This article was downloaded by: [Moskow State Univ Bibliote]On: 06 December 2013, At: 04:50Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: MortimerHouse, 37-41 Mortimer Street, London W1T 3JH, UK

Roeper ReviewPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/uror20

Thinking styles and the giftedRobert J. Sternberg a & Elena L. Grigorenko ba IBM Professor of Psychology and Education , Yale University , Box 11A Yale Station,New Haven, CT, 06520b Doctoral student in the Department of Psychology , Yale University ,Published online: 20 Jan 2010.

To cite this article: Robert J. Sternberg & Elena L. Grigorenko (1993) Thinking styles and the gifted, Roeper Review,16:2, 122-130

To link to this article: http://dx.doi.org/10.1080/02783199309553555

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”)contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensorsmake no representations or warranties whatsoever as to the accuracy, completeness, or suitabilityfor any purpose of the Content. Any opinions and views expressed in this publication are the opinionsand views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy ofthe Content should not be relied upon and should be independently verified with primary sources ofinformation. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands,costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly orindirectly in connection with, in relation to or arising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any substantial orsystematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distributionin any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found athttp://www.tandfonline.com/page/terms-and-conditions

Featured Article

Thinking Styles and the Gifted

Robert J. SternbergElena L. Grigorenko

This article considers the interaction between styles of thinking and gift-edness. In particular, it is argued that decisions regarding identification,instruction and programming need to take into account not only theabilities of the gifted, but their styles of thinking. The article opens withan introduction to some of the main issues of gifted education, and thenis divided into four major parts. First, the article presents the theory ofmental self-government, a theory of thinking styles and their relations togiftedness. Secondly, some data are presented testing the theory onvarious groups of students. Third, the article shows how thinking stylesare relevant to decisions regarding identification, instruction and pro-gramming. Fourth and finally, the article closes with some conclusionsabout thinking styles and the gifted.

Robert J. Sternberg is IBM Professor of Psychology and Educationand Elena L. Grigorenko is a doctoral student in the Department ofPsychology, Yale University.

When effort on top of effort to solve a problem fails,we have to consider the possibility that we are

seeking to solve the wrong problem. In such cases, often thesolution lies in a reformulation of the problem rather than infurther futile efforts at problem solving (Sternberg, 1986).

Such is the situation we confront with respect to variousproblems stemming from programming decisions for thegifted. Consider just a few of the problems that remain unre-solved:

Which is better for gifted children, acceleration or enrich-ment? Are conventional multiple-choice intelligence tests anacceptable means for identifying gifted children? Are giftedchildren different only quantitatively from ordinary children,or do they differ qualitatively—in kind? Is childhood gifted-ness a qualitatively different thing from adult giftedness, and ifso, are we misidentifying the children who will be gifted laterin life? Are gifted children better left on their own to find theirown interests, talents, and ways of doing things, or do theyneed guidance to realize their interests, talents, and preferredmodes of working? Is cooperative learning a useful teachingmethod for gifted children, or do gifted children work better ontheir own than with other children? What forms of testingwork best for the gifted?

These are only a few of the many questions that arise informulating identification and instruction procedures for thegifted. In this article we suggest that one reason our solutionsto these problems have been indefinite is because we have nottaken into account a moderator variable that distinguishes dif-ferent subgroups of gifted children, namely, thinking styles.The "right" answer with regard to each of these, and otherproblems, depends in part upon the set of styles of each giftedchild. What is "right," therefore, is a function in part of how achild is able to express his or her gifts.

Many people believe that the line between gifted and non-gifted children is drawn on the basis of their IQ scores.However, beyond this line there is a huge amount of individualvariance among the gifted. One can say that this variance canbe accounted for by the variety of talents demonstrated bygifted persons, and this explanation has merit. Nevertheless,much of this difference can also be explained by the variety ofthinking styles among the gifted.

It is important to emphasize that styles are not abilities:Whereas a strong case can be made that giftedness

means that a gifted student is "better" than the large majorityof students with respect to some ability or abilities, in the caseof styles, it is not clear what "better" would mean (Grigorenko& Sternberg, 1993a, 1993b). Styles are not better or worse—they are different. Of course, for a given task, a given stylemay be preferable. But for another task, another style may bepreferable. Thus, styles should not be used in identifying giftedchildren, but rather in helping us to understand how best tomaximize their accomplishments, given their particular kindsof giftedness.

The remainder of this article is divided into five parts.First, the theory of mental self-government (Sternberg, 1988),a theory of thinking styles, is summarized. The possibility ofusing this theory in the context of gifted education is pointedout. Second, the relevance of this theory for the six issues ofidentification, programming, and instruction raised above isdiscussed. Third, the issue of testing is considered. Fourth,some tests of the theory of mental self-government is de-scribed. Fifth and finally, we draw some conclusions about theroles of thinking styles in giftedness.

Mental Self-Government: A Theory ofThinking Styles

What are Styles?

When we think about characteristics of the gifted, wemost often think of abilities, and in particular, those abilitiesthat distinguish gifted children from other children. Differ-ences in performance among students do not all devolve fromabilities, however. One important source of differences is thatof thinking styles. Styles are propensities rather than abilities.They are modes of thought that individuals finds comfortableand suitable for themselves.

Styles, like abilities, are not etched in stone at birth. Theyare in large part a function of the environment, and they can bedeveloped. An individual with one proclivity in one task or sit-uation may have a different proclivity in another task or situa-tion. Moreover, some individuals may have one preferred styleat one stage of life and another preferred style at another stage.Styles are not fixed, but fluid. We need to recognize preferred

Manuscript submitted February, 1993.Revision accepted June, 1993

122/Roeper Review, Vol. 16, No. 2

Dow

nloa

ded

by [

Mos

kow

Sta

te U

niv

Bib

liote

] at

04:

50 0

6 D

ecem

ber

2013

styles in ourselves and in others, and to recognize that thesepreferences may change.

