A Comparison of Two Methods of Assessing Partial Knowledge on Multiple-Choice Tests.

Autor/inn/en	Tollefson, Nona; Chung, Jing-Mei
Titel	A Comparison of Two Methods of Assessing Partial Knowledge on Multiple-Choice Tests.
Quelle	(1986), (17 Seiten)Verfügbarkeit
Sprache	englisch
Dokumenttyp	gedruckt; Monographie
Schlagwörter	Achievement Tests; Comparative Analysis; Evaluation Methods; Graduate Students; Guessing (Tests); Higher Education; Multiple Choice Tests; Reliability; Scoring Formulas + Suchen Sie Ihr Suchwort? Achievement test; Achievement; Testing; Test; Tests; Leistungsbeurteilung; Leistungsüberprüfung; Leistung; Testdurchführung; Testen; Graduate Study; Student; Students; Aufbaustudium; Graduiertenstudium; Hauptstudium; Studentin; Erraten; Hochschulbildung; Hochschulsystem; Hochschulwesen; Multiple choice examinations; Multiple-choice tests, Multiple-choice examinations; Multiple-Choice-Verfahren; Reliabilität; Scoring rubrics; Auswertungsbogen
Abstract	Procedures for correcting for guessing and for assessing partial knowledge (correction-for-guessing, three-decision scoring, elimination/inclusion scoring, and confidence or probabilistic scoring) are discussed. Mean scores and internal consistency reliability estimates were compared across three administration and scoring procedures for multiple-choice tests: (1) number right; (2) Coombs elimination response method; and (3) inclusion. It was hypothesized that the mean scores of subjects administered an achievement test under experimental inclusion-of-correct- and elimination-of-incorrect-alternatives directives would be significantly different from the mean score of subjects administered the same test under number-right directives. Seventy-five graduate students in an introductory statistics class were randomly assigned to one of three administration and scoring procedures. A 36-item multiple-choice test with 4-response alternatives per item was used to assess the effect of the different methods. A comparison of mean scores indicated no significant between-group effect. Elimination and inclusion scoring produced higher internal consistency reliability estimates than did number-right scoring. Two data tables conclude the document. (TJH)
Erfasst von	ERIC (Education Resources Information Center), Washington, DC