reconceptualizinggenderdifferencesinachievement amongthe … · 2019-08-26 · reconceptualizing...

Reconceptualizing Gender Differences in Achievementamong the Gifted

DAVID LUBINSKI AND CAMILLA PERSSON BENBOW*

Iowa State University, Ames, Iowa, U.S.A.

CHERYL E. SANDERS

University of Colorado, Boulder, Colorado, U.S.A.

Introduction

Overthe past 30 years, manyof the unenlightenedbarri-ers preventing gifted women from achieving educational

credentials and occupational status commensurate withtheir abilities have been removed. In many educationalprograms, comparable gender representation quicklyensued, esyecially in areas like law where many kindsof 4-year ciegrees are acceptable for admissions. Excep-tional performances by womenonbar exams,law schoolgrades anti honors followed, just as the protagonistswho worked so hard to remove the aforementionedbarriers had predictedall along. Gender-comparabilitiesin medical schools, both in representation andin perfor-mance, followed shortly thereafter. This trend servedto reinforce further the well-grounded arguments forremoving vender-discriminating educational barriers tobegin with. That is, argumentsinitially stemming primar-ily from pclitical-ideological concerns now became but-tressed by economic and psychological justification: notonly were women performing admirably in these areas,the disciplines themselves were benefiting from a moreable student population. As a consequence of thegreater number of women with exceptional academiccredential: entering law and medicine, both disciplineshave insur2d that their future leaders and practitionerswill have greater competencies and sophistication.With such progress in mind, the attention naturally

has shifte:i to the physical sciences (our area of con-centration:, where pronounced gender disparities stillremain (Lick & Rallis, 1991; Eccles & Harold, 1992;Maple & Stage, 1991). Could similar benefits accrue forthese discilines if more women entered and maintaineda commitment to physical science educational/voca-

*Addresi correspondence to either Camilla P. Benbow orDavid Lubinski, Co-Directors of the Study of MathematicallyPrecocious Youth (SMPY)at ISU, Departmentof Psychology,

tional tracks? Why has comparable representation in

the math/science pipeline not been achieved? Have

we completely removed from the physical sciences

the barriers that previously prevented women from

entering law and medicine? Are there factors unique

to the math/science pipeline that discourage womenfrom entering and excelling within it? These questions,

among many others, are being investigated throughour research. Here we focus specifically on factorsrelating to educational/vocational choice, exceptionaleducational/vocational achievements and gender dif-ferences within the gifted population. Our research,however,is also aimed at program experimentation andrefinement of well-known educational interventions.Thatis, in working with intellectually talented students,individually and in groups, we attempt to find andprovide environments wherein their talents can bestblossom and cometo their full fruition. Understandingwhat those environments consist of and learning howto provide them are two of the more central goals ofour applied research. We shall draw upon that workas well.Our work with mathematically and verbally preco-

cious youth is particularly relevant to ascertaining thecritical determinants of gender differences in math/sci-ence achievement. Noteworthy professional achieve-mentsin the sciences tend to be within the exclusive pur-view of the highly able—people located within the topfew percentage points of the distribution of intelligence.Given this, our Study of Mathematically PrecociousYouth (SMPY) provides a data bank especially wellsuited to speak to male/female differences in educationalachievement and choice, inasmuchasit contains largeproportionsof individuals, of both genders, who possess

Iowa State University, Ames, Iowa 50011-3180. This work was

supported by a grant from the National Science Foundation(MDR 8855625).

693

Reconceptualizing Gender Differences

TABLE1

The SMPYLongitudinal Study.Its Cohorts of Subjects

When Age when AbilityCohort N identified identified SATcriteria level

Verb. = 370 or 1%1 2188 1972-1974 12-13 Math = 390

Top 1/3 of Talent 0.5%2 778 1976-1979 12 Search Participants

Math = 700 0.01%3 423 1980-1983 <13 Verb. = 630

1983 12 SAT-M + SAT-V = 540 5%Comparison Group

1982 12 500-590 Math 0.5%600-690 Verb.

Math = 500 or 0.5%4 = 750 1987 12 Verb. = 430

Cohort 5 includes 750 students enrolled in top-ranked graduate departments in theU.S.in variousscientific disciplines; they were surveyed at age 23-25 in 1992.

maintained. Finally, as a validity check on ourselectionprocedures, as well as enhancing the longitudinal designfeatures of SMPY, we are conducting a retrospectivestudy of the characteristics and developmental paths ofgraduate students in the top math/science departments inthe U.S. As noted earlier, this is our Cohort 5. Do suchstudents differ in substantive ways from the students inthe SMP’'Ylongitudinal study? If so, we soon will be ina position to detect how.Thus, over 5000 students are currently participating

in the SMPY longitudinal study and soon will increaseto over “000. All of the students in the five cohortsare being surveyedatcritical junctures throughouttheiryouth and adult lives. Moreover, each cohort will besurveyed at the same age to ensure developmentalcomparallity of our cross-cohortfindings. This serves asa rough sketch of the composition of the SMPY databaseand how:t constructed.

The Theoretical Structure Guiding SMPY Research

The conceptual framework for our research draws onthree already existing theoretical perspectives (e.g.,Dawis & Lofquist, 1984; Tannenbaum, 1983, 1986;Zuckerm.in, 1977). We also incorporate someof whatisalready k nown about the developmentoftalent and per-sonal pre “erences for contrasting educational/vocationalpaths. Pr marily, our work is based upon a well-knownmodel of vocational adjustment, the Theory of WorkAdjustme nt (TWA), a model that has been developedover the last 30 years by Rene V. Dawis and LloydH. Lofquist (Dawis & Lofquist, 1984; Lofquist &Dawis, 1969, 1991) at the University of Minnesota.

Oneespecially attractive feature of this modelis that itis readily extendedto critical antecedents of vocationaladjustment, such as choosing a college major.According to TWA,to assess optimal learning and

work environmentsit is useful to parse an individual’swork personality and the environmentinto two broad,but complementary, subdomains. An individual’s workpersonality is primarily comprised of his/her: (1) rep-ertoire of specific skills or abilities and (2) personalpreferences for the content found in contrasting edu-cational/vocational environments. In contrast, differ-ent environmental contexts (educational curricula andoccupations) are classified in terms of: (1) their abilityrequirementsand (2) their capability to reinforce. Opti-mal learning and work environments are then viewedas requiring two levels of correspondence, labeledsatisfactoriness and Satisfaction.