Of course, there are many alternative theories of styles(e.g., Gregorc, 1985; Holland, 1973; Kagan, 1976;

Myers, 1980; Renzulli & Smith, 1978; Witkin, 1978). Forexample, Gregorc distinguishes between sequential andrandom thinkers on the one hand, and between concrete andabstract thinkers on the other. Almost simultaneously, suchconstructs as reflection-impulsivity (Kagan, 1958), field de-pendence-independence (Witkin, 1962), and analytical-syn-thetical cognitive styles (Gardner, 1953) were suggested, andall of them were used to present and to explain individual dif-ferences in accomplishments in a way that does not involveIQ. Other researchers have focused attention on cognitive in-formation-processing patterns, such as DeCecco's andCrawford's (1974) work on conceptual tempo and selectionstrategies and Kolb's (1978) on concrete versus abstract think-ing abilities. Other specific styles have been studied as well, ofcourse (see Kogan, 1983, for a review). These theories containa great deal of overlap, as well as some distinguishing features.

Suggestions for matching teaching and learning styleshave been made in the past. According to O'Neil, learningstyles theory offers one way to expand teaching methods andcurricula to reach more students (1990). Kagan (1965) sug-gested that new pedagogical procedures should acknowledgethe interactions between the dispositions of the learner and thematerial, and should tailor presentation to the preferred strate-gy of the child. Renzulli and Smith (1978) have suggested thatindividuals have various learning styles, with each style corre-sponding to a method of teaching: project, drill and recitation,peer teaching, discussion, teaching games, independent study,programmed instruction, lecture, and simulation.

Despite the large number of theories dealing with eitherlearning styles, teaching styles, or both, none of them

can be perceived as universally accepted. Some of the theories,based on very important practical findings, are lacking in theo-retical background whereas others, even though they are con-sistent and attractive, do not have any clear applicability. Thetheory presented here is unique in its use of a metaphor ofmental self-government.

At the heart of the theory is the notion that people needsomehow to govern or manage their everyday activities, inschool as well as outside (Steinberg, 1988, 1990). However,there are many ways of doing so. Whenever possible, peoplechoose styles of managing themselves with which they arecomfortable. Most people are at least somewhat flexible intheir use of styles and try, with varying degrees of success, toadapt themselves to the stylistic demands of a given situation.The flexible use of mind for mental self-government accountsfor the variety of thinking styles.

Kinds of Styles

The Functions of Mental Self-GovernmentJust as governments carry out legislative, executive, and

judicial functions, so does the mind. The legislative functionof the mind is concerned with creating, formulating, imagin-ing, and planning. The executive function of the mind is con-cerned with implementing and with doing. And the judicialfunction of the mind is concerned with judging, evaluating,and comparing. Mental self-government involves all threefunctions. However, in each person, one of the functions tendsto be dominant.

The legislative style characterizes individuals who enjoycreating, formulating, and planning problem solutions. Suchindividuals like to create their own rules and do things in theirown way. They prefer problems that are not prestructured orprefabricated. A student who must develop his own paper topicwill benefit from having a legislative style. For the studentwith a legislative style, most of the fun of doing a paper is incoming up with the topic and with ideas about the structureand the organization of the paper, as opposed to actuallywriting it.

Individuals with an executive style are implementers.They like to follow rules, figure out which of already knownways they should use to get things done, and prefer problemsthat are prefabricated or prestructured. Whereas the individualwith the legislative style prefers to come up with the topic, theindividual with the executive style prefers to be given thetopic, and then to do the best possible job with it. She wouldprobably prefer, however, a structured test to a paper, and islikely to eschew courses that require a lot of independentwork.

The judicial style is seen in those people who like toevaluate rules and procedures, who like to judge

things, and who prefer problems in which one analyzes andevaluates existing things and ideas. As students, they willprefer analytical essays, such as ones in which they compareand contrast two points of view, or evaluate a point of view.They will also prefer analytical test questions to straightfor-ward memory ones (which are more likely to be preferred byexecutive stylists).

The Forms of Mental Self-GovernmentJust as the functions of mental self-government resemble

those of different branches of government, the forms of mentalself-government resemble forms of government: monarchic,hierarchic, oligarchic, and anarchic.

In the monarchic form, a single goal or way of doingthings predominates. People with a monarchic style tend tofocus single-mindedly on one goal or need at a time. They areeither oblivious to obstacles or are able to cast them aside. Astudent who must turn in a paper by noon tomorrow and is de-termined to do so, but who may face potential distractions, willbenefit from a monarchic style. That student will not allow thedistractions to get in the way of accomplishing his overridingpriority.

The hierarchic form allows for many goals, each ofwhich may have a different priority. People with a hierarchicstyle tend to enjoy dealing with multiple goals, to recognizethat not all goals can be fulfilled equally well, to recognize thatsome goals are more important than others, to have a goodsense of priorities, and to be systematic in their approach tosolving problems. When a student needs to allocate time toseveral homework assignments of varying importance, thatstudent will benefit from having a hierarchic style.

The oligarchic form allows for multiple goals, all ofwhich are equally important. A student who has an oligarchicthinking style will do well in a course that includes severaltests, each weighted equally, or that requires a major paperthat will count just as much as the final exam. Oligarchic indi-viduals relish dealing with multiple and often-competinggoals of equally perceived importance. But they experienceconflict and tension when they have to assign priorities tovarious tasks. Competing goals keep oligarchic individualsfrom completing tasks, because everything seems equally im-portant to them.

December, 1993, Roeper Review/123

Dow

nloa

ded

by [

Mos

kow

Sta

te U

niv

Bib

liote

] at

04:

50 0

6 D

ecem

ber

2013

For individuals with an anarchic style, rules, procedures,and guidelines are anathema. Students who have an anarchicthinking style tend to perform best when tasks and situationsare unstructured, when there are no clear procedures to be fol-lowed, or when the problems they confront are most readilysolved through insights that represent a departure from existingmindsets. People who have an anarchic style generally enjoydealing with a potpourri of needs and goals that are often diffi-cult to sort out. They tend to take a random approach toproblem solving, often seem intolerant or unaware of the needfor rules and regulations, and tend to resist authority.

Levels of Mental Self-GovernmentGovernment also functions at multiple levels, for example,

federal, state, county, and city. In general, one can distinguishbetween more global, a broadly based, and more local, or nar-rowly focused levels. Corresponding to these levels are twoaspects of mental self-government.

A person with a global style prefers to deal with relativelylarge and abstract issues and to ignore details or have someoneelse deal with them. The global person likes to conceptualizeand work in the world of ideas. Metaphorically, he tends to seethe forest, but not always the trees within it. A student with aglobal style will excel on assignments that require the studentto see a big picture—such as in a novel or in an historicalevent or in a scientific theory—but will do much less well onassignments that require the student to deal with the localdetails of the novel, historical event, or theory.

