You are on page 1of 11

C O G N I T I V E C H A R AC T E R I S T I C S O F T H E G I F T E D I N M AT H

Cognitive and Motivational Characteristics


of Adolescents Gifted in Mathematics:
Comparisons Among Students
With Different Types of Giftedness
Eunsook Hong Yvette Aqui
University of Nevada, Las Vegas

the field of gifted education. With the transition, gifted


ABSTRACT education has shifted its emphasis from identifying
youngsters with high intellectual ability to discovering a
Although numerous studies have compared cogni- wide variety of specific talents in young people in its
tive and motivational characteristics between gifted efforts to provide special educational opportunities to
versus nongifted students, research comparing those the gifted.
characteristics among different types of gifted stu-
dents has not kept up with the theoretical develop-
ment that saw a transition from unidimensional to PUTTING THE RESEARCH
multidimensional conceptualizations of giftedness. TO USE
This study compares cognitive and motivational
characteristics of high school students who are aca- Adolescents academically gifted, creatively talented,
demically gifted in math, creatively talented in math, and nongifted in the mathematics area were different
and nongifted. Whereas no differences were found in many of the cognitive and motivational character-
among the three groups in their beliefs about ability, istics examined in this study. The findings point to
most of the other characteristics examined in the the need for differentiation among types of gifted-
study distinguished the three groups. Academically ness when conducting research and in education.
gifted female students reported expending more Research on giftedness should indicate the type of
effort than did academically gifted male students. giftedness that will be examined, ref lecting the mul-
Creatively talented males put forth more effort than tidimensional conceptualization of giftedness. For
academically gifted males, and the creatively talented example, intellectually gifted, academically gifted, or
in general used more cognitive strategies than the creatively talented, as well as the domain of interest,
academically gifted. Overall, students who were should be clearly defined in the study. Educators
either academically gifted or creatively talented in should recognize various types of giftedness when
mathematics perceived that they were self-effica- identifying students for gifted programs and use
cious in general, used cognitive strategies, perceived knowledge of differences in cognitive and motiva-
their math ability and math self-efficacy to be high, tional characteristics among these students when
and valued learning math more so than their assisting them. Some empirical findings defy com-
nongifted age peers. mon assumptions. For example, beliefs about ability
were not different among students in the three
The transition from unidimensional (Terman, groups examined; gender differences were manifest-
1925; Terman & Oden, 1947, 1959) to multidimension- ed in a different manner in different categories of
al conceptualizations of giftedness (e.g., Gardner, 1983, giftedness; and some cognitive characteristics were
1999; Guilford, 1956, 1967; Marland, 1972; Milgram, evident in only one gifted category. Thus, under-
1989, 1991; Renzulli, 1986; Sternberg & Davidson, standing gifted students can be problematic under
1986; Torrance, 1984) has had a significant impact on assumptions not examined empirically.

G I F T E D C H I L D Q U A RT E R LY • S U M M E R 2 0 0 4 • VO L 4 8 NO 3 191
Downloaded from gcq.sagepub.com at MCMASTER UNIV LIBRARY on February 20, 2015
C O G N I T I V E C H A R AC T E R I S T I C S O F T H E G I F T E D I N M AT H

In an effort to provide empirical evidence for a mul- achieving in school math. One might expect that gifted
tidimensional conceptualization of giftedness, Hong and individuals in these two categories might share some gen-
Milgram (1996) distinguished four different types of gift- eral cognitive characteristics, considering that mathemat-
edness: general intellectual ability, academic ability, gen- ics is highly intellectual in nature and requires certain
eral creativity, and specific creative talent. A sample of types of cognitive processes whether individuals engage
students in seventh through ninth grade with a wide in math activities for school-related or out-of-school self-
range of abilities and talents was used in their study. The selected purposes. On the other hand, there may be some
underlying rationale for the selection of such a sample differences in their cognitive and motivational character-
was that a student might be gifted in one category, but istics due to different types and levels of interest, curiosi-
not gifted in the others, thus requiring a sample with a ty, and task commitment required to be excellent in the
wider range of abilities not limited to high intellectual two different categories of giftedness.
ability as a baseline requirement for sample selection. Cognitive and motivational characteristics have been
Among the models tested in their study, the proposed studied in general student populations. Hong and O’Neil
four-factor giftedness model best fit the data. Their study (2001) conducted a construct validation study examining
examined the domain of literature for the academic abil- self-regulation in academic performance. Self-regulation
ity and the creative talent categories. Their findings was composed of metacognition and motivation. The
demonstrated that creative talent in literature measured metacognition component consisted of planning and self-
by out-of-school accomplishment in literature had a low checking. Students who used metacognitive strategies
correlation with general intelligence and a low to moder- tended to plan strategies and checked their solution
ate correlation with academic ability in literature. processes and answers while solving problems.
Other studies that have examined the relationship Motivation consisted of self-efficacy and effort. Students
among different types of giftedness suggest similar con- who were intrinsically motivated tended to be self-effica-
clusions. Yager (1989), for example, contended that high cious and also expended their efforts while they were
achievement in scientific subjects, overall intellectual engaged in tasks. Their findings supported previous stud-
ability, and excellence in science are not necessarily relat- ies indicating strong relationships among motivation,
ed to one another. Milgram and Milgram (1976) argued self-efficacy, and strategy use (e.g., Shih & Alexander,
that creative talent measured by creative accomplishment 2000).
was not related to academic achievement measured by Zimmerman (1990), in his social-cognitive theory,
school grades. Wallach and Wing (1969) implied from asserted that a student’s use of cognitive and metacogni-
their findings that students who will succeed at creative tive strategies enhances perceptions of self-efficacy,
writing, science, painting, or some other endeavor can- which in turn are assumed to provide the motivational
not be predicted by grades or test scores alone and that basis for further self-regulation during learning. Self-effi-
nonacademic accomplishments should be examined and cacy beliefs contribute to motivation by determining the
credited when searching for talented students. goals people set for themselves, how much effort they
Although the previous research by Hong and expend, and how long they persevere (Bandura, 1986,
Milgram (1996) provided the conceptual clarification of 1993; Schunk, 1987).
diverse abilities and talents in the literature domain, The causal attributions that individuals make about
whether the distinction of the constructs can be found in the success or failure of their actions (Weiner, 1994)
other domains needs to be examined. In addition, the inf luence motivation and performance largely through
previous study did not examine cognitive and motiva- the mediational role of self-efficacy (Bandura, 1995;
tional characteristics of the gifted. To understand what Schunk, 1991, 1994). In general, students who are confi-
distinguishes different types of giftedness, various charac- dent of their ability intensify their efforts when failure
teristics of the gifted need to be studied. occurs and persist until they succeed (Pajares, 1996;
This study examined cognitive and motivational Pajares & Kranzler, 1995). Highly perceived ability,
characteristics of adolescents who were academically gift- learning goals, intrinsic value, and attainment together
ed or creatively talented in the mathematics domain. The are likely to support optimism and persistence during
academically gifted in this study were students who were learning (Harrison, Rainer, Hochwarter, & Thompson,
high achieving in school math; by contrast, the creatively 1997).
talented were those who were highly interested, active, Of the various theories of intelligence, entity (fixed
and/or accomplished in math, but not necessarily high ability) versus incremental (f lexible ability) theories

