Heinrich Stumpf

A hierarchical structural analysis of perfectionism and its relation to other personality characteristics

This study examined components of perfectionism and their relation to other personality constructs, based on data on the Frost et al. [Frost, R. O., Heimberg, R. G., Holt, C. S., Mattia, J.S., & Neubauer, A. L. (1993). A comparison of two measures of perfectionism. Personality and Individual Differences, 14, 119–126; Frost, R. O., Marten, P., Lahart, C., & Rosenblate, R. (1990). The dimensions of perfectionism. Cognitive Therapy and Research, 15, 245–261] Multidimensional Perfectionism Scale (MPS), the Rosenberg Self-Esteem Scale, the Brief Symptom Inventory, the NEO Five-Factor Inventory and the Myers-Briggs Type Indicator in a population of academically talented students. The six scales of the MPS had been well recovered factor-analytically before. On a somewhat higher level of generality, four intercorrelated factors, Concerns and Doubts, Personal Standards, Organization and Parental Pressure were found to underlie self description on the MPS. This finding was consistent with previous observations. Two orthogonal higher-order factors were extracted from the intercorrelations of these four dimensions. These two factors could be clearly interpreted as healthy and unhealthy perfectionism; similar factors had been observed before in analyses of the scale scores on the MPS. This hierarchical structure of four first-order and two second-order factors could be well replicated in a second sample of subjects. The various factors on both levels showed distinct patterns of correlations with a large range of personality characteristics as measured by the other questionnaires mentioned above. In particular, healthy perfectionism was correlated with conscientiousness, unhealthy perfectionism with lack of self-esteem. Given this pattern of correlations, healthy and unhealthy perfectionism appear not to be opposite poles of one continuum, but distinct, independent factors, which are embedded in the overall personality in different ways.

An Examination of the Multidimensional Perfectionism Scale with a Sample of Academically Talented Children

The Multidimensional Perfectionism Scale (MPS; Frost, Marten, Lahart, & Rosenblate, 1990) was constructed and normed on an all-female college student sample. The present study is a psychometric examination of the instrument for use with a mixed-gender sample of 855 academically talented children who were participating in the longitudinal Developmental Study of Talented Youth conducted by the Center for Talented Youth of Johns Hopkins University. The internal consistency of the MPS total score was .87; coefficient alphas for the subscales ranged from .67 to .90. The matrix of intercorrelations of MPS subscales was consistent with that obtained with the college sample used to develop the instrument. A confirmatory item factor analysis demonstrated an acceptable degree of fit. A factor analysis of raw scores on the MPS subscales yielded two factors. The first factor appears to represent dysfunctional perfectionism, the second factor reflects healthy perfectionism. The scores on these factors were correlated with the five-factor domain scores obtained on the NEO-FFI. The first factor correlated most strongly with N (Neuroticism); the second factor correlated most strongly with C (Conscientiousness).

Gender differences in academically talented young students' mathematical reasoning: Patterns across age and subskills.

When a sample of academically talented students in Grades 2-6 was given a test of mathematical reasoning ability, boys performed better overall than girls. The gender differences for mathematical ability appeared as early as 2nd grade in samples tested over a 7-year period but varied somewhat according to mathematical subskills. There were no substantial gender-related differences on tasks requiring students to identify whether enough information was provided to solve a task; however, boys performed better than girls on tasks requiring application of algebraic rules or algorithms, as well as on tasks in which the understanding of mathematical concepts and number relationships was required

A validation of the five-factor model of personality in academically talented youth across observers and instruments

Abstract Four instruments—the NEO-Five-Factor Inventory (NEO-FFI), the Adjective Check List (ACL), the Myers–Briggs Type Indicator (MBTI) and the California Child Q-Set (CCQ)—and two data sources—self reports and parental ratings—were used to assess personality dimensions according to the five-factor model of personality in 870 academically talented youth (mean age = 13.77) from the U.S.A. Data from the parents of 565 of these children were also analysed. The factor structure of the self and parent reports, the convergence between the two sources of reports and the correlations of the adjectives in the ACL with the scores on the NEO-FFI were all consistent with previous results obtained from adults. The findings largely support the notion that the five-factor model as derived from data from adults is applicable to academically talented youth.

Gender-related differences on the College Board's Advanced Placement and Achievement tests, 1982–1992.

This study explored gender-related differences in participation in and scores on the College Boards Advanced Placement Program and Achievement examinations. In both sets of tests, the numbers of male and female high school students were unequal for several subject areas. The French, Spanish, and Modern Hebrew examinations attracted many more female than male students, whereas many more male than female students took the Physics, Chemistry, and Computer Science examinations. There was little change in most of the patterns of enrollment from 1982 through 1992. Male students had moderately higher scores on the Physics, Chemistry, and Computer Science tests, whereas female students had a slight advantage on some of the language examinations. Although the patterns of gender-related score differences were stable, there were some indications that gender differences are narrowing, especially in the areas of American history and computer science.