The local sort of person prefers more concrete problemsrequiring detail work, and tends to be pragmatically

oriented. This person relishes the matters that the global personeschews and likes to work on problems that contain manydetails. A student with a local style may enjoy a homework as-signment with many exercises on small problems, but wouldlike less a conceptual essay that requires global analysis of aphenomenon. This person, metaphorically, tends to see thetrees, but not always the forest of which they are part.

The Scope of Mental Self-GovernmentGovernments need to deal both with internal, or domestic,

affairs, and with external, or foreign, ones. Similarly, mentalself-governments need to deal with both internal and externalissues. People who are more internal in style tend to be intro-verted, task-oriented, aloof, and interpersonally less aware thanexternalists. They also like to work alone. Essentially, theirpreference is to apply their intelligence to things or ideas inisolation from other people. An internalist will be happyworking on his or her own completing an assignment but willprobably be less than thrilled with a cooperative learning activ-ity, in which working with other people is a requirement.

People who are more external in their style tend to be ex-troverted, people-oriented, outgoing, and interpersonally moreaware than internalists. They like to work with others and toseek problems that either involve working with other people orthat are about other people. The group or cooperative learningexperience that is so disagreeable to the internalist is exactlywhat the externalist will seek out and enjoy.

The Leaning of Mental Self-GovernmentGovernments can have various political leanings.

Optimally, these leanings are represented on a continuum, suchas right wing to left wing, but for present purposes, two major"regions" of leanings will be distinguished, conservative andliberal. They refer, though, to styles, not politics!

Individuals with a predominantly conservative style liketo adhere to existing rules and procedures, minimize

change, and avoid ambiguous situations when possible. Theyprefer familiarity in life and work. A conservative style doesnot bar a simultaneously legislative one. A person may be bothlegislative and conservative if the person likes to come up withnew ideas and ways of doing things that are essentially conser-vative in bent, drawing heavily upon what has been done in thepast. Great conservative thinkers, such as Edmund Burke,would be examples of conservative legislative types.

A person with a liberal style likes to go beyond existingrules and procedures, maximize change, and find ambiguousand uncertain situations. This person becomes bored whenthings never seem to change. A person may be both executiveand liberal, as in the case of an underling who advocates a veryprogressive point of view. This person likes the progressivestance, but follows rather than leads in it.

Relevance of Styles for Issuesin Gifted Education

How are the various styles of mental self-government rel-evant to important issues regarding the gifted? Consider thefirst of the questions posed near the beginning of the article,and the relevance of thinking styles for addressing each ques-tion. If our goal is to capitalize on students' thinking styles,then we need to take these styles into account in planning andprogramming.

Acceleration versus Enrichment

Almost anyone in the field of gifted education is familiarwith the acceleration versus enrichment debate (e.g., Renzulli,1977; Stanley, 1976). Some who have followed the debatehave concluded that there is no one right answer—thatsomehow, it depends. But on what? From the standpoint ofthinking styles, the answer depends on the style of the student.

With respect to the functions of mental self-government,typical schooling clearly places at an advantage the child withan executive style. The curriculum basically structures whatthe child will learn and how she will learn it. The good tradi-tional student learns quickly and well what she is told to learn.A gifted student with an executive style will prosper in an ac-celerated course, because the student is doing more quicklywhat she does best—working within a structure imposed bysomeone else.

Consider, however, the student with a legislative style.This student probably is no great fan of traditional

schooling, which he sees as keeping him in a lockstep. For thisstudent, enrichment allowing him to do just what he wants willsuit his style extremely well. On the other hand, an acceleratedcourse will not fit this student well at all: It will just be aheavier dose of exactly the same kind of schooling that he hasnever enjoyed.

Another extremely important question is whether weshould place a gifted student in a context that fosters his or herpreferred style or whether we should expose this student to op-portunities to develop other styles. As an example, we canpoint to a gifted executive student placed in a very legislative-like research group in the context of a graduate program.According to her self-report, this experience was definitely en-riching for her, even though she will search for a permanentposition requiring executive and local styles.

124/Roeper Review, Vol. 16, No. 2

Dow

nloa

ded

by [

Mos

kow

Sta

te U

niv

Bib

liote

] at

04:

50 0

6 D

ecem

ber

2013

More generally, legislative individuals often most enjoybeing producers of knowledge, whereas executive individualsmore often enjoy being consumers of knowledge. Accelerativelearning tends to promote a consumer mentality, whereas en-richment tends to promote a producer mentality. Once again,which kind of course a student is most likely to profit from isnot a matter of right or wrong, but of right or wrong forwhom?

Identification of the Gifted

Consider next the identification of the gifted. How ade-quate are conventional multiple choice intelligence tests as in-struments of identification?

The theory of thinking styles has implications for theanswer to this question. With respect to the functions of mentalself-government, the executive stylist will probably benefitfrom the high degree of structure in a multiple-choice test, es-pecially if the test primarily emphasizes recall and straightfor-ward understanding. The test gives the structure, and thestudent works within it. A judicial student can still do wellwith multiple choice if there is emphasis on judgment andevaluation as part of the selection of responses. But a legisla-tive student is likely to feel extremely confined by this testingformat, and is not likely to show to best advantage on this kindof test. Such a student would do better with the push in con-temporary measurement toward performance, portfolio, andproduct assessment. At the same time, however, note thatwhereas this new direction in assessment is ideal for the leg-islative student, it is not ideal at all for the executive one. Weare about to replace one mistake with another, benefiting somestudents and harming others. There is not one right assessmentfor everyone: Different kinds of assessments are needed toassess giftedness in students with different styles.

Quantitative versus Qualitative Difference

Is the difference between gifted students and other stu-dents quantitative or qualitative? From the standpoint of thepresent analysis, the answer depends upon what we shall referto as a synergy between task, ability, and style.

Atask, ability, and style are synergistic when they worktogether to mutual benefit, or are compatible.

Consider, for example, the task of writing a short story for anEnglish class. This is a task that would tend to benefit someonewho is creatively intelligent and who has a legislative style.Were the task a critical essay rather than a short story, then an-alytical ability and a judicial style would be key instead. Aperson whose abilities and style(s) fit a task is in a position toperform that task qualitatively better than a person for whomthere is no match, or a match only in ability, only in style. Aperson with a large number of mathematical computation prob-lems to complete would benefit from mathematical ability andan executive, local style. The absence of either the ability orthe style would render performance on the task less thanoptimal. Obviously, synergy is not required to do a task, oreven to do it well. The argument here is simply that for excel-lence, synergy is required.