192 G I F T E D C H I L D Q U A RT E R LY • S U M M E R 2 0 0 4 • VO L 4 8 NO 3
Downloaded from gcq.sagepub.com at MCMASTER UNIV LIBRARY on February 20, 2015
C O G N I T I V E C H A R AC T E R I S T I C S O F T H E G I F T E D I N M AT H

(Dweck, 2000) have been discussed in regard to their Although these studies have compared cognitive or
relationships with educational constructs such as attribu- motivational characteristics between gifted versus
tion theory (e.g., Hong, Chiu, Dweck, Lin, & Wan, nongifted students, to our knowledge, research compar-
1999; Stipek & Gralinski, 1996). While both entity and ing those characteristics among different types of gifted
incremental theorists recognize the importance of ability students or research examining interaction effects
in task performance, incremental theorists (those stu- between gender and type of giftedness in various cogni-
dents holding a f lexible ability or incremental theory of tive and motivational characteristics has not been con-
intelligence) also stress effort. Further, when faced with ducted.
unsatisfactory performance, incremental theorists tend to Therefore, the current study compares cognitive and
put forth more efforts to take remedial action to improve. motivational characteristics among the three groups of
On the other hand, entity theorists (those students hold- students (i.e., the gifted in two different categories and
ing a fixed ability or entity theory of intelligence) are their nongifted age peers) across gender. We investigated
more concerned with demonstrating their ability or not two major areas of cognitive and motivational characteris-
revealing their lack of it, and they tend to attribute their tics: One measured student perceptions of their character-
failure to lack of ability, rather than effort (Hong et al.). istics in general task-oriented situations and the other in
Studies with gifted students have indicated that these stu- the mathematics domain. Specifically, the study examined
dents tend to hold an incremental view, rather than an interaction and main effects of the type of giftedness and
entity view of ability (Feldhusen & Dai, 1997; Hsueh, gender in (a) their theories about ability and their percep-
1997). tions about their own effort expenditure, self-efficacy,
Cognitive and motivational characteristics of the and cognitive strategy use in general task-oriented situa-
gifted have been examined by comparing them with tions; and (b) their perceived math ability, effort expendi-
those of their nongifted peers. Gifted children are cogni- ture in math, value of math, and math self-efficacy.
tively more competent and are more intrinsically moti-
vated than their nongifted peers (Fehrenbach, 1991;
Gottfried & Gottfried, 1996; Vallerand, Gagné, Senécal, Method
& Pelletier, 1994; Zimmerman & Martinez-Pons, 1990).
Gifted students tend to be more strategic (Montague & Participants
Applegate, 1993; Shore & Dover, 1987); are more likely
to have conscious control over the solution process Participants were 90 tenth or eleventh graders (42
(Shore & Carey, 1984); use more strategies for organizing males, 47 females, 1 unspecified) who were taking a math
and transforming information (Zimmerman & Martinez- course (Algebra II) in three high schools located in a large
Pons, 1990); use learning strategies more effectively and school district in the southwestern United States.
transfer these strategies to novel tasks when trained Math teachers who taught Algebra II were contacted
(Risemberg & Zimmerman, 1992); and use more reread- through the chair of the math department to determine
ing, inferring, analyzing structure, predicting, and evalu- their interest in assisting data collection with students
ating strategies (Fehrenbach, 1991) when compared to taking their course. Seven math teachers from three
their nongifted counterparts. schools volunteered to help with the project. Parental
Tallent-Runnels et al. (1994) found that junior high consent and student assent forms were collected from
female students scored higher than male students in students who were willing to participate in the study.
motivation. Female gifted students also reported greater There were 216 volunteers for the study. However, only
confidence in English (Li & Adamson, 1995). However, 90 students met the grouping criteria for the current
Junge and Dretzke (1995) found that, in mathematics, investigation (see Grouping Procedure below).
male high school students had stronger self-efficacy
expectations than females on more than one fourth of the Measures
items on the mathematical self-efficacy scale they used,
whereas females reported stronger self-efficacy expecta- Self-Assessment Questionnaire (SAQ). The SAQ
tions on only a few items that involved stereotypical (Hong, 2001) measured participants’ perceptions about
female activities. On the other hand, Vallerand and his domain-general and domain-specific cognitive and moti-
associates (1994) found no gender effects in motivation vational constructs. Mathematics was the domain for the
and perceived competence in elementary students. latter. Some of the SAQ items were created by the first

