Evaluation of Bilingual Education Programs: Problems and Some Solutions.

Autor/in	Swain, Merrill
Titel	Evaluation of Bilingual Education Programs: Problems and Some Solutions.
Quelle	(1976), (22 Seiten)Verfügbarkeit
Beigaben	Tabellen
Dokumenttyp	gedruckt; Monographie
Schlagwörter	Tagungsbericht; Bilingual Education; Bilingual Schools; Bilingual Students; Bilingual Teachers; Educational Assessment; English; Evaluation Methods; French; Program Evaluation; Research Design; Research Problems; Second Language Learning; Statistical Analysis; Test Interpretation; Testing; Testing Problems + Suchen Sie Ihr Suchwort? Bilingual teaching; Bilingualer Unterricht; Bilingual scholl; Bilinguale Schule; Education; assessment; Bewertungssystem; English language; Englisch; Französisch; Programme evaluation; Programmevaluation; Forschungsdesign; Forschungskritik; Zweitsprachenerwerb; Statistische Analyse; Test analysis; Testauswertung; Testdurchführung; Testen
Abstract	One of the overriding problems of any psychometric study is that aspects of behavior that are nonquantifiable or difficult to quantify are not considered. The other side of this coin is that those aspects of behavior that are easiest to quantify tend to be what are examined. In evaluating bilingual education programs, problems are found in the research design, the students tested, the nature of the tests used, the methods of analysis employed, ano the interpretation of findings. Research design has usually involved comparison of bilingual program students with students in a traditional program. This is easier than analyzing the goals and objectives of the program and what a student should achieve, but not always accurate. Selection of a control group is difficult because: (1) the characteristics of bilingual education students are not always known; (2) significant characteristics may be difficult to quantify; and (3) some characteristics may not be present among students in regular programs. There are three main problems in test selection: (1) neither standardized tests nor specially developed tests give accurate comparisons between the special and regular program students; (2) the content measured in tests is limited, and their data are used too extensively; and (3) tests tend to be the only accepted means of obtaining performance data. Statistical comparisons are needed but are overused. The methods of analysis most commonly used do not take into account possible interrelationships between separate test scores in a battery. Misunderstanding of statistics and, most importantly, a lack of information about the program being evaluated are two problems with the interpretation of findings. Many evaluation problems can be overcome by supplementing the data with case studies of students, observational data, and specific information about program structure, progress and content. (CHK)
Erfasst von	ERIC (Education Resources Information Center), Washington, DC
Update	2004/1/01