• Calculating the correlation coefficient between selected ability and achievement tests using validity generalization

      Sabers, Darrell; Shiroma, Paul Shigeo, 1960- (The University of Arizona., 1989)
      The correlation coefficient between an ability test and achievement test is necessary in order to estimate the effects of the regression of IQ on achievement when calculating a discrepancy between them for the purpose of classifying a child as learning disabled. Weighted mean correlations and their respective variances were computed across studies for one of 11 ability-achievement test pairs using a meta-analysis procedure called Validity Generalization. The results indicated that there is no "global correlation coefficient" that will accurately represent the relationship between all ability and achievement tests. Furthermore, critical data necessary to adjust correlations and their respective variances for statistical artifacts were not reported in the literature. Thus, the results obtained from the subgroups could be due to capitalization on chance.
    • Gender identity and educational achievement: Correlations of children's self perception with educational achievement

      Sabers, Darrell; Warren, Annmarie Maione, 1968- (The University of Arizona., 1993)
      The Children's Self-Perception Scale (CSPS), an instrument to measure self-perceived gender identity in school age children, was developed based on the Bem Sex Role Inventory (BSRI; Bem, 1974). The children's scores on the CSPS were correlated with two measures of educational achievement, grade point average (GPA) and Iowa Test of Basic Skills (ITBS; Hieronymus, et al., 1986) to determine the relationship between femininity and masculinity in males and females and educational achievement. The subjects were 950 third, fourth, and seventh grade students from two school districts. The majority of students was Hispanic. The CSPS was not able to differentiate reliably between masculinity and femininity; reliabilities for the CSPS ranged from .17-.78. Although results were inconsistent across grades, significant (p<.05) positive correlations were found between femininity and GPA, .25, and masculinity and achievement (.23-.31) for seventh graders in only one district. Several possible explanations for the non-significance and inconsistencies are explored.
    • Reliability of the Arizona Clinical Interview Rating Scale: A confirmatory study

      Sabers, Darrell; Fulginiti, John Vincent (The University of Arizona., 1988)
      Reliable measurement of student capability for a skill allows educators to verify student mastery. A major part of a physician's ability to gather information involves patient interviewing, and instruction of this skill is a substantial portion of a medical curriculum. Since 1974, the University of Arizona College of Medicine has employed patient-instructors (PIs), lay persons who function in the roles of patient and teacher for training of interview skills in the Preparation for Clinical Medicine (PCM) program. PIs provide "real" patient-interview experiences and immediate feedback to the students. The PCM program currently has four topic areas: Adult, Pediatric, Geriatric, and Psychiatric. The Arizona Clinical Interview Rating (ACIR) Scale was developed in 1976 to measure the technical performance aspects of interviewing. This study was undertaken to determine reliability of the ACIR. Implication of the results are discussed and suggestions made for the continued application of the ACIR Scale. (Abstract shortened with permission of author.)
    • Self-concept and interpersonal relations in Hispanic and non-Hispanic adolescents

      Sabers, Darrell; Peterson, Jennifer Bandy, 1963- (The University of Arizona., 1991)
      This study examines the self concept and interpersonal relations of 409 Hispanic and Non-Hispanic adolescents in grades five and eight. Results reveal no significant differences in global self concept of Hispanic and Non-Hispanic adolescents. Fifth graders have significantly better global self concepts than eighth graders. There are no significant differences in the interpersonal relations of the Hispanic and Non-Hispanic adolescents. Fifth graders report significantly better Total Interpersonal Relations, as well as better relations with Mother and Teacher. Relations with Father are significantly better for males and for fifth graders. Significant grade interactions are seen for relations with Male Peers and Female Peers. A multiple regression revealed no clear shift from parents to peers as reference group influencing self concept.
    • Truncation and its effect on standard error of correlation coefficients

      Sabers, Darrell; Durney, Ann Wells (The University of Arizona., 1990)
      A Monte Carlo study was conducted to investigate the effect of truncation of score distributions on systematic bias and random error of correlation coefficient distributions. The findings were twofold: Correlation decreases systematically due to increasing truncation; and the standard error of the correlation coefficient, which is a measure of random error, increases due to increasing truncation.
    • The use of two measures with young gifted children

      Sabers, Darrell; Visvanathan, Thara, 1966- (The University of Arizona., 1992)
      This study examined the performance of young gifted children on two measures. The sample included 25 children. The measures used were the DIAL-R and the WJ-R. Results obtained on the DIAL-R were used to confirm the development of the children. Scores on the DIAL-R were correlated with those on the WJ-R. Rankings of students by their teachers was also obtained as added information. Teacher preference of the two measures was used to determine which tests were more likely to be employed by the school in the future. Only 4 children were identified "potential advanced" on the language area and 3 of them with respect to the sum of the area scores on the DIAL-R. Most teachers preferred using the WJ-R over the DIAL-R with respect to the information it provided.