Synergy applies to the world of work as well as the worldof school. For example, a novelist would benefit from the samesynergy as the writer of the short story for the English class,and a literary critic or professor of literature would benefitfrom the same synergy as the writer of the critical essay in theEnglish class.

The most gifted individuals are probably those who showa good match between their abilities and their styles. Suchpersons pursue tasks that are well-matched to what they can dobest. In other words, they like to do what they do well. In sum,we can say that excellence involves quantitative differences interms of abilities and quantitative differences in terms of amatch between abilities and styles.

Giftedness in Childhood and AdulthoodIs giftedness the same in childhood as in adulthood? We

lack uniformity of opinion on this matter. On the one hand,scholars such as Bamberger (1986) and Feldman (1986) haveargued that at least some forms of giftedness in childhood arevery different from giftedness in adulthood. Renzulli (1986)has distinguished between "schoolhouse" giftedness and theforms of giftedness more typical of adults, which may havelittle or nothing to do with what leads to high grades in school.Other, such as Feldhusen (1986) and Sternberg (1986), havetaken more intermediate positions, saying that whether gifted-ness is the same or different depends on the kind of giftednessbeing addressed. Consider some specific examples.

Consider the styles of thinking that are most conduciveto success in elementary, secondary, and even college

science. For the most part, students benefit by having an exec-utive, local style. Some science courses consist largely ofmemorization of so-called "facts" in textbooks. Others involvesolving fairly trivial problems at the end of chapters. Even lab-oratory courses often involve little more than the mindless rep-etition of experiments that have been done countless timesbefore. But these styles will not serve the individual in particu-larly good stead as a genuine scientist. Scientists are judgedlargely on the importance, scope, and breadth of the problemsthey choose to address (Zuckerman, 1977). A good scientist isnot only an able problem solver, but an able problem finder(Arlin, 1988). Thus, a real scientist will tend to be served bestby a legislative, global style, rather than by the executive, localstyle that works so well in school. We may be derailing exactlythose students who might be most gifted as scientists, andpassing through our system of filters students who will not beparticularly able scientists. Thus, in the sciences, what consti-tutes childhood and adulthood giftedness will almost certainlybe different.

Unfortunately, the same problem of style mismatch arisesin other areas as well. Anyone who has taught mathematicsknows that the children who do well in mathematics, even upthrough high school and possible early in college, are not thestudents who will necessarily become good mathematicians.The reasons are exactly the same as those for scientists.Mathematicians need to find important problems; they are notsimply given problems at the back of a chapter.

The reader may conclude that the executive and local styleare useless for anything but school. But such is not the case atall. Accountants, for example, will do well with these styles, aswill certain kinds of lawyers. We wish to emphasize again thatstyles are not better or worse, but merely different, and differ-entially appropriate for various kinds of tasks.

Whereas in some domains, the styles needed forsuccess in school seem to be quite different from, or

even opposed to, those required for success in the workplace,in other domains, there is compatibility between styles at thetwo levels. For example, literature is for the most part taught inschools in a way that is compatible with what critics will dolater on. For the most part, students are not taught some "right"interpretation, but rather are encouraged to write essays and

December, 1993, Roeper Review/125

Dow

nloa

ded

by [

Mos

kow

Sta

te U

niv

Bib

liote

] at

04:

50 0

6 D

ecem

ber

2013

express their own interpretations. Even at the elementaryschool level these days there is a tendency to move away frombasal readers toward trade books, and to emphasize the analy-sis of literary texts. Thus, the judicial and hierarchical stylesthat are appropriate in school are the same ones that are appro-priate later on in related professional activity.

In sum, whether adulthood giftedness is the same aschildhood giftedness seems to depend: It depends on

whether the stylistic (and ability) requirements within a disci-pline change or not from childhood to adulthood. When the re-quirements change, the two forms of giftedness are different,but when they do not, the two forms of giftedness may be quitesimilar or even identical.

Independence versus Guidance

Are gifted children best left on their own to find their owninterests and ways of doing things? Or do they need guidanceand direction to do these things? Once again, the answer seemsto depend, at least in part, on styles of thought.

A student with a legislative style will generally seek andenjoy a great deal of independence. This student is likely tofeel frustrated if told what to do, or if given a lot of unwantedand unnecessary guidance. In contrast, the student with an ex-ecutive style will prefer to be given advice, and will thriveunder a system of instruction that involves guidance. Similarly,whether a student profits most from discovery or expositorylearning is likely to depend at least as much upon style as itdoes upon level of ability.

Of course, all students need some guidance. One wouldscarcely want to use styles as an excuse for a laissez-faire ap-proach to teaching. But the optimal amount and type of guid-ance will depend upon the student's pattern of style. Forexample, a student with a monarchic style may need guidancethat will direct her to areas of pursuit other than that which in-terests her most at the moment. She needs to learn not to over-invest in one particular activity at a time when schooling re-quires investment in many activities. A student with an oli-garchic style may need guidance in setting priorities, and inlearning how to form criteria for deciding what kinds of activi-ties are worth more, and less, in terms of time spent in pursu-ing them. An anarchic student may need guidance in harness-ing potential so that it is directed toward particular goals,rather than being spread out all over the place. Such a studentprobably has more potential than some others to become anti-social because of the lack of fit of the school environment tohis style. Appropriate interventions early on might make thedifference between a prosocial and an antisocial future.

Individual versus Cooperative Learning

"Cooperative learning" is one of the latest in a series ofbuzzwords that have stalked the educational establishmentfrom time immemorial. There has long been, and continues tobe, a notion that panaceas and buzzwords can somehowreplace a thoughtful and balanced approach to education. Animportant implication of the approach advocated here is thatoverall panaceas simply don't exist. What is best for onestudent will not be best for another, and so education needs tobe tailored to the individual.

A student with an external style will take easily and wellto cooperative learning. A student with an internal style,however, may well find cooperative learning irksome andcounter productive. Thus, the question is not whether coopera-

tive learning works for gifted students, but rather, for whichgifted students does it work best?

Of course, we do not want only to teach to strengths. Astudent with an internal style may benefit from working withothers and learning how to interact with them, just as a studentwith an external style will benefit from learning how to workalone. The goal is to teach students not just to capitalize ontheir preferred styles, but to compensate for and explore theirless preferred styles.