Satisfactoriness denotes the degree of correspondencebetween abilities and the ability requirements of aparticular environment (viz., occupation or educationalcurriculum), whereas satisfaction denotes the degreeof correspondence between the preferences and thetypes of reinforcers provided by an occupation oreducational track. Collectively, satisfactoriness (howthe environment will respond to the individual) andsatisfaction (howthe individualis likely to respond to theenvironment)are useful for predicting the length of timeindividuals are likely to remain in various educationalor career tracks. They are continuous as opposed todiscrete concepts and one’s psychological adjustmentto any given environment at any point in time is, toa large degree, a joint function of these two broadcorrespondence dimensions (see Figure 1). One impli-cation of the modelis that assessing the environment

695

D. Lubinskietal.

[correspondencef— t

Abilities ~ —~ AbilityRequirements

Values |——= Reinforcer

Pattern

Correspondence

PromoteSatisfactoriness

Transfer

AbilityRequirements

Reinforcer

Pattern

Satisfaction

FIGURE1. Thetheory of work adjustment — a depiction

for its requirements and reward capabilities is just asimportantas assessing the individual’s learning and workpersonality (i.e., abilities and preferences). This modelalso stresses the importance of assessing both abilitiesand preferences, concurrently, to ascertain the readinessof a given individual for a particular educational orcareer track (cf. Lubinski & Thompson, 1986).

Personality Structure: Assessing Critical Dispositionsfor Learning Readiness and Efficient Work

ABILITIES

Before assessing abilities, one needs to determine howintellectual abilities are best conceptualized. There isactually a remarkable degree of consensusthatintellec-tual abilities are quite adequately depicted by Guttman’s(1954) early formulation of the Radex (cf. Ackerman,1987; Carroll, 1985; Humphreys, 1979; Lubinski &Dawis, 1992; Snowet al., 1984); a Radex representationof intellectual abilities, taken from Snow’s work,is pro-vided in Figure 2. In this organization, cognitive abilitiesare differentiated along two dimensions, complexity(viz., sophistication of the intellectual repertoire, gen-eral intelligence, or “g“) and content (viz., lower-orderfactors composed of three relatively distinct symbolicsystems: verbal/linguistic, numerical/quantitative andspatial/pictorial). Both are important to assess. In ourwork with the gifted, for example, we have foundituseful to assess the complexity dimension to determine

696

the extent to which educational acceleration is warranted(to provide a more correspondent learning environ-ment), plus lower-order factors to ascertain the precisenature of the acceleration required (thus providing amore individualized and optimal learning environment,responsive to students’ unique strengths). Different“types” of gifted students, for example, verbally vs.mathematically precocious, assimilate certain coursework at different rates and more optimal learningtranspires if curricula are responsive to such individualdifferences.

PREFERENCES

In our research (and as part of our summer programs),the assessment of personal preferences is teamed withability assessment to paint a more comprehensivepic-ture of the unique aspects of each student and ofhow these features of their personality might factorinto educational and career decision making. Studentsalso have found this information useful in consid-ering educational and career possibilities with highschool counselors and parents. Two of the more use-ful schemes for analyzing educational/vocational inter-ests and values are Holland’s (1985) hexagon (con-sisting of Investigative, Artistic, Social, Enterprising,

Conventional, and Realistic vocational interests; seeFigure 3) and the Allport, Vernon, and Lindzey’s (1971)Study of Values (SOV), which is comprised of six valuedimensions(or evaluative attitudes), sharing appreciable


Auditor Letter SpanVisual Number Span

W-Digit Span ForwardW-Digit Span Backward

™ QuantitativeAchievement

LetterVocabulary|Vocabulary) Series

R iti Definitionecognition eriniti Verbal

Anagrams

ConceptFormation

Verbal Listening

Comprehension

Code

Learning ParagraphRecall

ReadingComprehension

Word Beginningsand Endings

Prefixes Suffixes

Synonyms

Addition-Multiplication Subtraction

DivisionNumerical Judgement

Arithmetic ReasoningNumberSeries

Necessary

Arithmetic

Operations

NumberAnalogies

Raven Surface Development Form

Verbal Analogies

GeometricAnalogies

Achievement|

1

Number ComparisonW-Digit Symbol

~ Finding A's- ~ Identical Figures

Paper

1 Board

Paper Folding |

CIC Plotting

W-Block

Design

Breech

Block

W-ObjectAssembly

W-PictureCompletion

FIC‘URE 2. Hypothetical radex map showing suggested ability and learning simplexes and the content circumplex“w” identifies subtests of the Wechsler Adult Intelligence Scale. (After Snow, Kyllonen & Marshalek, 1984.)

overlap with Holland’s model(viz., theoretical, esthetic,social, economic, religious, and political). We assessthese attri.utes as they are useful for identifying optimallearning environments (those likely to be most enjoy-able and rewarding) for gifted students of comparableabilities but who differ in nonintellectual attributesultimately related to career choice.

Environment Structure: Assessing Critical FeaturesofEnvironmental Ecologiesfor Learning and Work

Up to ths point we have talked about the person-ality structure of the individual (abilities and prefer-ences). School and work environments also can beanalyzed ising analogous dimensions. Educational/vo-cational environments may be construed as molecularecologies defined by: (1) their capability to reinforce cer-tain prefeences and (2) the response requirements (orthe abiliti:s) that they demandof individuals. In physicalscience environments the response requirements par- —ticularly iivolve high mathematical and spatial/mechani-cal reasoning abilities but also strong verbal ability, whileinvestigative interests and theoretical values are amongthe most salient personal preferences for gravitatingtoward sc.entific environments, finding the content ofthese disc: plines reinforcing (for developing one’s intel-

lectual talent) and maintaining a commitment towardsuch disciplines (Dawis & Lofquist, 1984; Holland, 1985;Lubinski & Benbow, 1992; MacKinnon, 1962; Roe,

1953; Southern & Plant, 1968). These environmentsrequire intense abilities and preferences for manipulat-ing and working with sophisticated things and gadgetsfor lengthy periodsof time. Individuals with pronouncedor relatively higher social values (or stronger need forpeople contact), in contrast, are not as readily reinforcedin such environments.The above is what we and others have found to be

the person—-environment correspondence structure forengineering and the physical sciences (Dawis, 1991;Dawis & Lofquist, 1984; Lubinski & Benbow, 1992;Holland, 1985; Roe, 1953). Although students are notformally selected for advancedscientific training basedon their theoretical values, their investigative interests,or their spatial and mechanical reasoning abilities (butthey are on mathematical reasoningability), they appearto self-select scientific careers based on all of theseattributes, whether they are explicitly aware of theirabilities and preferences or not (Humphreys, Lubinski,& Yao, 1993). Moreover, an individual will remain in -the sciences to the extent that congruenceis establishedbetween (1) his/her abilities and preferences and (2)the skill requirements and reinforcers provided bythe scientific environment, respectively. Satisfaction

697

D. Lubinskietal.

698

<«

SocialService——»

C Records &

Communications

—_—

—__—pP

B Management

& Planning A Marketing& Sales

WPersonal/

Customer

Services

1 UEducation

& Related

Services VSocial & DATAGovernment $

Services

D Financial

Transaction

E Storage &

Dispatching

NM

T General

Health

Care

PEOPLE

S Applied Arts +>(Written &Spoken)

Q AppliedArts

(Visual)

P Social

Sciences

10

R Creative/ O NaturalPerforming Sciences &Arts Mathematics

I Agriculture

F Business Machine/

Computer Operation

G Vehicle Operation& Repair 6

H Construction &

Maintenance

- .

m ceeneural ——. J Crafts &

Z —— Related@ KHome-Business Services

pauipment L Industrial. “pair EquipmentN Medical O ;«ase: perationSpecialities & Repair 7

& Technology PM Engineering& RelatedTechnologies

R

=[eotuyooy—e

locations, most are located near the point shown.

and THINGS (~—®*) or DATAand IDEAS( { )

* The World-of-Work Map arranges job families (groups of similar jobs) into 12 regions.Together, the job families cover all U.S. jobs. Although the jobs in a family differ in their

° A job family's location is based on its primary work tasks—working with DATA, IDEAS,PEOPLEand THINGS. Arrows show that work tasks often heavily involve both PEOPLE

¢ Six general areas of the workworld and related Holland types are indicated around the edge ofthe map. Job Family Charts (available from ACT) lists over 500 occupations by general areajob family, and preparation level. They cover more than 95% of the labour force.

