Search

The principal aims of this study, a conceptual replication of an earlier investigation of the TOEFL[R] computer-based test, or TOEFL CBT, in Buenos Aires, Cairo, and Frankfurt, were to assess test takers' reported acceptance of the TOEFL Internet-based test, or TOEFL iBT[TM], and its associations with possible determinants of this acceptance and with test performance; evaluate differences in the pattern of results for test takers from different countries; and compare the findings with those for the TOEFL CBT. A questionnaire concerning attitudes about the test and other relevant variables was administered by the Internet to TOEFL iBT examinees in 4 diverse countries with large testing volumes: China, Colombia, Egypt, and Germany. Overall attitudes about the TOEFL iBT were moderately positive in most countries, but neutral in Germany; attitudes about the Listening and Writing sections of the test were uniformly favorable in every country; but attitudes about the Speaking section were consistently less favorable in all countries and were unfavorable in Germany and Colombia. Attitudes about the test had similar patterns of relationships in the 4 countries: moderate correlations with attitudes about admissions tests in general, slight and inconsistent correlations with TOEFL performance and computer anxiety, and minimal correlations with other variables. All in all, these results were very similar to those in the earlier investigation, with 1 exception: the previous study found uniformly moderate positive attitudes about the TOEFL CBT in the 3 countries surveyed. (Contains 7 tables.)

This study explored individual differences in educational disadvantage-- deficits in formal and informal education in the school, home, and elsewhere--in the SAT test-taking population. Factor analysis identified six educational disadvantage factors--four concerning the students' schools and two the students' nativity and parenting--and one family socioeconomic status factor, race/ethnicity, high school grades, and SAT scores. The individual-differences perspective on disadvantage appears to be a viable one, and educational disadvantage seems to be a meaningful and useful construct. The following are appended: (1) Letters; and (2) Questionnaire.

Laboratory experiments by Steele and Aronson (1995) found that African-American subjects' performance on difficult verbal items, described as a verbal problem-solving task, was adversely affected when they were asked about their ethnicity just before working on the items. These results were attributed to stereotype threat. Asking about ethnicity primes African-American subjects' concerns about fulfilling the negative ethnic stereotype about their intellectual ability, thereby disrupting test performance. The present field experiment assessed the effects of asking community college students taking the Computerized Placement Tests (CPTs), in an actual operating setting, about their ethnicity and sex. This inquiry had no statistically and practically significant effects on how well the examinees did on the tests or how long they worked on the tests

Steele and Aronson (1995) found that the performance of African-American subjects on test items portrayed as a problem-solving task, in a laboratory experiment, was adversely affected when they were asked about their ethnicity. This outcome was attributed to "stereotype threat". Performance was disrupted by the subjects' concerns about fulfilling the negative stereotype concerning African Americans' intellectual ability. Extending that research, this field experiment evaluated the effects of inquiring about ethnicity and sex on the performance of examinees taking the Advanced Placement Program Calculus AB Examination in an actual test administration. With a minor exception, this inquiry had no statistically and practically significant effects on the test performance of African-American, female, or other subgroups of examinees. Answer sheets are appended.

This study examined three questions about measures of accomplishments--notable attainments that have been publicly recognized: their pseudoipsativity, the correspondence between quantity and quality scores, and their dimensionality. Comparable samples of graduate students described their accomplishments on a questionnaire or judged the similarity of the same accomplishments. Sample 1 consisted of 204 first-year graduate students, and sample 2 consisted of 75 students from four colleges. The Accomplishments Questionnaire consisted of 52 items in 6 areas. Accomplishments in the same field were positively correlated, while accomplishments in different fields varied in their correlations, some being positive and others being negative; these results are inconsistent with the predominantly negative correlations that would occur with ipsative measures. Measures of the quantity and quality (importance, rarity) of accomplishments correlated highly, after correction for attenuation, and appeared to assess the same thing. Analyses of self-report data found that accomplishments were factorially complex, with many of the same factors identified in other kinds of measures appearing. The factors observed in the judgment data largely corresponded to the semantic features of the accomplishment items (Contains 4 tables and 25 references.) (Author/SLD)

This study appraised the validity of SAT scores, in combination with grades in high school courses and the number and difficulty level of these courses, in predicting college grades in various fields of study, with the objective of providing SAT takers with predictions of their academic performance in different fields for guidance purposes. The possible impact of this feedback on the flow of students into specific major fields was also assessed. Data on an entering class at a large state university provided the basis for the study. The SAT and collateral variables were able to predict college grades in different academic areas by taking into account marked variations in grade distributions among fields of study. These predictions could be potentially useful in academic guidance. Students' predicted grades in the various fields and their intended majors were unrelated. Most students received their highest predicted grades in the humanities, although few intended to major in this area. Hence, such feedback would probably stimulate students to take courses or major in the humanities. Table A-1--Standardized Regression Coefficients for Predicting Actual and Imputed GPAs for Three Fields of Study--is appended.