G I F T E D C H I L D Q U A RT E R LY • S U M M E R 2 0 0 4 • VO L 4 8 NO 3 193
Downloaded from gcq.sagepub.com at MCMASTER UNIV LIBRARY on February 20, 2015
C O G N I T I V E C H A R AC T E R I S T I C S O F T H E G I F T E D I N M AT H

author, and others were modified from the self-regulato- were measured using eight math-related items from the
ry inventory developed and validated by O’Neil and his 80-item Activities and Accomplishments Inventory
colleagues (O’Neil, Baker, Ni, Jacoby, & Swigger, 1994; (AAI; Milgram & Hong, 2001). The AAI taps specific
O’Neil & Herl, 1998), from Generalized Self-Efficacy by challenging nonacademic activities and accomplishments
Schwarzer (1993) and Wegner, Schwarzer, and Jerusalem in 10 specific activity domains. The 10 domains are sci-
(1993) and from the self-efficacy portion of the ence, computers, mathematics, literature, social activi-
Motivated Strategies for Learning Questionnaire ties, drama, music, art, dance, and sports, with each
(MSLQ) by Pintrich and his associates (Pintrich & De domain consisting of eight items. Two of the eight items
Groot, 1990; Pintrich, Smith, Garcia, & McKeachie, were similar across domains: “Have you ever taken part in
1993). a [e.g., mathematics/drama] competition/contest?” and
The domain-general SAQ items measured partici- “Have you ever won a prize in a [e.g., mathematics/
pants’ (a) views about ability (5 items; e.g., “Intelligence drama] competition/contest?”
cannot be changed,” “If you are smart, you don’t have to This instrument was developed on the basis of earli-
study very hard to get good grades”), (b) perceptions er work by Holland and his associates (Holland, 1961;
about their own effort expenditure (7 items; e.g., “I work Holland & Austin, 1962; Holland & Nichols, 1964;
as hard as possible on all tasks,” “I don’t give up even if Holland & Richards, 1965) and Wallach and Wing
the task is hard”), (c) general self-efficacy (6 items; e.g., (1969). Hong, Whiston, and Milgram (1993) reported
“When I am confronted with a problem, I usually find evidence of moderate to good construct validity for a
several solutions,” “I always manage to solve difficult longer version of the inventory. Predictive validity of the
problems if I try hard enough”), and (d) cognitive strate- instrument scores was also investigated in a longitudinal
gy use (8 items; e.g., “I use multiple thinking techniques study. A significant relationship was found between the
or strategies to solve a problem,” “I select and organize domain of out-of-school activity in adolescent and the
relevant information to solve a problem”). The internal domain of vocational activity in adulthood. In addition,
consistency estimates (coefficient alpha) of the domain- the adult participants whose adolescent activities matched
general SAQ subscales in the current study were .62 for adult occupation had a higher level of work accomplish-
ability, .87 for effort, .76 for self-efficacy, and .85 for cog- ment than those for whom a match was absent.
nitive strategy. The attainments and activities are designated as
The domain-specific SAQ items measured percep- nonacademic because they are freely chosen and not done
tions specifically related to mathematics in four areas. to fulfill school requirements or to earn grades or credits.
They included participants’ (a) perceptions of their own The instructions of the AAI questionnaire read: “This
math ability (5 items; e.g., “I have a mathematical mind,” questionnaire is about the activities that students do in
“I can solve any math problems if I put my mind to it”), their free time. Please read the questions and mark your
(b) perceived effort exerted in math tasks (5 items, e.g., “I answers on the questionnaire. If you do an activity, circle
work as hard as possible on math tasks,” “I don’t give up the answer YES next to the question. If you don’t, circle
even if the math problem is hard”), (c) perceived value of the answer NO. There are no right or wrong answers to
math (6 items; e.g., “It is important for me to learn the these questions.” Examples of the items in the mathe-
course material in this math class,” “The content taught matics domain include: “Do you like games that are
in this math course is valuable for me”), and (d) perceived based upon simple mathematical activities such as card
math self-efficacy (6 items; e.g., “I believe I will receive games, puzzles, or mathematical computer games?”;
an excellent grade in this math class,” “I expect to do well “Have you ever devised new and original mathematical
in this math class”). The internal consistency estimates rules that were based on existing rules?”; and “Have you
(coefficient alpha) of the domain-specific SAQ subscales ever taken part in a mathematics competition?” A student
in the current study were .88 for math ability, .91 for with a high score in the math AAI was considered cre-
math effort, .93 for math value, and .93 for math self- atively talented in mathematics in this study (see
efficacy. Participants responded to each SAQ item by rat- Grouping Procedure).
ing themselves on the following four-point scale: 1 = not Math achievement. Math achievement scores from a
at all, 2 = somewhat, 3 = moderately so, and 4 = very math midterm exam were collected from the participat-
much so. ing teachers. Because the participating schools used dif-
Activities and Accomplishment Inventory. Participants’ ferent tests, the reported math scores were standardized
interest, activities, and accomplishment in mathematics before subsequent data analyses and the grouping proce-

194 G I F T E D C H I L D Q U A RT E R LY • S U M M E R 2 0 0 4 • VO L 4 8 NO 3
Downloaded from gcq.sagepub.com at MCMASTER UNIV LIBRARY on February 20, 2015
C O G N I T I V E C H A R AC T E R I S T I C S O F T H E G I F T E D I N M AT H