FIGURE3. World of work map


(need-reinforcer correspondence) and satisfactoriness(ability-ability requirement correspondence) are essen-tial for optimal intellectual development; achieving bothis the central goal of SMPY’s programmatic work withgifted youth.

OPTIMA!1. EDUCATIONAL CORRESPONDENCE FOR

THE EXTREMELY GIFTED

A proper response to this topic requires, first of all,a full appreciation of the range of interventions thatneed to be considered when creating optimal learningenvironments for the exceptionally gifted. That is, ifgiftedness is arbitrarily defined as being in the top1%, individual differences in IQs among the giftedrange from approximately 135 to over 200 (roughlyone-third of the entire IO range). Paralleling this vastability rarige is an equally wide spectrum ofideal learningenvironments. Because learning environments can rangefrom discorrespondent to optimally correspondent, akey component of our research is designed to uncoverunique ways to enhance the learning experiences andintellectual development of the exceptionally able—tomake it us optimally correspondent as possible. Wesuggest that the work of Harriet Zuckerman (1977)provides lues for how to enhance educational corre-spondence: among the exceptionally able. For this partof our thinking, we blend Zuckerman’s theory on theaccumulation of advantage with Tannenbaum’s (1983,1986) work on the critical elements for world-classachievement.Zuckerman (1977) studied the career paths of Nobel

Laureates and occupants of the “forty-first chair” (sci-entists generally acknowledged to have doneresearch ofNobel prize quality, but not awardedthe prize). Theseindividuals almost universally show promise extremelyearly in their careers and this evidenced precocityappears not only to respond to but also to creategreater Opportunities for intellectual development. Forexample, most Laureates receive an advantagein gradu-ate work by attending the most distinguished universities(10 universities produced 55% of the laureates) andby studying with the best minds of the day—otherNobel Laureates or occupants of the 41st chair—therebybegetting a pattern of eminence’s creating eminence.Zuckermin claims that the development ofscientifictaste, standards andself-confidence are the most benefi-cial resul:s of the Laureate’s apprenticeships (cf. JulianC. Stanlev, 1992). |

Moreover, future Noble Laureates obtain degreesand start publishing earlier and more copiously thanother scientists. Soon, by the quality of their scientificcontributions, they become distinguished from theirage-equivalent peers. This opens up further oppor-tunities for their development. Zuckerman suggestedthat the descriptions of Nobel Laureates’ careers fitswell with the modelof “the accumulation of advantage:the spiraling of augmented achievements and rewardsfor individuals and a system ofstratification that is

sharply graded” (p. 249). Moreover, almostall futureNobel Laureates were “active” in creating this beneficialenvironment(cf. Scarr & McCartney, 1983).

Thus, among the gifted, it would seem that those

who have the personal potentialities for manifestingexceptional achievementrequire special encounters withthe appropriate environmentto facilitate the emergenceof world-class accomplishments. Consistent with thisview, Bloom (1985) noted from interviews of talentedperformersin a variety of disciplines that special experi-ences, sometimes interventions, are important for thedevelopmentof talent. Moreover, Tannenbaum (1983,

1986) postulated that great performance or productivityresults from a rare blend of superior general intellect,distinctive special aptitudes, the right combination ofnonintellectual traits, a challenging environmentand thesmile of good fortune at crucial periodsoflife. (The firstthree components seem to parallel the abilities and pref-erences discussed in the Theory of Work Adjustmentand the latter two the work of Zuckerman.) Accordingto Tannenbaum, success depends upon a combinationof facilitators, whereas failure may result from even asingle deficit. By virtue of its “veto” power, then, everyone of the five qualifiers is a necessary requisite of highachievement and noneof them hassufficient strength toovercome inadequacies in the others.The abovediscussion presents the scaffolding for our

work on the dispositional determinants of contrastingeducational/careerpaths of the gifted and, thus, leads tothe conclusion that individuals whoare ideally suited forcareers in the physical sciencesare gifted individuals withhighly developed mathematical and spatial/mechanicalreasoning abilities and intense investigative/theoreticalpreferences. It is these individuals who will choosecareers in the physical sciences and engineering andremain committed to them. Gifted individuals withother ability and preference profiles will choose careersin other areas. Given this line of reasoning, a naturalconsequence of educational and career counseling basedon abilities and preferences will be disparate male/fe-male ratios in academic and vocational choices, not onlyin the math/science disciplines but other disciplines aswell. Moreover, these differences should intensify atthe higher educationallevels. Evidence supporting theseconclusionsis presented next.

Gender Differences in Abilities/Preferences: TheirImplications According to TWA

Abilities

Recent reports seem to indicate that certain genderdifferences in cognitive abilities are steadily diminishingin normative samples (Feingold, 1988; Hyde, Fennema,& Lamon, 1990; Rosenthal & Rubin, 1982). Thatis,males and females appear to be converging toward acommon meanon a variety ofintellectual abilities. Thesetrends, however, have not been noted among the most

699

D. Lubinskietal.

able (Benbow & Lubinski, 1992, in press; Benbow &Stanley, 1980, 1983; Lubinski & Benbow, 1992; Stanleyet al., 1992). Amongthegifted, there are sizable genderdifferences at age 13, favoring males, in mathematicalreasoning and in spatial and mechanical reasoningabilities—the very abilities required of the physicalsciences. Moreover, at the end of high school and

college, these differences remain and accompany genderdifferences favoring males in math/science achievementtest scores (as well as other test scores), whereas femalestend to doslightly better than males on a number ofverbally oriented achievement tests (Stanley et al.,1992). Before profiling these differences in somedetail,we will address first the question: how can normativemale/female means be converging while gender differ-ences amongthe gifted remain pronounced? There areat least two possible explanations and probably bothoperate to a degree: test construction practices andgenderdifferencesin ability dispersion.

