttjut fcr D--fcnse .inalses
1117 Viv Ccloto 40O0 Artiy !!.vy L-iv&-
Pz'1cs Vcrcs Est,-tc:., c! l Y? Arlirgtcn, VA ?2202
1 Dr. Jill Lar (in 1 Dr. Sy-cu- ,,. P~p:rt IDepartment C,' P~y.ij, 'c r~y 1'assaclus: tt! !nstt2' f7Carnegie t'-?!cr nv'st Artifici.)]I~1i L~bPittsburrKh FA 1'3?13 L5Tc' cy tw
1 Dr. A1.-n I-s-cl-lLearnin, F0 Ce.nt,-! 1 Dr. J~,?s t. P,-ulzcr.Univcrsity cf Pitt= u-; . Pcrtl~n1l Stt U'.ztPittsbur h, PA 1Q?.P.O. Ecx 751
PcrtlanJt, OP. 72,^7I Dr. Charlc - Lev-'s
Fpculteit Sccinle I&:sI 1 . LUIC2' P-ULL2Fijksuiverrit-2it ,. ,i 21431 !:. EDrrE.,TD 7E
Ou~ £t~rrstr.tAP.LINGT()',. VA 22Z2'7O r. i n j,1:7ETHERL'D' 1 Dr. I!Frthim Pclscn
D-par'r,r.t c~f Psych: 2c;)yDr. Jam~es Lm~der Univcr-nity Cf Cclrr;-crcpartment cf Psyc.iclcly . iculder, CO RC--C2Univerzity cf Wez-.ri. Austr:itNedlandF %4.A. 6C0"9 1 DR. PETEP P'OLSONAUSTRALIA DEPT. OF PnYOK)LrZY
U':IVZRf5:TY OF COLORAIXP)Dr. tCari: 'iller ECULDER, CC V309Ccrnputer Science L-lhcr;,trryTexas 1rnstrwjicnts, Tnc. I Dr. Steven E. PcltrCCkMiail Staticn 371. P .O. Ecx 2'159*, DoFjrt'1cnt cf Psych~ccyDallas. TY 75P65 University Cf Denve.r
D,.nvcr,C0 E02.13Dr. Allen l'unrcPehnvlcral Technclczy L-obcr; tcrics 1 DR. DIAI:E M~. RAMSEY-KLEE124I5 Elena tve.. Fcurtl' Flccr R-!K RESFIAfCI! & "pZT--. r:s:ciNedendc Pea~ch. Ct 90277 3947 RIDGEMMIYT rRIVE
NP~LTBU, CA 9C 255
YALE/STRI.fl3FRG Novellbcr 4, 1930~ Pa~ge 11
NCfl CCVt tic n Gov t
1 IINRAT M. L. RAUCHI I Commnittee e~n Ccjgnitiv(. Fesearchp 11 JJ % Dr. Lonnie R. She.rrcAlPUNJDE7'*'I?!STERI1J1. !VER VERTEIDIG'JNG Sccial Science Resezrch CcuncilPOSTFACII 1328 0~5 Third AvenueD-53 SOW! 1. GER!IANIY Mew Yrk, NY 10016
1 Dr. T1-irk D. Reckase 1 Rcb,,rt S. SieglerEduczticnal Psychclc.,y Dept. Associate PrcfesscrUniversity of Misscuri-Cciurnbi: Carncgie-1iellcn Univ,.rsityI i Hil~l I'--ll Dep3rtnent cf Psychclc .yCcltribia, 11r1 65211 Schcnley~ Par),
Pittsburgh, PA 15213I Dr. Fred Reif
SESA' E 1 Dr. Rcbert Smithc/c Pnysics Duprt.ent De-partnient cf Cciputcr $,cienceUnivcrsity cf Cclifcrrnia Rutgers UniversityScrkely. CA 04720 flew Erunsvick , NJ C"9'-
1 Dr. Andrew !'. Rcse 1 Dr. Richard Sncwkrieric~n Tstitutes fcr Fesearcn Schccl cf E'Juticr1055 ThSJ fferz-c 3t. N'., Stanfcrd University1V;shin,7tcr., DC 2CO7 Stanfcrd, CA 9~43n~5
1 Dr. Ernst 7. Rcthkcpf 1 DR. ALBFPT -PTEVENSBell Labcr--tcries EOLT BERANEY & ~ IZ6 0 0' ou n 'r A ver.u~ 50 'r!ULTON' STREET!'urrzy Hill, 1VJ 079741 CA",BRIDGE, !:A 0l2112'
1 PROF. FI.'711:0 SAM-EJIP4 A 1 Dr. Thomas G. Sticht.DEPT. 0?r P-y:H7LC'y Directcr, Bzasic Skills 'ivisicnU!1'l VE-.Fj-PY 7 TE':-!Er7FE Hul!SPOKNIlYVILL-, 7'; 379j1( 3fl0 N. Wshin.7tcr. S*tr,?t
tAlexarndria,VA 2231L,1 Dr. Irwin, E rascn
D.-p3rtrit of Psycl.clcgy 1 Dr. David Stcnet~ivrstj f s ,ntcn ED 2-16
Seattli,. WA 9PI195 SUNY, AlbanyAlb7?ny, NY 1222?