The Interface Between Thinking Stylesand Assessment

How do the proposed styles interface with assessment ofthe gifted? We consider this issue here.

Assessments are usually targeted at measuring abilities,whether general or particular, or at measuring achievements. Inreality, of course, all assessments measure at least some ofeach. In assessments, there will always be unsystematic mea-surement error, which test-constructors seek to minimize butwhich they can never eliminate. More serious is the existenceof consistent biases, which unlike error of measurement,always affect the same groups in the same ways, helping someand hurting others. The main thesis here is that type of test in-teracts with styles such as to create consistent biases. Byalways using tests of the same kind or kinds, we are alwaysbenefiting people with certain styles at the expense of peoplewith other styles. In essence, we are confounding thinkingstyles, which are psychologically distinct from mental abilitiesand achievements, with these abilities and achievements. Inother words, we are spuriously distorting the test scores in aconsistent fashion.

Consider the various kinds of testing instruments wemight have, what kinds of styles they would favor and disfa-vor, and how these effects might impact what happens in ourschools. We wish to state in advance what the conclusions ofthis analysis will be: Good testing does not use one format oranother, but rather, a combination of formats. Conventionalmultiple-choice tests have unfairly benefitted children withcertain styles but not others; but replacement of such tests withnew, more "authentic" assessments will also benefit childrenwith certain styles and not others. We will be replacing onemistake with another. Testing, like any other kind of scientificassessment, is best when it uses converging operations.

Multiple-choice/short-answer. A widely used form of as-sessment is the conventional multiple-choice/short-answerformat found in almost all standardized tests (both group- andindividually-administered) and most teacher-made tests at theprimary and secondary levels.

Various contemporary scholars have claimed that suchtests cannot measure certain abilities in anything ap-

proaching an adequate way (e.g., Gardner, 1983). Regardlessof the validity of this assertion, we believe that the mainproblem in measuring abilities via multiple-choice tests is notso much inadequacy in the measurement of abilities, but ratherconsistent biases in favor of certain styles.

Conventional tests strongly favor test-takers with an exec-utive, local style, and favor somewhat those with internal andconservative styles as well. Why? First, the examinee is placedin a highly structured, usually very circumscribed kind of situ-ation, where he is given the problem explicitly and told tosolve it within the constraints presented. This is the kind oftesting situation that is ideally crafted for the executive stylist.

126/Roeper Review, Vol. 16, No. 2

Dow

nloa

ded

by [

Mos

kow

Sta

te U

niv

Bib

liote

] at

04:

50 0

6 D

ecem

ber

2013

The problems are usually quite narrow in scope, at least, rela-tive to the larger problems of the discipline, favoring the localstylist. Students are not allowed to work with others; indeed,collaboration is usually considered to be cheating, and there-fore, the internal stylist is favored. Finally, the scoring is ac-cording to a prefixed answer key that represents whatever theconventional wisdom may be for a certain kind of problem, fa-voring the conservative stylist.

Of course, there is nothing wrong with this particularprofile of styles. Styles are not better or worse—just

different. The problem is that if virtually all testing is of theconventional kind, then individuals with one profile of styles,or a profile close to it, will be systematically benefitted, andwill appear "smarter" than individuals with other profiles.Students will be given credit for attributes having nothing todo with ability, or really, even achievement.

What kind of effect could we expect this pattern of fa-voritism to have on our schools? We culd expect to funnelthrough the system students who do what they are told, andusually do it well; who are best at working with problems inthe small rather than in the large; who prefer to work individu-ally rather than with others; and who are fairly conservative intheir approach to problems. And this description fits quite wellthe description of the prototypical "good student" encounteredby many professors who teach "better" students at both the un-dergraduate and graduate levels. But especially at the graduatelevel, we need to ask ourselves whether these are the kinds ofstudents we ideally want. They may have been very good attaking tests well and at getting good grades, but they are un-likely to be the students who are most creative when it comesto generating new ideas in their research, regardless of disci-pline (see Sternberg & Lubart, 1991a; 1991b). Creative stu-dents are more likely to show profiles of styles that are legisla-tive and global, which are styles that are not favored by theconventional tests.

To the extent that more recent tests emphasize criticalthinking and analysis, there will be a shift away from benefit-ing executive stylists toward benefiting judicial stylists. On theone hand, there is clearly value, in its own right, of measuringthe ability to think with the material that is learned. After all,what good is it to know something if one is unable to use it?On the other hand, we are still benefiting certain styles at theexpense of others so long as we do not use converging opera-tions in our measurements.

Essay tests. Who is benefitted by essay tests will dependlargely not upon the test itself but upon how it is scored.Consider three emphases a teacher (or testing company) mighthave in scoring.

One emphasis is on facts presented. An essay test mayreally amount to a recall rather than a recognition

version of a conventional test. If the scorer is basically seekingrecall of "facts" rather than understanding, use, or organizationof these facts, the test then benefits the executive, local, con-servative stylist.

A second emphasis is on interpretation and analysis. Here,the scorer looks for the examinee's ability to criticize and oth-erwise analyze the material learned. Essay examinations in theform of "Compare and contrast..," "Analyze..," and"Evaluate..," are of this form. Here, the judicial style benefitsover the legislative or executive. The global student may alsobenefit if the kinds of evaluations requested are on a larger asopposed to a smaller scale. To the extent that the scorer countsorganization as well, which typically is considered importantin evaluating essays of this type, students with a hierarchical

style will also benefit: Such students tend to organize materialin a way that emphasizes major points, downplays minor ones,and that places the material in a logical sequence.

A third emphasis is on creative synthesis. Here, the scoreris concerned less with the student's ability to repeat back whatshe has learned than with her ability to go beyond the informa-tion learned. Can she add something creative that shows shehas achieved some kind of higher order synthesis that extendsbeyond what was learned in class or in a book? Such an em-phasis is likely to benefit legislative and possibly global stu-dents, if sufficient room is allowed for the student to expressherself on larger issues. Again, to the extent that the scorercounts organization, the hierarchical student will tend tobenefit, or the oligarchic student if the points to be made are ofroughly equally importance.