GENDER-RELATED DIFFERENCES IN SPATIAL ABILITY AND THE K FACTOR OF GENERAL SPATIAL ABILITY IN A POPULATION OF ACADEMICALLY TALENTED STUDENTS

Abstract Spatial ability is an area of cognitive functioning in which evidence of gender-related differences in performance tends to emerge relatively often. Previous research has found, however, that different types of spatial tasks differ in their amounts of gender-sensitivity. Mental rotation tasks, for instance, have often shown large gender-related differences in favour of males, whereas there are small differences or advantages for females on visual memory tasks. Therefore, the present study examined the profile of differences across a wide array of types of spatial tasks. Two forms of a spatial test battery containing 14 types of items each were administered to academically talented middle and high school students. Factor analysis yielded a strong general factor (termed “ k factor”) underlying performance in both forms. The gender-sensitivity of the item types depended to a considerable extent on the loadings of the subtests on the k factor. When the k factor was partialled out of the variances of the tests, gender-related differences on the various types of items were reduced, often to insignificance, and some tests that had initially exhibited little gender-related variance showed advantages for females. Mental rotation and visualization of perspectives, however, consistently showed substantial gender-related variance beyond the variation explained by the k factor.

Group Data on High School Grade Point Averages and Scores on Academic Aptitude Tests As Predictors of Institutional Graduation Rates

For every 4-year college in the United States listed in the 1998 College Handbook of the College Board, the percentages of students graduating within 6 years of entering and of students having high school grade point averages (GPAs) of at least 3.00 were recorded. The authors also obtained the College Board Scholastic Assessment Test I (SAT I) Verbal and Math and the American College Test (ACT) scores at the 25th and 75th percentiles of the distributions of scores of the enrolled freshmen. The SAT I Verbal and Math and the ACT scores at the 25th and 75th percentiles proved to be good predictors of the percentage of students graduating from the same institution that admitted them as freshmen (rs ranging from .62 to .73), as did the percentage of freshmen having high school GPAs of 3.00 or higher (r = .49). The correlations of the group percentages and means with the criterion were considerably higher than the predictive-validity coefficients of the SAT I and ACT scores for individual graduation as reported in the literature.

The Five-Factor Model of Personality in South African College Students:

The Revised NEO Personality Inventory (NEO-PI-R) was administered to 408 South African college students. When the instrument was factor analyzed at the facet level with varimax rotation, the structure of the five-factor model was well reproduced for the entire sample as well as the White and Black subgroups. Parallel analysis indicated the appropriateness of the five-factor solution. Whereas the structure of personality was highly similar across race, there were statistically significant differences in mean scores of some domains and facets by race. The greatest difference was in Openness to Experience, particularly in the Openness to Feelings facet, with the White subgroup scoring relatively high, the Black subgroup scoring relatively low, and the Indian subgroup scoring in an intermediate range. The authors speculate that these differences are primarily the result of social, economic, and cultural differences between the races rather than the direct product of race itself.

Gender-related differences in cognitive abilities: Evidence from a medical school admissions testing program

Abstract In an investigation of gender-related differences in cognitive ability factors, analyses were undertaken of a series of administrations over a 9-yr period of a comprehensive test battery used to select medical school applicants in West Germany. Fifteen correlation matrices based on a total of 96,968 males and 90,142 females were factor analysed. Three factors were extracted in every case and rotated to an orthogonal simple structure using the Varimax procedure. In every instance, the three factors were identified as reasoning, perceptual speed, and memory with congruence coefficients across administration ranging from 0.89 to 0.99. Highly similar factors were also identified when the data of males and females were factored separately. In all 15 analyses, males scored higher on the reasoning factor than did females, and females scored higher than males on the memory factor, in each case about one-half of a standard deviation. Clear changes over the years were not in evidence, except for a tendency for the female advantage in memory to decline.

Performance factors and gender-related differences in spatial ability : another assessment

To explore whether test-taking styles (performance factors)cancontribute to explaining gender. related differences on tests of spatial ability, 15 spatial tests were administered to three samples of subjects. On each test, number-correct scores and ratio scores (number of items solved divided by the number of items attempted) were computed. In accordance with previous research findings, the use ofratio scores significantly reduced the magnitude ofthe gender-related differences on the Mental Rotations Test. For most of the remaining tests, however, the reduction of the gender-related score difference was small. It was concluded that the difference reduction for the Mental Rotations Test was specific to the format of this test. In common spatial tests, performance factors may account for a small portion of gender-related variance, but the bulk of this variance must be attributed to other factors.