1 PD. W1,LTEF :3CH'1EID*.PD:PT. ('7 P3Y:'10LM1Y 1 DR. PATRICK SUiPPESUNIVF".71TY OFF L?;: INST1TUTF FOR IA'ETCLSTUI!Z
CH~P :?:*L 616-? THE SOC IfiL :r:LSTA'.cTID U1VIVER31TY
1 Dr. Alcn chcenfeld STA'FOED, CA 94175Deprt~~r;..of 'aticr-nitics
H-i,iltcn CcllegeClintcn, !JY 13323
YA Lt/."T7- 4~r Ncvcvnber ii, 1980 Page 12
Vocn Gcvt 1.4 n Gc vt
I Dr. Pirihiri-i Swmin~i.lh~n I Dr. Ikird VIPinerLabor;cry c f Psyc*-x-nc1.rlc 4ind I'urcz,u cf' t~cia1 SCiencv hesearch~
Ev.:!1uj.ticn V~sonrch 1990 '4 Street, N. 11.Schc~c) cf [ducatirll Vazihinetcn. DC 200361llnivcrzity cf Wvissa!chusetts
In~e~t* I0t10,)3 1 r. Pty2 ]is 1IA:?verGr.,dujt-a Sct~ccl ef Educ;.ticn
I Dr. ViIli:--)i Titsucka IV!jrvzord UniviersieyCc!i,uli2.-r r:!zz Fduzticn Resu'.rch 2f)') Lars-en UzaI1. Appizcn Vay
I. iambridyc. MA P213F25? En ;ij--rinn Re r. rch Labcr~jtcrytUnivcrz: *Y Cf IlIljnzis I Dr. Do4vid J. WeissUrbv't;, IL c1101 t6iff 'E11iott Fmll L
Univ,-rsity cf fVinnescItaDr. T):vjl 7T-isscn 75 tT. Fiver PoadIr-!prt~i-!-, .f Psyrl.clclty tinneppclls, [IN? 55455Urivtrrz-ty cf n~~Lrwrnc. f;n4 I Dr. Keijth 1. We!r curt
Infcrnrticn Sciences Dept.I Dr. Jz--hn Tnh 7s 7be Rn Ccrpcr,-ticn
112' Tri -i; s J. '.%tsc!) Irscarch Cefnt~r 1700 "-.in St.P.O0. I:Lx Sai:ta rcnica, CA 904()rYc rktr-~ w i&ts !:Y 1C,50
I P~R. Zl2A1 E. WHYTELY1 D. PEFFY wl :Y' P:-YC:':LOOY DEPARTIF1:7T! i-- FA:C r -- iA; Ir U~l:!I;'R3TTY OF NMAZS1703 ?'A::. 777FFT LAJVE^ZE, KANISAS 66044
I Dr. Christcpher Wijckens1 r 1::z 7cvne Drpzrt-nent of Psyc~clc.,y
Uri v. c f Fc. Ca- I ifc rrA University cf IllincisPehavir- ?2chnclc~y L7bs Chimpazi~n. IL 61F201?45 S. E'lcr, Avc.PI.dcndc ~:. CA 5p277 I Dr. J. trthur Vccd6 r:!
Dcp.-rtrient of Psy~iclc ,yI Dr. J. 'J:Ai'r University cf Cliftrniz
Perccptrcrics. Inc. Lcs Angeles, CA 90024J
5?71 Variol Avenu,
I Dr. Benl., J. VLtcr.r.cdDept. of lPsychclrc-yf.'crthw-!stern Lfiv-rs3,tyEvanstcn, IL 6021
1 Dr. WIit- P~. UttalUniversity cf rVichi~nnInstitulte fcr Socitl Res!-arcliAnn Arbcr, III 41lr