dure took place. The transformed math achievement level of an independent variable. When main effects were
scores for the 90 students involved in the current study significant, necessary post hoc multiple comparisons
ranged from 17.64 to 71.43, with a mean and standard were performed using the Tukey method.
deviation of 48.19 and 12.53, respectively. The sample size of 90 was reduced to 89 due to 1 stu-
dent in the second group having no gender specification.
Grouping Procedure Due to the small sample sizes that resulted from selecting
participants representing three types of giftedness,
Three groups were created to represent adolescents assumptions for MANOVA procedures were carefully
who are (a) academically gifted, but not creatively talent- examined to determine if they were reasonable analytical
ed in mathematics (Group 1); (b) creatively talented, but procedures for the current data. No within-cell outliers
not academically gifted in mathematics (Group 2); and (c) were found at p < .001. Skewness measured for each vari-
neither academically gifted nor creatively talented in able for each cell was not extreme, mostly smaller than
mathematics (Group 3). For the purpose of the study, |1| with the largest value of 1.86. Scatterplots for each
participants with a math achievement score higher than pair of dependent measures within each group revealed
59 (78th percentile of the entire sample) and with 3 (76th no cause for concern about linearity.
percentile) or less “yes” responses to AAI were assigned Box’s M test of equality of covariance matrices indi-
to the first group (n = 25). Participants with 4 (89th per- cated that the assumption of homogeneity of variance-
centile) or more “yes” responses to AAI items and with a covariance matrices was met for the domain-general
math achievement score lower than 59 were assigned to measures, F(50, 10543) = 1.35, p > .05. Univariate tests of
the second group (n = 21). Participants with a math homogeneity of variance also confirmed equality of vari-
achievement score lower than 42 (22nd percentile) and ances for individual dependent variables except for effort
an AAI score of 2 (59th percentile) or less were placed in p < .05. For the domain-specific measures, Box’s M test
the third group (n = 23). The grouping procedure was significant, although only math self-efficacy indicat-
excludes students in the middle range of mathematics ed nonequality of variance, p < .05. Thus, Pillai’s criteri-
scores to accomplish group distinction (high and low lev- on was used to evaluate multivariate significance (Mertler
els of academic mathematics) with a reasonable sample & Vannatta, 2001; Olson, 1979). The results of overall
size in each group. Due to high difficulty levels of math evaluations of assumptions were considered satisfactory
activities and accomplishment items included in the AAI to continue with the planned statistical procedures.
that was used to measure creative talent in mathematics,
only 3 participants scored 6 or above (out of 8). Thus,
activity scores between 2 and 4 were used as cutoff scores Results
for this category.
Effects on Domain-General Cognitive
Data Analysis and Motivational Variables: Theories of Ability,
Effort, Self-Efficacy, and Cognitive Strategy
Two multivariate analyses of variance (MANOVAs) Use in General Task-Oriented Situations
were performed using the four domain-general and the
four domain-specific SAQ measures as dependent vari- The combined domain-general measures were sig-
ables and gender and group (three types of giftedness; nificantly affected by a disordinal interaction between
henceforth, group) as independent variables. To deter- gender and group, F(8, 162) = 3.63, p = .001, with partial
mine which of the four domain-general and four η2 = .15 ref lecting a weak association between the com-
domain-specific measures contributed to group differ- bined measures and the interaction of the two independ-
ences, follow-up univariate analyses of variance ent variables. The gender and group main effects were
(ANOVA) were conducted. Bonferroni corrections for also statistically significant in the multivariate analyses, p
probability (α = .0125) were applied due to assessing a < .01 and p < .0005, respectively, indicating that some of
large number of group comparisons (four for domain- the individual measures might have been affected by the
general and four for domain-specific measures). When group or gender difference.
interaction effects were found significant, further analy- Follow-up univariate analyses indicated that the gen-
ses were conducted to compute simple effects and simple der-group interaction effect was found only for the effort
contrasts for comparing group differences within each variable, F(2, 83) = 11.69, p < .0005, partial η2 = .35.

G I F T E D C H I L D Q U A RT E R LY • S U M M E R 2 0 0 4 • VO L 4 8 NO 3 195
Downloaded from gcq.sagepub.com at MCMASTER UNIV LIBRARY on February 20, 2015
C O G N I T I V E C H A R AC T E R I S T I C S O F T H E G I F T E D I N M AT H

Table 1
Means, Standard Deviations of the Domain-General Effort Scores
in the Academically Gifted, Creatively Talented, and Nongifted Group by Gender

Academically Gifted Creatively Talented Nongifted

n M SD n M SD n M SD

Male 12 2.23 .43 15 2.92 .51 15 2.28 .33


Female 19 3.15 .57 12 3.02 .31 16 2.19 .26

Table 2
Means, Standard Deviations of the Four Domain-General Cognitive Measures
in the Academically Gifted, Creatively Talented, and Nongifted Group

Academically Gifted Creatively Talented Nongifted

M SD M SD M SD F(2, 63) Partial η2

Ability 2.26 .67 2.28 .49 2.21 .56 .75 .01


Efforta 2.79 .69 2.97 .43 2.23 .30 12.13** .35
Self-efficacy 2.81 .54 3.01 .47 2.51 .39 8.71** .17
Cognitive strategy 2.37 .61 2.74 .52 2.26 .35 8.31 .17

Note. * p < .01. ** p < .0005.


aInterpretation of the effort main effect not meaningful due to the significant interaction effect between gender and group.

Thus, simple effects for gender and those for group were nificant, p > .80. Male students in Group 2 had a higher
computed for the effort variable. A statistically significant mean effort compared to Group 1 and to Group 3 (see
gender difference was found only in Group 1 (academi- Table 1 for means and standard deviations for the three
cally gifted), F(1, 83) = 38.00, p < .0005, with η2 = .31, groups across gender).
indicating a moderate association between gender and For female students, effort scores were significantly
effort scores. The mean effort score of female students (M different among all three groups, ps < .05. Female stu-
= 3.15, SD = .57) was higher than that of their male coun- dents in Group 1 had a higher mean score compared to
terparts (M = 2.23, SD = .43). Gender differences were Group 2, F(1, 83) = 4.60, p < .05, with η2 = .02, indicat-
not found in the other two groups, ps > .40. ing that the group difference was statistically significant,
Tests of simple effects for the group variable (type of but not substantially meaningful with only 2% of vari-
giftedness) indicated that, for male students, statistically ance of effort accounted for by group membership. The
significant group differences were shown among the difference between Group 1 and Group 3 was also signif-
three group means, F(2, 83) = 11.21, p < .0005. A similar icant, F(1, 83) = 44.68, p < .0005, η2 = .36, as well as the
result was found for female students, F(2, 83) = 24.51, p < difference between Group 2 and Group 3, F(1, 83) =
.0005. Simple contrasts were performed for pairwise 27.49, p < .0005, η2 = .22 (see Table 1).
comparisons within males and females separately. For We next examined main effects on univariate
male students, the mean effort scores were significantly domain-general variables that were not inf luenced by the
different between Group 1 and Group 2, F(1, 83) = 17.14, group-by-gender interaction. Statistically significant dif-
p < .0005, η2 = .16, and between Group 2 and Group 3, ferences were found only on self-efficacy and cognitive
F(1, 83) = 17.11, p < .0005, η2 = .17. The difference strategy use among the three types of giftedness, but not
between Group 1 and Group 3 was not statistically sig- for gender. The mean self-efficacy scores were signifi-