Projects. Projects are assessments that are done over aperiod of time and that involve some degree of independenceand initiative on the part of the student. The more they requireof each of these, the more they will benefit the legislativestudent. Not all projects are of this kind, however. If theteacher specifies what the project will be on, or gives thestudent only a fairly narrow range of topics, then the legisla-tive student may not be left with much room for initiative. Forexample, Sternberg once went to a "project fair" involving hisson, in which the projects were almost all on states of theunion. Moreover, they all followed roughly the same format,obviously one preset by the teacher. Such projects were scarce-ly one that would allow much room for student initiative.

B e c a u s e there is so much possible information tohandle in doing a project, such an activity will usually

benefit a hierarchic student. But to the extent that there aremany competing demands on the student and his time, so thatit is difficult to allocate to the project the full set of resources itmay ideally demand, the project format may benefit themonarchic student more than it benefits students of the othertypes. If such a student decides that the project is sufficientlyimportant, then He or she will make sure that more than suffi-cient resources are allocated, regardless of competing demandson his time. The same principle holds true for other kinds ofassessments that are resource-intensive but that compete forscarce resources, such as a take-home exam with a tight timelimit.

Individual projects tend to benefit internal stylists, butprojects that are assigned to groups will benefit external styl-ists. Most internalists are uncomfortable with group work, andif the group consists wholly or even mostly of internalists, theymay have difficulty marshaling and integrating resources tomake a collaborative venture a fully viable one.

Portfolio assessment. There is nothing approaching fullconsensus on how portfolio assessment should be done, oreven regarding what should go into a student's portfolio. Someportfolios are little more than collections of test papers,whereas others may represent a collection of students' bestprojects. Moreover, sometimes the student decides whatshould go in the portfolio, other times the teacher, and stillother times, both.

Who will be benefitted from portfolio assessment there-fore depends on exactly how the assessment is done. To theextent that a portfolio is simply a collection of conventionaltest papers, it will benefit the same student who is benefittedby conventional tests. But to the extent that a student is active-ly involved in sorting out a variety of possible materials for in-clusion in the portfolio, and especially project materials, thisform of assessment will benefit the legislative and judicial, hi-

December, 1993, Roeper Review/127

Dow

nloa

ded

by [

Mos

kow

Sta

te U

niv

Bib

liote

] at

04:

50 0

6 D

ecem

ber

2013

erarchical student. The hierarchical style will be key in settingpriorities for what should and should not go in the portfolio.The oligarchic student will find this type of assessment partic-ularly troublesome, in that his difficulties in setting prioritieswill render the task of deciding what to include in the portfolioa particularly daunting one.

Interview assessment. Interviews are routinely used foradmissions purposes, and some examinations are given in anoral form that resembles an interview. In terms of abilities, in-terviews downplay written verbal skills while emphasizingoral verbal skills. But what changes, from the point of view oftheory of styles, when assessment is oral rather than written?What primarily changes is a shift from a more internal to amore external stylistic focus. When writing, one needs to takeinto account the point of view of a potential reader—but at adistance. One can do it in one's head. In an interview situation,one is managing a demonstration of knowledge or abilitythrough a social vehicle, and one inevitably needs to attend tothe social as well as the cognitive aspect of the interview.Having done a great deal of interviewing and at one time evenanalyzed college-admissions interview procedures (Sternberg,1973), Sternberg is convinced that the social aspect of an inter-view is extremely important. Interviewers tend to prefer inter-viewees who are relaxed, who put the interviewers at ease,who are socially as well as verbally facile, and who have somedegree of interpersonal sparkle. The person who sparkles in hiswriting may not in his speech, and vice versa. Thus, the inter-view tends to place the external stylist at an advantage, in con-trast to most other testing media.

Assessment of Thinking Styles

We have discussed some general issues regarding the in-terface between thinking styles and assessment. The questionnow arises as to how thinking styles themselves can be as-sessed.

We have developed a number of converging operationsfor measuring styles, both in students and in teachers(Grigorenko & Sternberg, 1993a). We will describe each inturn.

The Thinking Styles Inventory is a self-report measure inwhich students (or other examinees) rate themselves on a 1-9scale, where 1 is low and 9 is high, for a number of behaviors.Consider examples of some of the items on the inventory:

/ like tasks that allow me to do things my own way[legislative].

I like situations in which it is clear what role I mustplay or in what way I should participate [executive].

I like to evaluate and compare different points ofview on issues that interest me [judicial].

I like to complete what I am doing before startingsomething else [monarchic].

When undertaking some task, I like first to come upwith a list of things that the task will requie me to do andto assign an order of priority to the items on the list [hi-erarchic],

I usually know what things need to be done, but Isometimes have trouble deciding in what order to dothem [oligarchic].

When working on a written project, I usually let mymind wander and my pen follow up on whateverthoughts cross my mind [anarchic].

Usually when I make a decision, I don't pay muchattention to details [global].

I like problems that require engagement with details[local].

I like to be alone when working on a problem [inter-nal].

I like to do things in new ways, even if I am not surethey are the best ways [progressive].

In my work, I like to keep close to what has beendone before [conservative].

We also devised a Thinking Styles Questionnaire forTeachers, which measures teachers' preferences in styles forstudents for seven of the styles. Examples of items in thisquestionnaire are:

/ want my students to develop their own ways ofsolving problems [legislative].

I agree with people who call for more, harsher dis-cipline, and a return to the *good old ways' [conserva-tive].

A third measure is the Set of Thinking Styles Tasks forStudents. The idea here is to measure students' preferences forstyles in actual tasks. Consider two examples:

When I am studying literature, I prefer to(a) follow the teacher's advice and interpreta-tions of authors' positions, and to use theteacher's way of analyzing literature [executive](b) to make up my own story with my own char-acters and my own plot [legislative](c) to evaluate the author's style, to criticize theauthor's ideas, and to evaluate characters'actions [judicial](d) to do something else (please indicate in thespace below)

You are the mayor of a large northeastern city. Youhave city budget this year of $1 million. Below is a listproblems currently facing your city. Your job is todecide how you will spend the $1 million available toimprove your city. Next to each problem is the projectedcost to eliminate a problem entirely. In the space on thenext page list each problem of which you will spend citymoney and how much money you will budget for thatproblem. Whether you spend money on one, some, or allproblems is up to you, but be sure your plan will notexceed the $1 million available. Whether you spend allthe money to solve one or a few problems or divide themoney partially to solve many problems is up to you...

(1) Drug problem ($1 million)(2) Roads ($250 thousand)(3) Landfill ($250 thousand)(4) Shelters for the homeless ($500 thousand)...