The relationships between examinee background characteristics and performance on the Graduate Record Examinations (GRE) General Test were appraised by a structural equation modeling analysis. The initial characteristics of 3,145 examinees (i.e., sex, ethnicity, parental education, geographic region, and age) had modest relationships with their test performance. Of these, parental education had the most consistent and strongest association. Sex also had an appreciable association, but it was limited to the quantitative score. College-related characteristics (e.g., college major and the institution's public vs. private control, Carnegie classification, selectivity, and Ph.D. productivity) and undergraduate grade point average (GPA) generally had stronger and more pervasive relationships with test performance than did the examinees' initial characteristics, not only by mediating the associations of the examinees' initial characteristics with test performance but also by making independent contributions in their own right. The associations were especially strong for school quality (e.g., a composite of public vs. private control, selectivity, and Ph.D. productivity), college major, and undergraduate GPA. Five tables and three figures illustrate the analysis. (Contains 38 references.) (Author/SLD)

This study compared the effectiveness of several existing and proposed methods for statistically adjusting college grade point averages (GPAs) for course and departmental differences in grading standards, using first-semester grades from an entire entering class at a large state university (4,351 students), in 1988. Most of the adjusted GPAs produced by these methods functioned similarly and, despite high correlations with actual GPA, had greater internal-consistency reliability than actual GPA and were more predictable from Scholastic Aptitude Test (SAT) scores and high school rank (HSR). Most of the adjusted GPAs also functioned similarly with regard to sex differences in over-underprediction. The adjusted GPAs and actual GPAs exhibited the same small but significant sex differences in over-underprediction by SAT scores, but the adjusted GPAs displayed smaller differences than actual GPAs in over-underprediction by SAT scores and HSR. Seven tables present analysis results. (Contains 44 references.) (Author/SLD)

The usefulness of response latency data for biographical inventory items was assessed for improving the inventory's validity. Focus was on assessing whether weighting item scores on the basis of their latencies improves the predictive validity of the inventory's total score. A total of 120 items from the Armed Services Applicant Profile (ASAP) were administered via computers to 1,090 male Navy recruits at the Recruit Training Center in San Diego (California); the regular score, latency-weighted scores, and measures of deviant latencies were obtained. The latency-weighted scores did not improve the ASAP's validity in predicting 6-month retention beyond that obtained with the regular score; the deviant latency measures did not function as suppressor or moderator variables to increase the ASAP's validity. Subgroups of items with differing latencies varied systematically in their internal-consistency reliability (with increased reliability for subgroups with shorter latencies), and a small subgroup of items with moderate latencies was almost as valid as the regular score, suggesting that latency data may be useful in writing and selecting inventory items. Five tables and four graphs complement the study. A 34-item list of references is included. (Author/SLD)

The aim of this study was to appraise whether different forms of the Scholastic Aptitude Test (SAT) used since the mid 1970s varied in their correlations with academic performance criteria in the same cohort of examinees. A 1975 form and a 1985 form were administered to equivalent samples of high school juniors, and self-reported grade-point average and high school rank were obtained. The SAT Verbal and Mathematical scores generally had similar correlations with the grade criteria in the two samples, but the Verbal score had a significantly higher correlation with school rank in the 1985 sample. The principal conclusion is that the 1975 form of the SAT does not have greater validity than the 1985 form in assessing academic performance, at least at the high school level. This outcome offers no support for the hypothesis that the decline in the SAT's ability to predict college grades since the mid 1970s, observed in recent research, is attributable to changes in the test. Three tables present analysis details. (Contains seven references.) (Author)

This study determines major dimensions of social stratification for whites and blacks and explores the existence of distinct social classes. It is based on a survey conducted on 225 white and 206 black household heads in Toledo, Ohio, using a highly structured interview that included variables reflecting major theoretical dimensions of stratification, important indexes, and well established correlates. Eighteen first-order factors were found for whites and 18 for blacks. Five factors matched in the two samples: social status, residence, organization activity, political activity and efficacy, and main support's social status. Other important factors, unique to each sample were: for whites, self-employed and majority group membership, and for blacks, class consciousness and method variance. The present findings indicate that stratification is more complex than anticipated by current conceptualization, previous research, and existing indexes. (Author/AM)

A study examined the relationship of native language and level of English proficiency to the structure of the Test of English as a Foreign Language (TOEFL). Using all of the information provided by various responses to the test's items (the four alternatives, omitted, and not reached), the items' interrelations were analyzed by three-way multidimensional scaling for samples of examinees systematically varying in native language and level of English proficiency. Four dimensions were identified: three corresponded to the sections of the test, and the fourth was an end-of-test phenomenon. The dimensions were predominantly defined by easy items and were most salient for low-scoring examinees. Native language had little influence on results. The major conclusions were that the TOEFL's construct validity is supported, the test's interpretation varies with the examinees' English proficiency, easy and difficult items differ in their potential for diagnosis and global screening, and the dimensionality of the TOEFL and of competence in English depends on the examinees' English proficiency. (Author/MSE)