196 G I F T E D C H I L D Q U A RT E R LY • S U M M E R 2 0 0 4 • VO L 4 8 NO 3
Downloaded from gcq.sagepub.com at MCMASTER UNIV LIBRARY on February 20, 2015
C O G N I T I V E C H A R AC T E R I S T I C S O F T H E G I F T E D I N M AT H

Table 3
Means, Standard Deviations of the Mathematics Effort Scores
in the Academically Gifted, Creatively Talented, and Nongifted Group by Gender

Academically Gifted Creatively Talented Nongifted

n M SD n M SD n M SD

Male 12 2.20 .64 15 2.87 .75 15 2.09 .45


Female 19 2.97 .60 12 2.70 .54 16 1.70 .45

cantly different among the three groups, F(2, 83) = 8.71, 1 (academically gifted), F(1, 83) = 16.33, p < .0005, η2 =
p < .0005, partial η2 = .17. The results of the post hoc .16. Female students (M = 2.97, SD = .60) scored a high-
analyses indicated that mean differences between Group er mean effort than their male counterparts (M = 2.20,
1 and Group 3, p < .05, and Group 2 and Group 3, p < SD = .64). Gender differences were not found in the
.0005, were statistically significant, with the creatively other two groups, ps ≥ .05.
talented in math (Group 2) reporting the highest mean Tests of simple effects for the group variable (type of
self-efficacy and Group 3 the lowest. The mean self-effi- giftedness) indicated that statistically significant group
cacy difference between Group 1 and Group 2 was not differences were found for both males, F(2, 83) = 7.62, p
statistically significant, p > .20 (see Table 2 for means and = .001, and females, F(2, 83) = 22.02, p < .0005. The
standard deviations). results of simple contrasts showed that, for male students,
For cognitive strategy use, F(2, 83) = 8.31, p = .001, the mean effort scores were significantly different
partial η2 = .17, again, students creatively talented in math between Group 1 and Group 2, F(1, 83) = 9.54, p = .003,
(Group 2) reported using cognitive strategies the most η2 = .09, and between Group 2 and Group 3, F(1, 83) =
and Group 3 the least (see Table 2). The mean differences 13.38, p < .0005, η2 = .14. The difference between Group
between Group 1 and Group 2, p < .02, and between 1 and Group 3 was not statistically significant, p > .40.
Group 2 and Group 3 showed were statistically signifi- Male students in Group 2 had a higher mean effort score
cance, p = .001. compared to Group 1 and to Group 3 (see Table 3 for
means and standard deviations).
Effects on Domain-Specific Cognitive For female students, the mean effort scores were sig-
and Motivational Variables: Ability, Effort, nificantly different among all three groups, ps < .05.
Value, and Self-Efficacy in Mathematics Female students in Group 1 reported a higher mean score
compared to Group 2, F(1, 83) = 4.75, p < .05, with η2 =
A statistically significant disordinal gender-group .02, indicating that the group difference was statistically
interaction was found on the combined domain-specific significant, but not practically significant. The difference
cognitive measures, F(8, 162) = 3.04, p = .003, with par- between Group 1 and Group 3 was significant, F(1, 83) =
tial η2 = .13. The gender and group main effects were also 40.70, p < .0005, η2 = .33, as well as the difference
statistically significant in the multivariate analyses, p < .05 between Group 2 and Group 3, F(1, 83) = 23.66, p <
and p < .0005, respectively, indicating again that some of .0005, η2 = .19. The pattern of findings from the analysis
the domain-specific measures might have been affected of math effort and of general effort was the same (see
by group or gender differences. These multivariate find- Table 3).
ings were similar to those found with the domain-gener- Group differences were shown in math ability, math
al variables. value, and math self-efficacy in the follow-up univariate
Univariate analyses indicated that the gender-group analyses, all ps < .0005, with partial η2 ranging from .25 to
interaction effect was only in the effort variable, F(2, 83) .31. Post hoc analysis indicated that both Group 1 and
= 8.28, p = .001, partial η2 = .17. This result was also sim- Group 2 were significantly different from Group 3, all ps
ilar to that with the domain-general variables. Tests of < .0005. However, no statistically significant differences
simple effects for gender on math effort indicated that were found between Group 1 and Group 2 on the three
significant gender differences were found only in Group measures. Gender effects for the three variables were not

G I F T E D C H I L D Q U A RT E R LY • S U M M E R 2 0 0 4 • VO L 4 8 NO 3 197
Downloaded from gcq.sagepub.com at MCMASTER UNIV LIBRARY on February 20, 2015
C O G N I T I V E C H A R AC T E R I S T I C S O F T H E G I F T E D I N M AT H

Table 4
Means, Standard Deviations of the Four Domain-Specific (Mathematics) Measures
in the Academically Gifted, Creatively Talented, and Nongifted Group.

Academically Gifted Creatively Talented Nongifted

M SD M SD M SD F(2, 63) Partial η2

Math ability 2.87 .69 2.70 .79 1.91 .65 15.58* .27
Math efforta 2.67 .72 2.79 .66 1.89 .48 18.96* .31
Math value 2.82 .75 2.74 .75 1.87 .74 13.97* .25
Math self-efficacy 3.02 .79 2.70 .85 1.91 .58 17.29* .30

Note. * p < .0005


a
Interpretation of the effort main effect not meaningful due to the significant interaction effect between gender and group.