The scoring of this problem is in terms of form of mental self-government. A monarchic person is likely to spend all oralmost all the money on one problem. A hierarchic person willdivide the money among problems, spending more money onproblems of perceived importance. An oligarchic person willspend roughly equal amounts of money on each of the prob-lems. An anarchic person will show no system at all, or willnot conform to the constraints of the problem.

128/Roeper Review, Vol. 16, No. 2

Dow

nloa

ded

by [

Mos

kow

Sta

te U

niv

Bib

liote

] at

04:

50 0

6 D

ecem

ber

2013

Finally, we have constructed a measure called Students'Thinking Styles Evaluated by Teachers. In these items, teach-ers evaluate the styles of their individual students. Sampleitems include:

S/he prefers to solve problems in her or his own way[legislative].

S/he likes to evaluate her or his own opinions andthose of others [judicial].

Again, the idea of having a variety of measures is touse converging operations to assess the various

styles. Not only do the kinds of measures differ, but who isdoing the assessment (e.g., student or teacher) differs as well.But do these measures, and the theory underlying them, haveconstruct validity? Consider some of the data we have collect-ed to address this question.

Data Testing the Theoryof Mental Self-Government

Construct Validation Data

An initial study was done with Marie Martin in order toassess the construct validity of the theory and an early versionof the Thinking Styles Inventory. Subjects were 75 collegestudents. The results were promising for a "first pass."

Scale reliabilities ranged from .56 (executive) to .88(global), with a median of .78. Most of the scale intercorrela-tions were low. The exceptions were ones that we anticipated.Correlations greater than .5 in absolute value were global withlocal (-.61), progressive with legislative (.66), conservativewith legislative (-.50), conservative with executive (.59), andprogressive with conservative (-.60). Thus, the legislative andprogressive styles tend to be associate, as do the executive andconservative ones. Global and local styles tend to be negative-ly associated.

Afactor analysis was generally although not totally sup-portive of the structure of the theory. Five factors ac-

counted for the 77% of the variance in the data. A first factorshowed high loadings (greater than .7 in absolute value) for theconservative (.87), executive (.58), progressive (-.81), and leg-islative (-.78) styles. Thus, this factor combined the legislative-executive distinction with the progressive-conservative one. Asecond factor loaded separately for judicial (.70), with a highloading as well for oligarchic (.70). Whereas the legislativeand executive styles are almost diametrically opposed, the ju-dicial is not diametrically opposed conceptually to either, andso its loading on a separate factor makes sense. A third factorcontrasted external (.72) with internal (-.80). The fourth factorcontrasted the local style (.92) with the global one (-.82). Andthe fifth factor showed a high loading for the hierarchic style(.86).

Correlations were computed with the MBTI (Myers-Briggs Type Indicator) as well as the Gregorc measure of mindstyles. For the MBTI, 30 of 128 correlations were statisticallysignificant, whereas for the Gregorc, 22 of 52 were significant.These correlations are well above the levels that would be ex-pected by chance, and suggest that the various style measurespartition a similar space of the intelligence-personality inter-face, but in different ways. In contrast, the correlation of themeasure of mental self-government with IQ was not signifi-

cant, nor was the correlation with GPA. Three styles correlatedsignificantly with SAT Math (judicial, global, and liberal, allpositively), but none with SAT Verbal. Thus, styles do indeedappear to be largely, although not wholly, distinct from intelli-gence or aptitudes.

Data on the Relevance of Styles to the Schools

We (Grigorenko & Sternberg, 1993a) have collected dataassessing the relevance of styles to the schools. We did threestudies in all.

In the first, we examined whether teachers' styles differ asa function of school. Participants in the study were 85 teachers,57 female and 28 male. They were in four middle schools ofwidely differing types. We found several interesting effects.

First, as we predicted, teachers at lower grade levelswere more legislative and less executive than teachers

at higher grade levels. In other words, the teachers at the lowerlevels are more encouraging of a style linked to creativity intheir work with the students (Sternberg & Lubart, 1991a;1991b).

Second, older teachers were more executive, local, andconservative than younger teachers. Of course, we do notknow whether this result is a cohort effect or whether it repre-sents a maturation process. But the study indicates that onaverage, the younger teachers have a style more encouragingof creativity than the older ones.

Third, teachers show some differences in styles acrosssubject-matter areas. Science teachers tend to be local, human-ities teachers more progressive. The latter finding was consis-tent with our expectation, but the former was a bit of a disap-pointment, suggesting as it does a more "molecular" viewtoward science than we would prefer among the science teach-ers.

Fourth, and perhaps most interestingly, we had an inde-pendent rater rate the ideology of each school for each of thestyle dimensions. The idea here was for the rater, who wasblind to our hypotheses, to rate the ideology of the school,using catalogs, faculty and student handbooks, statements ofgoals and purposes, curricula, and related information. Someschools, for example, were rated as ideologically more legisla-tive or more liberal than others. We found that six of sevenplanned contrasts were statistically significant when the actualstyles of the teachers in the school were compared to the ratedideology. In other words, teachers' styles tend to match theideology of their school.

In a second study, we looked at some of the demographicsof styles. We found that father's, but not mother's education,was negatively related to the judicial, local, conservative, andoligarchic styles. Father's occupational level was also nega-tively related to the judicial, local, conservative, and oligarchicstyles. Consistent with common beliefs, younger siblings werefound to be more legislative than older ones. Perhaps most in-terestingly, we found that students tended to match their teach-ers in style. As students could not possibly have been placed inclasses so as to achieve such a match, the results are consistentwith our notion that styles are partially socialized—theydevelop by internalization of styles observed in role models.

In a third study, we queried whether students benefit iftheir styles match their teachers'. In other words, we

know from the last study that there is tendency to match. Butdo students who match actually do better than students whodon't, independently of the students' abilities? We found thatstudents were more positively evaluated by and received better

December, 1993, RoeperReview/129

Dow

nloa

ded

by [

Mos

kow

Sta

te U

niv

Bib

liote

] at

04:

50 0

6 D

ecem

ber

2013

grades from teachers who matched their styles than from thosewho did not. Moreover, teachers tended to overestimate theextent to which their students matched them in styles. In otherwords, people think others are more like them than they reallyare!