First, Stanley et al. (1992) have remarked thatit isdifficult to assess changes in group performance oncognitive tests over the last few decades, inasmuch as anumberoftest publishers may have routinely culled fromtheir instruments items that characteristically generatethe most conspicuous gender differences—a procedurethat some refer to as correcting for “gender bias” or“equity in testing.” Thus, it is possible that the apparentconvergence of male/female group meansis duetotestconstruction practices as much as, or perhaps evenmore than, a genuine changein the cognitive attributespurporting to be assessed by these measures.

Second,if meta-analytic reviews are indeed detectinga degree of genuine gender-convergence, consumersof

meta-analytic reviews must keep in mind that thismeth-odology provides information only on groupdifferencesin overall level of the attribute under analysis. Meta-analytic reviews do not provide information on groupdifferences in other statistics such as those indexingability dispersion. There are other parameters on whichthe genders can differ and a critically important oneisvariability (cf. Benbow, 1988).Many lines of evidence have converged to suggest

that males are more variable than females on a varietyof intellectual variables and, interestingly, this appearsto hold even for variables on which females havesuperior means. This phenomenon has been observedover several decades in normative samples(cf. Feingold,1992; Lubinski & Benbow, 1992; Lubinski & Dawis,1992; Lubinski & Humphreys, 1990a, 1990b). Table 2,for example, consists of data from Project TALENT(Flanaganet al., 1962). Project TALENTcontains datafrom stratified random sample of U. S. high schools,collected back in 1960; this data bank contains fourgrades of students, 9 through 12, with approximately100,000 students in each grade. A number ofabilityand preference measures were administered to thesestudents over the course of several days of testing.Four composite measures (viz., English language, spa-tial visualization, mathematical reasoning, & generalintelligence) are assembled in Table 2; each meanand standard deviation represents approximately 50,000subjects. It is clear that even for the English languagecomposite, on which females are clearly superior as agroup, the males across all four grades were morevariable on this measure as wellas all the others.A more contemporary exampleis provided by Stanley

TABLE 2

Meansand Standard Deviations for Four Ability Composites.From Project TALENT for Grades 9-12 by Gender

English language Mathematics Spatial IntelligenceM. S. M, S.. M, S. M, S,

Grade 9Females 87.65 17.29 15.35 6.43 60.15 20.07 145.36 52.58Males 79.51 18.11 16.01 6.98 69.18 22.54 142.48 54.22


Grade 11

Females 96.63 16.91 17.77 8.02 66.22 20.94 169.97 52.13Males 89.28 17.89 20.69 8.90 79.01 22.96 173.56 53.95


Sample sizes for each cohort by gender follow: grade 9, females = 49,393, males = 49,968;grade 10, females = 47,119, males = 48,543; grade 11, females = 45,428, males = 43,851; grade12, females = 40,116, males = 38,392. Detailed descriptions of these ability composites may be

found in Lubinski and Humphreys (1990a).

700


et al. (1992). These investigators again report that malestend to be more variable on measures of cognitivefunctioning, including tests for which females havehigher means. For example, Stanley et al. (1992) notedthat the largest gender difference favoring females onthe Differential Aptitude Test (DAT) is observed inDAT-Speling. Grade 12 females score approximately.5 standard deviations above the males on this measure.Alternatively, one may state that only 30% of malesscore above the female mean; yet, because of greater

" male variability, there is a comparable male/female pro-portion among students within the top 1% in “spellingtalent.” This finding and the general phenomenon ofgender differences in ability dispersion has importantimplications for understanding male/female differencesat exceptional levels of achievement (cf. Lubinski &Dawis, 192). When assessing gender differences inachievement amongthegifted, it is the uppertail of theability dis‘ribution that we are evaluating; and this uppertail contains an inordinate number of males. Moreover,genderdifferences in dispersion and level often operatein concert to produce especially disparate male/femaleratios at the extremes, as we will illustrate next byreturning to SMPY’s work with the mathematicallytalented.

In nationwide talent searches in the U.S., discussedpreviousl”, gifted students taking the College BoardScholastic Aptitude Test (SAT) have consistently gener-ated the fisllowing pattern of scores. Gender differencesin SAT-V are typically small. Yet on SAT-M thedifferenc: between means approximates .4 standarddeviations, favoring males, and males are morevariablethan females. Together, these genderdifferencesin leveland dispersion produce the following male/female ratiosfor these | 2- to 13-year-olds: SAT-M — 500 (average scoreof college-bound 12th-grade males), 2:1, SAT-M — 600(83rd percentile of college-bound 12th-grade males),4:1, and SAT-M — 700 (95th percentile of college-bound12th-gracle males), 13:1 (Benbow & Stanley, 1983).Compara)le ratios have been replicated across the U.S.in a number of talent searches across several years, aswell as in other cultures.” Score ranges at SAT-M -500 are important for a 12-year-old to consider. Theyreflect important individual differences in quantitativesophistication (Benbow, 1992) and mark the level atwhich successful graduate workin the physical sciencesat the very best universities begins to become probable.The above gender difference in mathematical reason-

ing ability does not operate in isolation, however, tosolely produce the profound genderdisparities in edu-cational attainment and pursuits along the math/sciencepipeline. Gender differences in other abilities requiredby the physical sciences, especially spatial and mechani-cal reasoning, amplify the disparities. These abilities are

*Within Asian samples, male/female proportions at theextremes »f mathematical talent are narrower than those

reported for Caucasians (Stanley et al., 1989). For example,

frequently overlooked by investigators trying to cometo grips with the under-representation of women inengineering and the physical sciences. Table 3 containsdata that bear on this issue. They were collected onCohort 4 by SMPY at Iowa State University overthe last 4 years. In addition to the SAT, studentsin Cohort 4 are administered a variety of nonverbaltests including Raven’s Advanced Progressive Matrices,three-dimensional spatial visualization and mechanicalreasoning. (A numberof personal preference question-naires are administered as well, see below.) Gifted stu-dents at or abovethe cutting score for the top 1%in over-all mathematical reasoning ability display trivial genderdifferences in not only SAT-V but also in AdvancedRaven scores. Yet significant gender differences arerevealed for spatial ability and mechanical reasoning.This also parallels the findings of Stanley et al. (1992).These investigators analyzed gender differences on theDifferential Aptitude Test (DAT)in effect-size units,taken from a national sample of over 61,000 students.The most pronounced genderdifference in this batterywasobserved in Grade 12 on the Mechanical Reasoningmeasure, the male-female effect-size difference was

almost a full standard deviation (.89) favoring males.These gender differences in spatial and mechani-

cal reasoning abilities, combined with the well-knowngender differences in mathematical reasoning ability(Benbow,1988), help explain why disparate male/femaleproportions are observed all along the math/sciencepipeline. The satisfactoriness criterion for engineeringand manyof the physical sciences is not as frequentlymet by gifted females as males. This is only part ofthe picture, however. There are gender differencesin nonability personal attributes (vocational interestsand values), in addition to life style preferences, thatexacerbatedisparities stemming from genderdifferencesin satisfactoriness for the physical sciences. We turn tothem next.

