Literaturnachweis - Detailanzeige
Autor/inn/en | Noble, Julie; Sawyer, Richard |
---|---|
Titel | Alternative Methods for Validating Admissions and Course Placement Criteria. AIR 1995 Annual Forum Paper. |
Quelle | (1995), (28 Seiten)
PDF als Volltext |
Sprache | englisch |
Dokumenttyp | gedruckt; online; Monographie |
Schlagwörter | Stellungnahme; Admission Criteria; College Admission; Correlation; Courses; Decision Making; Evaluation Methods; Higher Education; Institutional Research; Regression (Statistics); Statistical Analysis; Student Placement; Validity Admission; Admission procedures; Zulassungsbedingung; Zulassungsverfahren; Zulassung; Hochschulzugang; Hochschulzulassung; Korrelation; Kursangebot; Decision-making; Entscheidungsfindung; Hochschulbildung; Hochschulsystem; Hochschulwesen; Institutionelle Forschung; Regression; Regressionsanalyse; Statistische Analyse; Schülerpraktikum; Gültigkeit |
Abstract | Correlational methods are compared to an alternative method based on decision theory and logistic regression for providing validity evidence for college admissions and course placement criteria. The advantages and limitations of both methods are examined. The correlation coefficient measures the strength of the linear statistical relationship between one or more measures (e.g., test score or high school grade point average) and college grade point average (GPA). However, correlations are difficult to interpret and provide little direct information about the practical implications of using particular measures and corresponding cutoff scores. They are also hampered by restrictive statistical assumptions and restriction of range problems. For example, course grades or college GPA carry values understood to range from A to F; however, experience with over 80 institutions shows that students seldom receive grades lower than a C. In addition, correlational methods are relevant only for students who are admitted to the institution or to the course, but are not readily applicable to students for whom admissions or placement decisions are yet to be made. Validity indices provide direct evidence of a measure's utility for making admissions or course placement decisions. The practical implications of using specific cutoff scores for all students of interest are shown. Validity indices are less affected than correlation coefficients by restriction of range problems in the measure(s) used to make admissions or placement decisions. This method can effectively use multiple measures for estimating students' chances of success, where success can be defined by a meaningful local standard. The data is presented in several figures and tables. (Contains 12 references.) (SW) |
Erfasst von | ERIC (Education Resources Information Center), Washington, DC |