statistically significant (see Table 4 for means, standard compared to their nongifted peers. Similarly, gifted stu-
deviations, and F statistics for the three groups). dents reported using cognitive strategies more often than
the nongifted, especially with the creatively talented
reporting significantly higher uses than the nongifted.
Discussion Although the difference between the academically gifted
and the creatively talented was not significant for self-
Students’ beliefs regarding ability (intelligence) were efficacy, the difference between the two groups in the use
similar across gender and the three types of giftedness. of cognitive strategies was significant, with the creatively
That is, whether gifted or not, or male or female, stu- talented reporting a higher mean than the academically
dents’ theories about intelligence were the same in gen- gifted. This finding might indicate cognitive resourceful-
eral. With the mean scores ranging from 2.21 to 2.28, ness of creatively talented students, even though they do
students in the current sample were neither particularly not do as well in school math tests. Further studies are
entity theorists nor incremental theorists. Since incre- needed to determine the source of high-level cognitive
mental theorists tend to place more emphasis on effort strategies (e.g., profound interest) found in this group. In
(Stipek & Granlinski, 1996) and since effort has been addition, studies that require math problem solving
shown to have a positive relationship with achievement would help us closely observe cognitive strategies used by
(e.g., Chen, 1996), this finding was somewhat surprising. students in these groups.
However, both types of theorists also acknowledge the Research findings regarding gender effects on moti-
importance of intellectual ability in task performance vational variables conf lict widely. Whereas Tallent-
(Hong et al., 1999). In their studies of gifted students in Runnels et al. (1994) found females scoring higher in
gifted programs, Feldhusen and Dai (1997) and Hsueh motivation, it was male students in Junge and Dretzke
(1997) found that gifted students tend to hold an incre- (1995) who showed stronger self-efficacy than females.
mental view of their abilities. Neither study, however, Moreover, Vallerand et al. (1994) found no gender main
compared gifted to nongifted students regarding their effects in motivation and perceived competence in ele-
beliefs about abilities. The lack of group difference in mentary students. In the current study with secondary
beliefs about intelligence found in the current study students, gender effects also were not found except as part
might ref lect the beliefs of such adolescents as those who of an interaction with the type of giftedness.
participated in this study. Studies with various age levels Furthermore, gender and group interaction effects were
and various cultures would shed light on whether there is indicated only in effort (both domain-general effort and
any developmental change or cultural inf luence on belief math effort). Female high achievers in academic mathe-
formation. matics reported expending more effort in general tasks
Adolescents academically gifted or creatively talented (e.g., working as hard as possible on all tasks) and in
in math reported being more self-efficacious in general mathematics (e.g., putting forth best efforts when solving

198 G I F T E D C H I L D Q U A RT E R LY • S U M M E R 2 0 0 4 • VO L 4 8 NO 3
Downloaded from gcq.sagepub.com at MCMASTER UNIV LIBRARY on February 20, 2015
C O G N I T I V E C H A R AC T E R I S T I C S O F T H E G I F T E D I N M AT H

math problems) compared to male high achievers. Vallerand et al., 1994), and were more strategic or used
However, gender differences were not found within the more strategies (Montague & Applegate, 1993; Shore &
creatively talented or the nongifted. Dover, 1987; Zimmerman & Martinez-Pons, 1990).
The gender difference found among academically Overall, the new findings from this study add to findings
gifted students might point to the tendency of female of previous research, especially the differences found
adolescents to examine their work meticulously even between students in the two types of giftedness in math-
when they attain high levels of achievement in school ematics. Among the findings that have not been dis-
math. On the other hand, male high achievers might cussed in previous studies are the gender difference found
believe that they can do well in school math without only in the academically gifted, but not in the creatively
exerting much effort. Interestingly, however, both male talented regarding their effort expenditure; creatively tal-
and female creatively talented students reported high lev- ented males putting forth more effort than academically
els of effort in the general domain, as well as in mathe- gifted males, but finding no difference among females in
matics, possibly because out-of-school math activities these two groups; creatively talented students using more
require a higher level of commitment and sustained effort cognitive strategies than academically gifted or nongifted
to excel. Student interviews along with a questionnaire peers; and no gender or group differences in their beliefs
would help clarify this phenomenon in future studies. about intelligence.
Whereas female students in either gifted category However, the lack of differences in the three math-
reported expending more effort than the nongifted, cre- related measures (ability, value, self-efficacy) between the
atively talented males stood out in the amount of effort gifted and talented groups was somewhat surprising and,
expenditure as compared to their academically gifted and at the same time, somewhat expected due to the area (i.e.,
nongifted counterparts. That the pattern of the findings mathematics) being investigated. In the study of the liter-
was the same for both domain general tasks and mathe- ature domain by Hong and Milgram (1996), low to mod-
matics domain might indicate that gender differences in erate factor correlations were found between these two
effort are consistent across domains. Studies using other gifted and talented groups, with .29 and .37 in the two
domains are needed to further examine this conclusion. samples used in their study. The previous findings indi-
The conf licting research findings regarding gender cated that school grades and out-of-school accomplish-
effects on motivational variables in previous and current ment scores can, to some extent, discriminate between
studies (i.e., males scoring higher than females, females the individuals who are characterized as highly gifted in
scoring higher than males, no gender differences, or dif- literature in school and creatively talented in out-of-
ferences in some components of motivational variables, school accomplishments in literature.
in various studies) suggest that more research studies are However, the lack of distinctions between the gifted
warranted to determine the pattern of gender difference and talented categories manifested in the mathematics
in various domains, as well as in various grade levels and domain might indicate that, in mathematics, the content
cultures. and quality of in-school and out-of-school activities may
In mathematics-related cognitive and motivational not be distinctly different. For example, although the
characteristics, the findings were consistent across the instructions printed in the Activities and
three measures (math ability, math value, and math self- Accomplishments Inventory explain that the question-
efficacy). Compared to nongifted students, gifted male naire is about the activities students do in their free time,
and female students, whether high achievers in academic the out-of-school mathematical activities may be largely
mathematics or creatively talented in mathematics, per- related to in-school math activities. Mathematics,
ceived themselves as highly able in math (e.g., “I have a whether practiced by a student as school math or freely
mathematical mind”), valued their math class (e.g., “It is chosen out-of-school math, might require similar cogni-
important for me to learn the course material in this math tive processes more so than other domains such as litera-
class”), and were confident in mathematics (e.g., “I ture or leadership. However, the findings might also
expect to do well in this math class”). indicate the need to enhance the empirical distinction
Previous studies demonstrated that the gifted, when between the gifted and talented categories by developing
compared to their nongifted counterparts, perceived better assessment approaches, with better definitions of
themselves to be cognitively more competent, attributed creative mathematical ability and performance that
their success and failure to effort (e.g., Chen, 1996; ref lect cognitive and metacognitive conceptions of gift-
Fehrenbach, 1991; Gottfried & Gottfried, 1996; edness. In addition to the type of assessment used in the