PREFERENCES

As noted earlier, physical scientists are characterizedprimarily by their high theoretical/investigative pref-erences (MacKinnon, 1962), coupled with a relativelylow need for people contact. Both mathematicallygifted males and females have, relative to their ownsex norms, strong theoretical values and investigativeinterests. Yet, there are prominent gender differencesin critical preferences for maintaining a commitmentto careers in the math/science pipeline that mirror theaforementioned spatial/mechanical abilities among thegifted at age 13. Mathematically talented males are moretheoretically oriented on the SOV (see Table 3). Fur-ther, their primary interests lie in the investigative and

at SAT-M > 700, the male/female Asian ratio is closer to 4:1,

whether students assessed live in the United States or not.

701

D. Lubinskietal.

TABLE3

Ability Profiles of Mathematically Gifted Students Attending a Summer Academic Program AcrossFiveSeparate Years by Gender

Year Gender

Age-Adjusted Advanced Mental BennettSAT-M SAT-V Raven’s Rotation Mechanical_ _ _ Test Reasoning

N XxX SD X SD N XxX SD N X SD N XX sp

» M 72 494 93 398 91 72 £246 41 #72 #42313 47.7 72 #498 #741999 F 45 458 66 396 82 45 243 44 45 23.7 88 45 £453 52

» M 84 486 91 395 88 85 244 42 83 316 7.5 83 49.9 7.0F 49 465 76 404 90 47 246 45 48 #241 89 49 456 52

» M 68 532 101 426 78 68 #4«2925:1 39 68 #=299 841991 F 51 480 87 418 87 51 258 43 51 £25.11 10.2

+ M 107 579 101 413 81 92 25.2 42 95 300 81F 67 472 85 418 80 58 25.9 42 63 24.1 10.0

»« M 69 537 100 415 79 69 #4245 65 69 202 94F 48 487 74 422 76 48 253 44 48 22.5 97

1990 + M 87 S45 96 415 79 82 246 68 80 298 8.8F 61 487 71 419 80 57 425.1 41 #56 £4216 9.4

»« M 20 585 86 441 98 20 273 44 20 249 #99 20 +402 94F 11 505 980 449 96 11 247 51 #11 #178 #421 «11 #42935.6 80

1989 »; M 43 593 95 446 78 21 270 44 40 238 9.7 42 42.2 10.0F 34 514-82) 455) 79s 247 584272 32s 35D——«O2'*G

» M 57 562 81 435 59 57 £42266 3.81988 F 32 491 65 424 80 32 £42251 «5,3

» M 72 S71 85 440 62 66 26.8 3.7 8 39.3 6.5F 39 500 64 425 76 36 #4225.3 5.3 9 29.0 7.2

* Students whotookall of thetests.+ All students whotook anyonetest.

(secondarily) the realistic sectors of Holland’s Hexagon(Benbow & Lubinski, 1992; Fox, Pasternak, & Peiser,1976). In contrast, mathematically talented females aremoresocially and esthetically oriented and haveintereststhat are more evenly divided amonginvestigative, socialand artistic pursuits (Table 3; Benbow & Lubinski,1992; Fox, Pasternak, & Peiser, 1976). Females aremore balanced andless narrowly focused in terms theirinterests and values. (One couldalso say this about theirabilities, cf. Lubinski & Benbow, 1992.) Consequently,the TWAsatisfaction criterion is less often achievedfor females than males when considering the physicalsciences.

Thus, at age 13 more males than females possessability and preference profiles that are congruent withchoices to pursue highly focused careers necessary fordistinction in the physical sciences. Due to their moreevenly distributed preferences and abilities, the careerchoices of mathematically gifted females and the amountof time they devote to scientific careers will be lessdistinguished than their male counterparts. Males willbe more exclusively committed to the sciences, whilefemales will have competing interests and will tend todevelop their talents in relatively equal proportionsacross artistic, social, and investigative educational/vo-

702

cational domains. That is exactly what is found inour educational programs designed for adolescents inthe top 1%in ability. Females enroll in courses inmath/science and English/foreign languagein essentiallyequal proportions, whereas males were approximatelysix times morelikely to enroll in math/science areas thanin English/foreign languages. TWA would predict thatthe samepattern will reveal itself when career choices,made at a later age, are examined. Indeed, this is thecase. Table 4 provides data on the secured educationalcredentials that mathematically gifted students in Cohort1 achieved (or are intending to achieve) 10 yearsfollowing their identification at age 13. Less than 1%of the females in the top 1% of mathematical abilityare pursuing doctorates in mathematics, engineering,or physical sciences (Lubinski & Benbow, 1992). Eighttimes as many males are doing so. Benbow and Lubinski(1992) presented similar data just collected for SMPY’sCohorts 2 and 3.An alternative way to capture the essence of these

gender differences in preferences takes us back toThorndike (1911) and one of the most celebrateddimensions of individual differences, “people versusthings.” In normative samples, females tend to gravitatetoward the former, while males gravitate towards the


TABLE4

Preference Profiles of Mathematically Gifted Students Attending a Summer Academic Program AcrossFive Separate Years by Gender

Year GenderTheoretical Social

N x SD X SD

Economic

Study of ValuesAesthetic _Political Religious

SD X SD X SD X SD

. M 7 46.7 7.1 35.7 68F 45 41.5 82 44.0 7.4

1992 , M 73 467 7.1 35.7 68F 45 41.5 82 44.0 7.4

. M 68 47.7 7.0 371 7.3F 51 420 68 43.2 81

1991 , M 77 476 69 37.1 7.0F 57 41.7 7.0 438 8.3

. M 69 4.6 88 38.4 7.81990 F 48 403 8.0 44.0 8.099 , M B 46.6 8.7 383 7.6

F 51 40.7 80 43.6 81

. M 2 493 7.4 35.4 5.9F i 39.0 91 423 9.1

1989 , M 43 50.0 6.8 348 7.5F 34 418 74 41.2 83

. M 57. 48.0 85 34.4 7.8F 32 423 7.5 40.7 8.0

1988 M 61 483 85 345 7.6F 33 425 7.4 40.9 8.0

41.637.8

7.1 36.7 7.1 440 67 33.2 10.96.7 43.6 67 37.4 5.9 34.2 10.07.1 366 7.1 440 67 33.5 11.06.7 43.6 6.7 37.4 5.9 34.2 10.0

7.2 364 82 42.9 66 34.2 10.469 42.6 7.1 39.0 7.2 35.4 10.269 365 83 43.1 68 33.8 10.17.0 42.88 7.5 38.7 7.0 35.6 10.3

8.2 38.4 84 425 69 33.4 11.147.1 42.1 64 40.1 67 375 8.18.1 37.8 8.7 42.7 68 33.9 11.37.2 42.8 7.1 40.1 66 37.1 8.4

94 373 80 45.0 7.8 30.8 11.196 406 5.2 404 9.3 366 12.5

42.2 8.2 37.0 7.7 441 82 30.9 10.777 43.9 82 39.2 7.2 343 10.9

449 7.6 35.3 8.1 45.2 82 32.4 12.87.5 43.6 84 40.1 6.2 349 10.37.4 35.0 80 448 8.3 32.9 12.7