In a fourth study, we tested whether gifted students' stylesdiffer from those of nongifted students. We found that on thetests that measure what students say they do, there was no dif-ference between gifted and nongifted children. However, ontests requiring children actually to do tasks, the gifted childrenproved to be more legislative, judicial, and liberal thannongifted children, but less executive.

In conclusion, the set of studies suggest construct validityfor the theory of mental self-government, and also suggest theusefulness of the theory in school settings.

Conclusions

To conclude, there are no "right" answers when it comesto programming for the gifted. The ideal identification and in-structional procedures need to take into account students'styles of thought, as well as their abilities. The answer to ques-tions about gifted programming is usually, "It depends," andwe have tried to show in this article that what it depends on, onpart, is style of thought. We have showed that the theory ofthinking styles applies to gifted children as well as to non-gifted ones, which means that one can identify and educategifted children based on the ideas presented in this article.

If there is a lesson here, it is that we must reward not onlythe styles that we believe pay off in the present, but also thosethat are likely to pay off in the future. Otherwise, we may endup training students to develop styles that will ultimately failthem. Without an understanding of thinking styles, we willcontinue to operate in ways that are educationally ineffective,and even counterproductive. But with such an understanding,we can shape our educational process so that it will benefit allstudents, both in the short run and over the long haul.

For students to benefit from being gifted, they mustlearn not only how to capitalize upon preferred styles

and to compensate for less developed ones, but also to be flexi-ble. Obviously, it would help to be able to measure thesestyles, and we have developed an instrument that does so (seeAuthor Note). Education should foster students' ability to shiftfrom one style to another as the situation warrants. This meansthat teachers must be trained to use, to teach and to modelstyles flexibly as well. To the extent that we take styles as wellas abilities into account in our programming for the gifted, wewill benefit all gifted students.

REFERENCESArlin, P. K. (1989). Problem finding and problem solving in young artists and young sci-

entists. In M. L. Commons, J. D. Sinnott, F. A. Richards, & C. Armon (Eds.),Beyond formal operations II: Comparisons and applications of adolescent and adultdevelopmental models. New York: Praeger.

Bamberger, J. (1986). Cognitive issues in the development of musically gifted children. InR. J. Sternberg and J. E. Davidson (Eds.), Conceptions of giftedness (pp. 388-413).Cambridge: Cambridge University Press.

Feldhusen, J. F. (1986). A conception of giftedness. In R. J. Sternberg and J. E. Davidson(Eds.), Conceptions of giftedness (pp. 112-127). Cambridge: Cambridge UniversityPress.

Feldman, D. H. (1986). Nature's gambit: Child prodigies and the development of humanpotential. New York: Basic Books.

Gardner, H. (1983). Frames of mind: The theory of multiple intelligences. New York:Basic Books.

Gregorc, A. (1985). Inside styles: Beyond the basics. Maynard, MA: Gabriel Systems, Inc.Grigorenko, E. L., & Steinberg, R. J. (1993a). Thinking styles in school settings.

Manuscript in preparation for publication.Grigorenko, E. L., & Sternberg, R. J. (1993b). Thinking styles and their relation to gifted-

ness. Manuscript in preparation for publication.Holland, J. L. (1973). Making vocational chokes: A theory of careers. Englewood Cliffs,

NJ: Prentice-Hall.Kagan, J. (1976). Commentary on 'Reflective and impulsive children: Strategies in infor-

mation processing underlying differences in problem solving.' Monographs of theSociety for Research in Child Development, 41(5), Serial No. 168.

Martin, M. (1988). Mind as mental self-government: Construct validation of a theory ofintellectual styles. Unpublished manuscript.

Myers, I. B. (1980). Gifts differing. Palo Alto, CA: Brooks/Cole.Renzulli, J. S. (1977). The enrichment triad model: A guide for developing defensible pro-

grams for the gifted and talented. Mansfield Center, CT: Creative learning Press.Renzulli, J. S., & Smith, L. H. (1978). Learning styles inventory. Mansfield Center, CT:

Creative Learning Press.Stanley, J. C. (1976). The case for extreme educational acceleration of intellectually bril-

liant youths. Gifted Child Quarterly, 20(1), 66-75.Sternberg, R. J. (1973). Cost-benefit analysis of the Yale admissions office interview.

College and University, 48, 154-164.Sternberg, R. J. (1986). Intelligence applied: Understanding and increasing your intellec-

tual skills. San Diego, CA: Harcourt Brace Jovanovich.Sternberg, R. J. (1988). Mental self-government: A theory of intellectual styles and their

development. Human Development, 31, 197-224.Sternberg, R. J. (in press). Thinking styles and testing: Bridging the gap between ability

and personality assessment. In Stemberg, R. J., Ruzgis, P. (eds.). Intelligence andpersonality. New York: Cambridge University Press.

Sternberg, R. J. (1990). Thinking Styles: Keys to understanding student performance. PhiDelta Kappan, 71, 366-371.

Sternberg, R. J. & Lubart, T. I. (1991a). An investment theory of creativity and its devel-opment. Human Development, 34, 1-31.

Sternberg, R. J. & Lubart, T. I. (1991b). Creating creative minds. Phi Delta Kappan, 608-614.

Witkin, H. A. (1978). Cognitive styles in personal and cultural adaptations: The 1977Heinz Werner Lectures. Worcester, MA: Clark University Press.

Zuckerman, H. (1977). The scientific elite: Nobel Laureates' mutual influences. In R. S.Albert (Ed.), Genius and eminence: The social psychology of creativity and excep-tional achievement (pp. 241-252). New York: Oxford University Press.

Author NoteResearch for this article was supported under the Javits Act Program

(Grant # R2O6R00001) as administered by the Office of Educational Researchand Improvement of the US Department of Education. Grantees undertakingsuch projects are encouraged to express freely their professional judgements.This article, therefore, does not necessarily represent positions or policies ofthe Government and no official endorsement should be inferred.

Information on the measures of thinking styles can be obtained from thesenior author. Elena L. Grigorenko was supported by a fellowship from YaleUniversity. Copies of the instruments may be obtained at cost from theauthors. Requests for reprints should be sent to Robert J. Sternberg,Department of Psychology, Yale University, Box 11A Yale Station, NewHaven, CT 06520.

130/Roeper Review, Vol. 16, No. 2

Dow

nloa

ded

by [

Mos

kow

Sta

te U

niv

Bib

liote

] at

04:

50 0

6 D

ecem

ber

2013