G I F T E D C H I L D Q U A RT E R LY • S U M M E R 2 0 0 4 • VO L 4 8 NO 3 199
Downloaded from gcq.sagepub.com at MCMASTER UNIV LIBRARY on February 20, 2015
C O G N I T I V E C H A R AC T E R I S T I C S O F T H E G I F T E D I N M AT H

current study (i.e., assessment of activities and accom- Holland, J. L., & Nicholas, R. C. (1964). Prediction of aca-
plishment), other assessments, such as analysis of cogni- demic and extracurricular achievement in college. Journal of
tive processing, would increase the understanding of the Educational Psychology, 55, 55–65.
distinction between the academically gifted and the cre- Holland, J. L., & Richards, J. M., Jr. (1965). Academic and
nonacademic accomplishment: Correlated or uncorrelat-
atively talented in mathematics. In addition, whether
ed? Journal of Educational Psychology, 56, 165–174.
similar results would emerge with profoundly gifted sam-
Hong, E. (2001). Self-assessment questionnaire. Unpublished doc-
ples in the two gifted categories would be an important
ument, College of Education, University of Nevada, Las
topic for a future study. Vegas.
Hong, E., & Milgram, R. M. (1996). The structure of gifted-
ness: The domain of literature as an exemplar. Gifted Child
References Quarterly, 40, 31–40.
Hong, E., Milgram, R. M., & Whiston, S. C. (1993). Leisure
Bandura, A. (1986). Social foundations of thought and action: A social activities in adolescents as a predictor of occupational
cognitive theory. Englewood Cliffs, NJ: Prentice-Hall. choice in young adults: A longitudinal study. Journal of
Bandura, A. (1993). Perceived self-efficacy in cognitive devel- Career Development, 19, 221–229.
opment and functioning. Educational Psychologist, 28, Hong, E., & O’Neil, H. F., Jr. (2001). Construct validation of
117–148. a trait self-regulation model. International Journal of
Bandura, A. (Ed.). (1995). Self-efficacy in changing societies. New Psychology, 36, 186–194.
York: Cambridge University Press. Hong, E., Whiston, S. C., & Milgram, R. M. (1993). Leisure
Chen, L. K. S. (1996). Motivational orientations and metacog- activities in career guidance for gifted and talented adoles-
nitive abilities of intellectually gifted students. Gifted Child cents: A validation study of the Tel-Aviv Activities
Quarterly, 40, 184–194. Inventory. Gifted Child Quarterly, 37, 65–68.
Dweck, C. S. (2000). Self-theories: Their role in motivation, person- Hong, Y., Chiu, C., Dweck, C. S., Lin, D. M., & Wan, W.
ality, and development. Florence, KY: Psychology Press. (1999). Implicit theories, attributions, and coping: A
Fehrenbach, C. R. (1991). Gifted/average readers: Do they use meaning system approach. Journal of Personality and Social
the same reading strategies? Gifted Child Quarterly, 35, Psychology, 77, 588–599.
125–127. Hsueh, W. (1997). A cross-cultural comparison of gifted children’s the-
Feldhusen, J. F., & Dai, D. Y. (1997). Gifted students’ attitudes ories of intelligence, goal orientations, and response to challenge.
and perceptions of the gifted label, special programs, and Unpublished doctoral dissertation, Purdue University,
peer relations. Journal of Secondary Gifted Education, 9, 15–20. West Lafayette, IN.
Gardner, H. (1983). Frames of mind: The theory of multiple intelli- Junge, M. E., & Dretzke, B. J. (1995). Mathematical self-effica-
gences. New York: BasicBooks. cy gender differences in gifted/talented adolescents. Gifted
Gardner, H. (1999). Intelligence reframed: Multiple intelligences for Child Quarterly, 39, 22–28.
the 21st century. New York: BasicBooks Li, A. K. F., & Adamson, G. (1995). Motivational patterns
Gottfried, A. E., & Gottfried, A. W. (1996). A longitudinal related to gifted students’ learning of mathematics, science,
study of academic intrinsic motivation in intellectually and English: An examination of gender differences. Journal
gifted children: Childhood through early adolescence. for the Education of the Gifted, 18, 284–297.
Gifted Child Quarterly, 40, 179–183. Marland, S. P., Jr. (1972). Education of the gifted and talented:
Guilford, J. P. (1956). The structure of intellect. Psychological Report to the Congress of the United States by the U.S.
Bulletin, 53, 267–293. Commissioner of Education and background papers submitted to
Guilford, J. P. (1967). The nature of human intelligence. New York: the U.S. Office of Education, 2 vols. Washington, DC: U.S.
McGraw-Hill. Government Printing Office. (Government Documents,
Harrison, A. W., Rainer, R. K., Jr., Hochwarter, W. A., & Y4.L 11/2: G36)
Thompson, K. R. (1997). Testing the self-efficacy-per- Mertler, C., & Vannatta, R. (2001). Advanced and multivariate sta-
formance-linkage of social-cognitive theory. Journal of tistical methods: Practical application and interpretation. Los
Social Psychology, 137, 79–88. Angeles: Pyrczak.
Holland, J. L. (1961). Creative and academic performance Milgram, R. M. (Ed.). (1989). Teaching gifted and talented learners
among talented adolescents. Journal of Educational in regular classrooms. Springfield, IL: Thomas.
Psychology, 52, 136–147. Milgram, R. M. (1991). Counseling gifted and talented children: A
Holland, J. L., & Austin, A. W. (1962). The prediction of the guide for teachers, counselors, and parents. Norwood, NJ:
academic, artistic, scientific, and social achievement of Ablex.
undergraduates of superior scholastic aptitude. Journal of Milgram, R. M., & Hong, E. (2001). Activities and accomplishment
Educational Psychology, 53, 132–143. inventory. Unpublished document, School of Education, Tel

200 G I F T E D C H I L D Q U A RT E R LY • S U M M E R 2 0 0 4 • VO L 4 8 NO 3
Downloaded from gcq.sagepub.com at MCMASTER UNIV LIBRARY on February 20, 2015
C O G N I T I V E C H A R AC T E R I S T I C S O F T H E G I F T E D I N M AT H