380 7.5 43.4 84 40.0 62 35.2 10.2

* Students who tookall of the tests.+ All students who took anyonetest.

latter (cf., Lubinski & Humphreys, 1990a) and thisparamete: of individual differences operates among thegifted as well (Lubinski & Benbow, 1992). Given thefemale preference within the sciences for biology andmedicine, however, compared to the physical sciences,as is evident in Table 4, perhaps it would be more preciseto state tht gender differences in vocational preferencesare structured around “organic” versus “inorganic”content domains (Benbow & Lubinski, in press). Itis not sc.ence, per se, that turns off many females,rather, it seems to be the inorganic nature of many ofits conten: domains. Weare currently investigating keyvalue con;igurations (high theoretical values, relativelylow socia!, values), which we believe more precisely mapthe individual differences that contribute to these careerdecisions. Our preliminary findings indicate that thehigher-order trend, theoretical minus social as assessedat age 13, has predictive validity for structuring choice ofcollege major and areas of graduate concentration.

Life-sty'e choices: Before leaving the domainof pref-erences, there is one critical gender difference in life-style preference that is essential to document (and onethat is typically not assessed on standardizedinterest orvalues questionnaires). This gender difference is likelyto exert i huge effect on gender differences even in

disciplines in which male/female ratios in achievededucational credentials are comparable: commitmentto full-time work as young adults. In ourfirst threecohorts, for example, about 95% of mathematicallytalented males versus less than 60% of such femalesplan to work full-time until retirement, a percentagethat has been stable over the past 20 years (Benbow& Lubinski, 1992, in press). This latter statistic wouldindicate that females, as a group,will tend to devotelesstime to their vocational developmentrelative to males.Further, in most research in this area as in our previousresearch, questions to respondents are typically framedin termsoffull-time versus variouspart-time options, notin terms of how muchtheyare willing to work. Thus, weare currently assessing how the gifted feel about 50- to70-hour work weeks, schedules morein line with peopleat the cutting edge of their discipline. This might revealfurther genderdisparities.

In sum, therefore, mathematically gifted females, inaddition to having a more multifaceted interest profileand a more complex mixture of value orientations forevaluating their experiences and structuring their life-style, prefer to devote less time to vocational pursuits.They have more to balance, more competing needsatcomparable intensities. Mathematically gifted males,

703

D. Lubinskietal.

however, are more focused on a theoretical/investigativeStyle of life with fewer competing pulls and prefer todevote a greater amountof time to vocationalpursuits.

Contemporary Research Trends: Viewed From theContext of TWA

Wehaveillustrated above how the personal attributesof females compared to maleswill lead them to choosescientific careers less frequently, as a group, and todistribute their educational developmentacrossartistic,social and investigative areas more evenly. Of course,conventionally purported barriers (Eccles, 1985; Eccles& Harold, 1992; Kerr, 1985; Noble, 1989; Reis &Callahan, 1989; Silverman, 1986) might still remain.We suggest, however, that these and other purportedbarriers be evaluated from a broadercontext using TWAand its key components,satisfaction andsatisfactoriness,to establish expectations on the degree to which com-parable gender representation might be anticipated.According to TWA,because of their differing abilityand preference profiles, highly able males and femalesachieve, as a group, Satisfaction and satisfactorinessthrough different means. Consequently, they respondto the environment differently, as does the environmentto them. Gender differences as a reflection of choiceneedto be factored into existing models and perceptions.Moreover, satisfactoriness and satisfaction baselinesmight be useful for appraising theoretical explanationsof gender differences in achievement moregenerally.For example, Eccles et al. (1983) introduced the

expectancy-value model of motivation, which they pro-posed as a framework for understanding the relation-ship between values/personality attributes and academicachievement, as well as genderdifferences therein. Thispopular model describes two primary factors affectingachievement behavior: (1) expectations for success andfailure and (2) subjective task value. Moreover, itconceptualizes gender differences in achievement, justas we do, from a choice perspective rather than adeficit perspective. It also views choices as being madefrom a variety of options presented within a complexsocial reality, wherein gender roles and stereotypesoperate (see Eccles & Harold, 1992, for a discussionof this model as it pertains to the gifted). Yet towhat extent do expectations and subjective task valuereflect realistic personal estimates of the degree to whichTWA’s correspondence dimensions, satisfactoriness andsatisfaction, respectively, are achievable in contrastingeducational and work environments? Is self-confidenceand self-efficacy, for example, merely a reflection of theextent to which oneis in a correspondent environmentalecology in the TWAsense?*

Inadequate math/science preparation is anotherfactoroften mentioned as curtailing women’s career options.

*Moreover, amongthe gifted, their self-concepts, especiallyacademic ones, are so high in both males and females that

704

Sells (1980), for example, perceived mathematicsin highschoolas a “critical filter,” screening out females fromengineering and science majors. Indeed, the numberofmathematics and science courses taken in high schoolis found to relate to choice of college major as wellas career (Berryman, 1985; Ethington & Wolfe, 1986);gifted females, even mathematically gifted ones, dotake somewhat less mathematics (and science) in highschool compared to such males (Benbow & Minor, 1986;Benbow & Stanley, 1982). Consequently, it has beenSuggested that more females would enter and remainin the math/science pipeline if they were required totake more mathematics and science courses in highschool. But to what extent is course-taking amongmathematically gifted females a reflection oftheir pref-erences and/or abilities? Preference and ability profilesare in place long before high school. Can they bechangedasa function of course-taking? Would requiringmathematically gifted females with intense preferencesfor social and artistic content to take more mathemat-ics, chemistry, physics, and computer science in highschool increase their representation in the math/sciencepipeline? Would requiring mathematically and spatiallygifted boys with intense preferences for building andmanipulating physical materials and inorganic things totake more high school courses in English increase theirrepresentation in the humanities?Some of the other factors thought to attenuate the

professional development of womeninclude those thatwomen “do to themselves.” These are, among others,the Queen Bee Syndrome(Staines, Tavis, & Jayaratne,1974), the Great Impostor Syndrome (Clance, 1985;Machlowitz, 1982; Warschaw, 1985), the CinderellaComplex (Kerr, 1985) and the Perfection Complex(Reis, 1987). The Queen Bee and Perfection Complexare similar in that females feel that they have tobe perfect in every way and in how they handlethe multiple roles of professional, mother, and wife.The Great Impostor Phenomenon captures how manysuccessful womenfeel they achieved their success—notthrough their hard work and ability but rather throughluck; and they are waiting to be found out. To avoidbeing “found out”they get caughtin a circle of workingeven harder, achieving greater success and developinggreater fear of being detected as the impostorthey trulybelieve they are (Kerr, 1985, 1991). To what extentis this depiction characteristic of a gifted female withseveral competing interests at comparable intensitiescoupled with the greater conscientiousness of femalesin general (Schmidt & Hunter, 1992)? These questionsand others like them will be pursued in our futureresearch. In all of our research, however, detailedassessments of well-known personal attributescriticalfor satisfaction andsatisfactoriness in specialized careersand advanced educationaltracks are conducted,not only

measures derived on normative samples are psychometrically

inadequate dueto ceiling effects (Swiatek, 1992).

