Richard J. Rankin

A Test of the Internal/External Frame of Reference Model at Different Levels of Math and Verbal Self-Perception

In this study, predictive relations were examined between math and verbal achievement and several different measures of motivation. The sample consisted of 348 sixth grade and 325 ninth grade Norwegian students. Motivation variables of self-concept, self-perceived aptitude, and self-perceived ability to learn within each domain were highly correlated and used as indicators to define math and verbal latent variables labeled general math and verbal self-perceptions. Subscales measuring self-perceived math and verbal skills were used to define another set of math and verbal latent variables labeled specific math and verbal self-perceptions. Both general and specific math and verbal self-perceptions were strongly related to corresponding achievement. Predictions from Marsh’s internal/external frame of reference model were confirmed for math and verbal general self-perceptions but not for specific self-perceptions. General self-perceptions strongly affected intrinsic motivation, self-perceived effort, and anxiety in the respective domains, whereas specific self-perceptions did not. The pattern of results was quite similar for the two grade levels.

Math, Verbal, and General Academic Self-Concept: The Internal/External Frame of Reference Model and Gender Differences in Self-Concept Structure.

In this study, predictions were tested from Marshs (1986) internal/external (I/E) frame of reference model (measuring cognitive dimensions of math and verbal self-concepts), and gender differences in the structure of academic self-concept were examined through path analyses of data from 231 Norwegian sixth-grade students (117 boys and 114 girls)

Math and Verbal Achievement and Self-Concepts Testing the Internal/External Frame of Reference Model

Predictions from the Marsh Internal/External Frame of Reference Model were tested on two groups of sixth-grade Norwegian students. Data were analyzed through path analysis for latent variables using structural analysis. The predictions were supported for students who perceived their level of math and verbal achievement as substantially different (n = 309): Math and verbal self-concepts were not correlated and there were significant negative correlations between verbal achievement and math self-concept and between math achievement and verbal self-concept. However, the predictions were not supported for students who perceived their level of achievement in the two areas to be relatively equal (n = 138): Math and verbal self-concepts were strongly correlated and there were no significant negative correlations between achievement in one area and self-concept in the other area. The results support the theoretical basis of the I/E Model but indicate that it offers different predictions for different students.

WPPSI Reliability and Predictive Validity With Disadvantaged Mexican-American Children.

Abstract The reliability and predictive validity of the WPPSI was investigated with a sample of lower SES Mexican-American children. Reliability was determined using split-half procedures parallel to those reported in the test manual. Predictive validity was determined by correlating WPPSI scores with scores from the Metropolitan Reading Test. The results revealed that the WPPSI has high reliability, as determined by internal consistency procedures, but validity coefficients were not significant. These results suggest that the WPPSI may be inappropriate for special program placement purposes with this population.

K-ABC content bias: comparisons between Mexican American and white children

Content (item) bias on the Mental Processing Scales and the Achievement Scale of the Kaufman Assessment Battery for Children (K-ABC) was investigated with 100 Mexican American and 100 White fifth- and sixth-grade boys and girls. All children were English-speaking and from similar socioeconomic status (SES) backgrounds. An item-group (partial correlation) method that controlled for age, sex, and ability was used to examine bias. On the Mental Processing Scales, 17 (14%) of 120 items were found to be biased—the strong majority against the Mexican American children. On the Achievement Scale, it was observed that 58 (63%) of 92 K-ABC items were biased—all against the Mexican American sample. This latter finding is discussed in the context of possible differences in learning opportunities (language based, SES based, and school segregation) experienced by the two ethnic groups.

Cultural Modification of Effect of Family Size on Intelligence

Previous studies of the relationship between family size and intellectual level suggest a negative relationship between these two variables. This study presents data to show that in at least one culture the relationship between family size and ability is neither linear nor negative. In a culture where large families are the norm, children from families close to mean size have higher ability than those from families at either extreme.

Evidence of Bias in Predictive Validity on the Kaufman Assessment Battery for Children in Samples of Anglo and Mexican American Children.

Possible bias in the differential predictive validity of the Kaufman Assessment Battery (K-ABC) was investigated with 76 Anglo and 90 Mexican American fifth- and sixth-grade boys and girls. All children were English-speaking and from similar socioeconomic backgrounds. The criterion variable was the Comprehensive Tests of Basic Skills (CTBS; Language, Reading, Mathematics, and Total Scores). Several statistical techniques were used to investigate test bias (examination of predictive validity coefficients; two methods of examining homogeneity of slopes of the regression lines). The results showed considerable evidence of bias in differential predictive validity, indicating that the global cognitive score of the K-ABC (the Mental Processing Composite) was less effective in predicting CTBS scores for the Mexican American group than for the Anglo group.

WISC-R factor structures among white, Mexican American, and African American children : A research note

The factor structure for the Wechsler Intelligence Scale for Children-Revised (WISC-R) was investigated for White (n = 183), Mexican American (n = 129), and African American (n = 139) nonreferred school-age children (6.8 to 14.6 years) of low-to middle-class socioeconomic background. Exploratory and confirmatory factor analyses were performed. The WISC-R factor structure model tested was Kaufmans (1975) three-factor solution that consisted of Verbal Comprehension (VC, Factor 1), Perceptual Organization (PO, Factor 2), and Freedom from Distractibility (FD, Factor 3). The results of the exploratory analyses showed Kaufmans three-factor solution for the three ethnic groups, but the order of Factor 2 and Factor 3 was reversed for the Mexican American and African American samples—thus raising questions about the comparability of the WISC-R factor structure across groups. The confirmatory analyses, which used the orthogonal nested factors approach by Gustafsson and Balke (1993), resulted in a model (for each of the ethnic groups) in which a G factor explained most of the variance (followed by the nested first-order factors—VC, PO, and FD). The results of the confirmatory analyses provide some support for David Wechslers (1974) original intentions that the WISC-R structure is best described as having a general, or global, factor.

Dimensions of Math and Verbal Self-Concept and the Internal/External Frame of Reference Model

A major recent development in the self-concept (SC) literature is the formulation of the Internal/External Frame of Reference Model (I/E model). The purpose of this study was to test the I/E model on a Norwegian sample both when measuring math and verbal selfesteem (general feelings of doing well or poorly in math and verbal areas) and math and verbal self-efficacy (expectations of being able to perform defined and concrete tasks in each area). Previous studies of the model, with the exception of a study by Marsh, Walker, and Debus (in press) referred to below, have either measured math and verbal self-esteem in North-American or Australian samples (Byrne & Shavelson, 1987; Marsh, 1986, 1990; Marsh, Byrne, & Shavelson, 1988) or math and verbal self-efficacy in a Norwegian sample (Skaalvik & Rankin, 1990). These studies have shown quite different results (see below). This calls for further analysis and elaboration of the model.

Focusing on the Selection Process: A Field Study of the Measures that Influence Admission Decisions:

This field study examines the measures used at multiple stages in a selection process employed by a universitys teacher training program. Specifically, the study utilizes LISREL structural analysis to assess the importance each measure assumed in determining admission and intermediate outcomes. Of the measures, which included ability tests and a situational interview, only the mathematics, language, and essay scores prophesied selection. Neither the pre-interview overall qualifications rating nor the interview score proved significant in predicting admission, suggesting a discrepancy between official and operative emphases. Age and race emerged as non-significant, and sex as of minor importance. An analysis of the results supports the position that demographic variables should be investigated in field settings in which raters possess substantial pertinent applicant information. The study concludes with a discussion of the need to elevate the level of analysis in selection research in order to understand the nature and validity of decisions made during the process itself.