Aviv University, Israel, and College of Education, Shih, S., & Alexander, J. M. (2000). Interacting effects of goal
University of Nevada, Las Vegas. setting and self- or other-referenced feedback on children’s
Milgram, R. M., & Milgram, N. A. (1976). Creative thinking development of self-efficacy and cognitive skill within the
and creative performance in Israeli children. Journal of Taiwanese classroom. Journal of Educational Psychology, 92,
Educational Psychology, 73, 568–572. 536–543.
Montague, M., & Applegate, B. (1993). Middle school stu- Shore, B. M., & Carey, S. M. (1984). Verbal ability and spatial
dents’ mathematical problem solving: An analysis of think- task. Perceptual & Motor Skills, 59, 255–259.
aloud protocols. Learning Disability Quarterly, 16, 19–32. Shore, B. M., & Dover, A. C. (1987). Metacognition, intelli-
Olson, C. L. (1979). Practical considerations in choosing a gence, and giftedness. Gifted Child Quarterly, 31, 37–39.
MANOVA test statistic: A rejoinder to Stevens. Sternberg, R. J., & Davidson, J. E. (Eds.). (1986). Conceptions of
Psychological Bulletin, 86, 1350–1352. giftedness. New York: Cambridge University Press.
O’Neil, H. F., Jr., Baker, E. L., Ni, Y., Jacoby, A., & Swigger, Stipek, D., & Gralinski, J. H. (1996). Children’s beliefs about
K. M. (1994). Human benchmarking for the evaluation of intelligence and school performance. Journal of Educational
expert systems. In H. F. O’Neil, Jr. & E. L. Baker (Eds.), Psychology, 88, 397–407.
Technology assessment in software applications (pp. 13–45). Tallent-Runnels, M. K., Olivárez, Jr., A., Candler-Lotven, A.
Hillsdale, NJ: Erlbaum. C., Walsh, S., Gray, A., & Irons, T. (1994). A comparison
O’Neil, H. F., Jr., & Herl, H. E. (1998, April). Reliability and of learning and study strategies of gifted and average-abili-
validity of a trait measure of self-regulation. Paper presented at ty junior high students. Journal for the Education of the Gifted,
the annual meeting of the American Educational Research 17, 143–160.
Association, San Diego, CA. Terman, L. M. (1925). Genetic studies of genius: Mental and physi-
Pajares, F. (1996). Self-efficacy beliefs in academic settings. cal traits of a thousand gifted children. Stanford, CA: Stanford
Review of Educational Research, 66, 543–578. University Press.
Pajares, F., & Kranzler, J. (1995). Self-efficacy beliefs and gen- Terman, L. M., & Oden, M. H. (1947). Genetic studies of genius:
eral mental ability in mathematical problem solving. Vol. 4. The gifted child grows up: Twenty-five years follow-up of
Contemporary Educational Psychology, 20, 427–444. a superior group. Stanford, CA: Stanford University Press.
Pintrich, P. R., & De Groot, E. V. (1990). Motivational and Terman, L. M., & Oden, M. H. (1959). Genetic studies of genius:
self-regulated learning components of classroom academic Vol. 4. The gifted child at mid-life: Thirty-five years follow-up of
performance. Journal of Educational Psychology, 82, 33–40. the superior child. Stanford, CA: Stanford University Press.
Pintrich, P. R., Smith, D. A., Garcia, T., & McKeachie, W. J. Torrance, E. P. (1984). The role of creativity in identification of
(1993). Reliability and predictability of the Motivated the gifted and talented. Gifted Child Quarterly, 28, 153–156.
Strategies for Learning Questionnaire (MSLQ). Educational Vallerand, Gagné, F., Senécal, C., & Pelletier, L. G. (1994). A
& Psychological Measurement, 53, 801–813. comparison of the school intrinsic motivation and per-
Renzulli, J. S. (1986). The three-ring conception of giftedness: ceived competence of gifted and regular students. Gifted
A developmental model for creative productivity. In R. J. Child Quarterly, 38, 172–175.
Sternberg & J. E. Davidson (Eds.), Conceptions of giftedness Wallach, M. A., & Wing, C. W., Jr. (1969). The talented student:
(pp. 53–92). New York: Cambridge University Press. A validation of the creativity-intelligence distinction. New York:
Risemberg, R., & Zimmerman, B. J. (1992). Self-regulated Holt, Rinehart, & Winston.
learning in gifted students. Roeper Review, 15, 98–101. Wegner, M., Schwarzer, R., & Jerusalem, M. (1993).
Schunk, D. H. (1987, September). Self-efficacy and cognitive Generalized self-efficacy. In Measurement of perceived self-effi-
achievement. Paper presented at the annual meeting of the cacy: Psychometric scales for cross-cultural research (R. Schwarzer,
American Psychological Association, New York. Trans.). Berlin: Forschung an der Freien Universitat
Schunk, D. H. (1991). Self-efficacy and academic motivation. Berlin.
Educational Psychologist, 26, 207–231. Weiner, B. (1994). Ability versus effort revisited: The moral
Schunk, D. H. (1994). Self-regulation of self-efficacy and attri- determinants of achievement evaluation and achievement
butions in academic settings. In D. H. Schunk & B. J. as a moral system. Educational Psychologist, 29, 163–172.
Zimmerman (Eds.), Self-regulation of learning and performance: Yager, R. E. (1989). Teaching science to gifted science students.
Issues and educational implications (pp. 75–99). Hillsdale, NJ: In R. M. Milgram (Ed.), Teaching gifted and talented learners
Erlbaum. in regular classrooms (pp. 223–248). Springfield, IL: Thomas.
Schwarzer, R. (1993). Measurement of perceived self-efficacy: Zimmerman, B. J., & Martinez-Pons, M. (1990). Student dif-
Psychometric scales for cross-cultural research. Berlin: Forschung ferences in self-regulated learning: Relating grade, sex, and
an der Freien Universitat Berlin. giftedness to self-efficacy and strategy use.

G I F T E D C H I L D Q U A RT E R LY • S U M M E R 2 0 0 4 • VO L 4 8 NO 3 201
Downloaded from gcq.sagepub.com at MCMASTER UNIV LIBRARY on February 20, 2015

You might also like