D. Lubinskietal.

responsiveto their individual differences and uniqueness |as possible. If one is to be all that one can be, to borrowMaslow’s phrase, one must be responsive to one’strue nature—a theme that cuts across many fulfillmenttheories and formulations aimed at construing optimalforms of human functioning, including those of GordonAllport, Carl Rogers, and Carl Jung. It might beadvisable for counselors and educators to keep thewisdom offered by these theorists in mind when workingwith clients and students searching for optimal directionfor their educational and vocational development, and,perhaps also, to remind clients of Jane Loevinger’s(1976) observation (contained in her treatment of egodevelopment) that “personality develops by acquiringsuccessive freedoms.” Yet all of this actually can beencompassed by TWA,an empirically based model ofpersonal fulfillment within the world of work (Dawis &Lofquist, 1984; Lofquist & Dawis, 1969, 1991).

References

Ackerman,P. L. (1987). Individual differencesin skill learning:An integration of psychometric and information proces-sing perspectives. Psychological Bulletin, 102, 3-27.

Allport, G. W., Vernon, P. E., & Lindzey, G. (1970). Manualfor the study of values. Boston: Houghton-Mifflin.

Benbow, C. P. (1988). Sex differences in mathematicalreasoning ability amongthe intellectually talented: Theircharacterization, consequences and possible explanations.Behavioral and Brain Sciences, 11, 169-232.

Benbow,C. P. (1992). Academic achievement in mathematicsand science between ages13 and 23: Are there differencesamong students in the top one percent of mathematicalability? Journal of Educational Psychology, 84, 51-61.

Benbow, C. P., & Lubinski, D. (1992). Gender differences

among intellectually-gifted adolescents: Implications forthe math/science pipeline. Invited address at the annualconvention of the American Psychological Society, SanDiego.

Benbow, C. P., & Lubinski, D. (in press). Consequencesof gender differences in mathematical reasoning ability:Some biological linkages. In M. Haug, R. E. Whalen,C. Aron, & K. L. Olsen (Eds.), The developmentof sexdifferences and similarities in behaviour. Kluwer AcademicPublishers in the NATO Series: London, England.

Benbow, C. P., & Minor, L. L. (1986). Mathematicallytalented males and females and achievement in highschool sciences. American Educational Research Journal,23, 425-436.

Benbow, C. P., & Stanley, J. C. (1980). Sex differencesin mathematical ability: Fact or artifact. Science, 210,1262-1264.

Benbow,C.P., & Stanley, J. C. (1982). Consequences in highschool and college of sex differences in mathematicalreasoning ability: A longitudinal perspective. AmericanEducational Research Journal, 19, 598-622.

Benbow, C. P., & Stanley, J. C. (1983). Sex differences in

mathematical reasoning ability: More facts. Science, 222,1029-1031.

Berryman, S. E. (1985). Minorities and women in mathematicsand science: Who choosesthesefields and why? Paperpres-

ented at the annual meeting of the American Associationfor the Advancementof Science, Los Angeles.

706

Bloom,B.S. (1985). Developing talent in young people. NewYork: Ballentine.

Carnap, R. (1950). Logical foundations of probability.Chicago: University of Chicago Press.

Carroll, J. B. (1985). Exploratory factor analysis: A tutorial.In D. K. Detterman (Ed.), Current topics in humanintelligence: Vol. 1, Research methodology (pp. 25-58).Norwood, NJ: Ablex.

Clance, P. R. (1985). The impostor phenomenon. NewWoman, 15(7), 40-43.

Cohn,S. J. (1991). Talent searches. In N. Colangelo & G. A.Davis (Eds.), Handbookofgifted education (pp. 166-177).Boston: Allyn & Bacon.

Dawis, R. V. (1991). Vocational interests, values and prefer-ences. In M. Dunnette & L. Hough (Eds.), Handbook ofindustrial and organizational psychology, 2nd edn, Vol. 2(pp. 833-871). Palo Alto: Consulting Psychologist Press.

Dawis, R. V., & Lofquist, L. H. (1984). A psychological theoryof work adjustment: An individual differences model anditsapplications. Minneapolis: University of Minnesota Press.

Dick, T. P., & Rallis, S. F. (1991). Factors and influences onhigh school student career choices. Journal for Research inMathematics Education, 22, 281-292.

Eccles (Parsons), J., Adler, T. F., Futterman, R., Goff, S.

B., Kaczala, C. M., Meece, J. L., & Midgley, C. (1983).Expectations, values and academic behaviors. In J. T.

Spence (Ed.), Perspective on achievement and achievementmotivation. San Francisco: W. H. Freeman.

Eccles, J., & Harold, R. D. (1992). Gender differences in

educational and occupational patterns amongthegifted.In N. Colangelo, S. G. Assouline, & D. L. Ambroson(Eds.), Talent development: Proceedings from the 1991 H.B. & J. Wallace national research symposium on talent

development (pp. 2-30). Unionville, NY: Trillium Press.Ethington, C. A., & Wolfe, L. M. (1986). A structural model of

mathematics achievement for men and women. AmericanEducational Research Journal, 23, 65-75.

Feingold, A. (1988). Cognitive gender differences are disap-pearing. American Psychologist, 43, 95-103.

Feingold, A. (1992). Sex differencesin variability in intellectualabilities: A new look at an old controversy. Review ofEducational Research, 62, 61-84.

Flanagan, J. C., Dailey, J. T., Shaycoft, M. F., Gorham, W.

A., Orr, D. B., & Goldberg, I. (1962). Design for a studyfor American youth. Boston: Houghton Mifflin.

Fox, L. H., Pasternak, S. R., & Peiser, N. L. (1976).Career-related interests of adolescent boys andgirls. InD. P. Keating (Ed.), Intellectual talent: Research anddevelopment (pp. 242-261). Baltimore: Johns HopkinsUniversity Press.

Grinder, R. E. (1985). The gifted in our midst: By their divinedeeds, neuroses and mental test scores we have knownthem. In F. D. Horowitz & M. O’Brien (Eds.), Thegifted and talented: Developmental perspectives (pp. 5—36).Washington, DC: American Psychological Association.

Guttman, L. (1954). A new approachto factor analysis: Theradex. In P. Lazarsfeld (Ed.), Mathematical thinking in the

social sciences (pp. 258-348). Glencoe, IL: Free Press.Holland, J. C. (1985). Making vocational choices: A theory of

vocational personalities and work environments (2nd edn).EnglewoodCliffs, NJ: Prentice-Hall.

Humphreys, L. G., Lubinski, D., & Yao, G. (1993). Utilityof predicting group membership: Exemplified by the roleof spatial visualization in becoming an engineer, physicalscientist, or artist. Journal of Applied Psychology, 78,250-261.