The aim of this study was to construct and validate a biographical inventory to measure personality traits that are predictive of leadership. The experimental inventory, consisting of tentative scales for Dominance, Emotional Stability, Need for Achievement, Self-Confidence, and Sociability, was administered to 642 incoming midshipmen at the Naval Academy. The tentative scales were item analyzed, and the validity of the final scales was appraised against subsequent peer ratings of the traits and of leadership, as well as against first-semester grades that may reflect leadership (Military Performance and Professional Military Quality Point Rating). The Sociability, Dominance, and Need for Achievement scales demonstrated convergent and discriminant validity and correlated with the leadership criteria; the Emotional Stability and Self-Confidence scales showed no signs of validity. Although the level of validity was often modest, the Sociability scale correlated appreciably (0.28) with the peer rating of leadership, suggesting that this measure may have practical value in assessing leadership potential. (Contains 1 figure, 15 tables, and 16 references.) (Author/SLD)

The Myers-Briggs Type Indicator is a Jungian-oriented self-report inventory which classifies people into dichotomous categories along four scales: extraversion-introversion, sensation-intuition, thinking-feeling, and judgment-perception. Introduced in 1942, the scoring and items have changed after several items analyses. Form F contains 51 phrase questions and 44 word pairs. A critical review of reliability and validity studies raises four issues to consider when evaluating the Indicator: (1) correspondence between Jung and Myers typology; (2) adherence to Myers' typology; (3) predictive validity; and (4) uniqueness. Myers has a different concept of the extroversion-introversion and judging-perceiving scales, but the scales do not reflect this concept. The Indicator generally does not reflect Myers' typology: classifications do not appear unchanging, interacting, or associated with different compounds of surface traits; in addition, type indeterminacy is unrelated to ineffective behavior and maladjustment. The Indicator may be useful for grade prediction in liberal arts schools, but its ability to identify vocational interest and creativity is limited. Finally, only the thinking-feeling scale is unique among existing scales; but factor analysis may uncover other unique properties. Norms are presented for high school, college, and graduate students; recently employed college graduates; and public school teachers. (CP)

The usefulness of Scholastic Assessment Test (SAT) scores in combination with grades in high school courses and the number and quality of these courses in predicting college grades in various fields was studied for 981 college freshmen. The SAT and collateral variables were able to predict grades in different areas and could be useful in guidance. (SLD)

The relations between examinee background characteristics and performance on the Graduate Record Examinations (GRE) General Test were studied through structural equation modeling for 3,145 examinees. Findings regarding examinee characteristics underscore the importance of socioeconomic status and intervening process variables in research on group differences in performance on cognitive tests. (SLD)

The aim of this study was to delineate departmental differences in the length of time that doctoral students take to receive their degrees and the institutional characteristics linked with it, being as comprehensive as possible in the institutional characteristics examined and disentangling the confounding effects of student characteristics. Variables describing graduate departments in three disciplines and their parent universities were obtained from 86 departments of Chemistry, 57 departments of English, and 62 departments of Psychology. In analyses of departments in each discipline, a relatively small set of institutional variables correlated with average time to the doctorate for each department, after controlling for characteristics of students in the department, and these relationships differed from discipline to discipline. The clearest and most extensive findings emerged for Psychology: the institutional variables identified primarily concerned faculty accessibility (student/faculty ratio and department size), with longer time required in departments with many students per faculty member or with many students. Student characteristics accounted for about half of the departmental variation in time in each discipline. An appendix contains four tables describing departmental characteristics. (Contains 6 tables, 17 figures, and 40 references.) (Author/SLD)

The role played by sex-related differences in prediction of grades from Scholastic Aptitude Test scores with regard to the grade criterion and the variables associated with academic performance was studied for an entire university first-year class (4,351 students). Sources of underprediction for women are discussed. (SLD)

This study examined the role that sex-related differences in the nature of the grade criterion and in variables associated with academic performance play in the over- and under-prediction of college grades by the Scholastic Aptitude Test (SAT) both when the test is used alone and in conjunction with high school grades. An entire freshman class (4,351 students) at a large state university was studied. Women's grade point average (GPA) was underpredicted, but only by a small amount. Adjusting the GPA for differences in grading standards for individual courses did not affect the underprediction, but controlling for sex differences in individual-difference variables concerned with academic preparation, studiousness, and attitudes about mathematics reduced or essentially eliminated the underprediction. Eight figures and 22 tables present results of the analysis. Two appendixes contain classifications and scores, and the study questionnaires. (Author/SLD)