Hyde, J. S., Fennema, E., & Lamon,S.J. (1990). Gender

differences in mathematics performance: A meta-analysis.Psychological Bulletin, 107, 139-155.

Kerr, B. A. (1985). Smart girls, gifted women. Dayton, OH:Ohio F'sychology Publishing Co.

Kerr, B. A. (1991). Educating gifted girls. In N. Colangelo& G. Davis (Eds.), Handbook of gifted education (pp.402-415). Boston: Allyn & Bacon.

Kimball, M. M. (1989). A new perspective on women’s mathachiev:ment. Psychological Bulletin, 105, 198-214.

Loevinger, J. (1976). Ego development: Conceptions andtheorié.. San Francisco: Jossey-Bass.

Lofquist, L. H., & Dawis, R. V. (1969). Adjustment to work.

New York: Appleton-Century-Crofts.Lofquist, I.. H., & Dawis, R. V. (1991). Essentials of per-

son environment correspondence counseling. Minneapolis:Unive: sity of Minnesota Press.

Lubinski, I)., & Benbow, C. P. (1992). Gender differences in

abilities and preferences among the gifted: Implicationsfor the math/science pipeline. Current Directions in Psy-cholog:cal Science, 1, 61-66.

Lubinski, [)., & Dawis, R. V. (1992). Aptitudes, skills and

proficiency. In M. D. Dunnette & L. M. Hough (Eds.),The k.indbook of industrial/organizational psychology,2nd ei, Vol. 3 (pp. 3-59). Palo Alto, CA: ConsultingPsychclogists Press.

Lubinski, D., & Humphreys, L. G. (1990a). A broadlybased analysis of mathematical giftedness. Intelligence,14, 327-355.

Lubinski, I)., & Humphreys, L. G. (1990b). Assessing spurious“moderator effects”: Illustrated substantively with thehypothesized (“synergistic”) relation between spatial visu-alization and mathematical ability. Psychological Bulletin,107, 3:7-355.

Lubinski, ])., & Thompson, T. (1986). Functional units ofhuman behavior and their integration: A dispositionalanalysis. In T. Thompson & M. Zeiler (Eds.), Analysis andintegration of behavioral units (pp. 275-314). Hillsdale,NJ: Erlbaum.

Lytton, H., & Romney, D. M. (1991). Parents’ differentialsocialization of boys and girls: A meta-analysis. Psycho-logica! Bulletin, 109, 267-296.

MacKinnon, D. W. (1962). The nature and nurture of creative

talent. American Psychologist, 17, 484-495.

Machlowit:., M. (1982). The great impostors. Working Women,7, 97-"8.

Maple, S. .A., & Stage, F. K. (1991). Influences on the choiceof math/science majors by genderand ethnicity. AmericanEducational Research Journal, 28, 37-60.

Noble, K. | 1989). Counseling gifted women: Becoming heroesof our own stories. Journal for the Education of the Gifted,12, 13 -141.

Reis, S. M (1987). We can’t change what we don’t recognize:Unde:standing the special needs of gifted families. Gifted

Child Duarterly, 31, 83-89.

Reis, S., &: Callahan, C. (1989). Gifted females: They've come

a long way—or have they? Journalfor the Education of theGiftec', 12, 99-117.

Roe, A. (1953). The making of a scientist. New York: Dodd,Mead.

Rosenthal, R. and Rubin, D. B. (1982). Further meta-analyticproce:iures for assessing cognitive gender differences.

Journ: of Educational Psychology, 74, 708-712.Scarr, S., & McCartney, K. (1983). How people make their

own environments: A theory of genotype—environmenteffects. Child Development, 54, 424-435.

Schmidt, F. L., & Hunter, J. E. (1992). Developmentof causalmodel of processes determining job performance. CurrentDirections in Psychological Science, 1, 89-92.

Sells, L. W. (1980). The mathematicsfilter and the educationof women and minorities. In L. H. Fox, L. Brody, &D. Tobin (Eds.), Women and the mathematical mystique.Baltimore: Johns Hopkins University Press.

Silverman, L. K. (1986). What happensin the gifted girl? InC. J. Maker (Ed.), Critical issues in gifted education:Defensible programsfor the gifted (pp. 43-89). Rockville,MD:Aspen.

Snow, R. E., Kyllonen, P. C., & Marshalek, B. (1984). The

topography of ability and learning correlations. In R. J.Sternberg (Ed.), Advances in the psychology of humanintelligence: Vol. 2 (pp. 47-104). Hillsdale, NJ: LawrenceErlbaum Associates.

Southern, M. L., & Plant, W. T. (1968). Personality character-

istics of very bright adults. Journal of Social Psychology,75, 119-126.

Stains, G., Tavris, C., & Jayaratne, C. (1974). The queen bee

syndrome. Psychology Today, 7, 55-60.Stanley, J. C. (1973). Accelerating the educational progress of

intellectually gifted youths. Educational Psychologist, 10,133-146.

Stanley, J. C. (1990). Leta Hollingworth’s contributions toabove-level testing of the gifted. Roeper Review, 12,166-171.

Stanley, J. C. (1992). A slice of advice. Educational Researcher,21, 25-26.

Stanley, J. C., Benbow, C. P., Brody, L. E., Dauber, S., &

Lupkowski, A. E. (1992). Gender differences on eighty-sixnationally standardized aptitude and achievement tests.In N. Colangelo, S. G. Assouline, & D. L. Ambroson

(Eds.), Talent development (pp. 42-65). Unionville, NY:Trillium Press.

Stanley, J. C., Feny, J., & Zhu, X. (1989). Chinese youths whoreason extremely well mathematically: Threat or bonanza?Network Newsand Views, 8, 33-39.

Stanley, J. C., Keating, D. P., & Fox, L. H. (Eds.)(1974). Mathematical talent : Discovery, description anddevelopment. Baltimore: Johns Hopkins University Press.

Tannenbaum, A. (1983). Gifted children: Psychological and

educational perspectives. New York: Macmillan.Tannenbaum, A. (1986). The enrichment matrix model. In

J. S. Renzulli (Ed.), Systems and models for developingprograms for the gifted and talented (pp. 391-428).Mansfield Center, CT: Creative Learning Press.

Thorndike, E. L. (1911). Individuality. Cambridge, MA:Riverside Press.

Warschaw,T. (1985). The “I-don’t-deserve-it” syndrome. New

Woman, 15, 134-137.

Zuckerman, H. (1977). Scientific elite: Nobel laureates in theUnited States. New York: Free Press.

Suggested Further Reading

Clark, B. (1983). Growing up gifted: Developing the potentialof children at home and at school. Columbus: Charles E.Merrill Publishing Company.

Swiatek, M. A. (1992). Academic and psychosocial perspectives

on giftedness during adolescence. Unpublished doctoraldissertation, Iowa State University, Ames, IA 50011.

707

reconceptualizinggenderdifferencesinachievement amongthe … · 2019-08-26 · reconceptualizing...

Documents