Word Association Data and the Assessment of Bilingual Education Programs

Abstract

A variation of the standard word association technique was used to investigate the relative language skills of a group of first grade children schooled bilingually and compared with their monolingually instructed counterparts. The results suggested that both language of instruction and language of testing affected the percentage of different responses, blanks, and other-language responses. The Ss generally performed better in their native language than in their second language. This technique may be a useful short-cut evaluative tool for researchers or teachers affiliated with bilingual education programs. Educators in diverse countries, including Canada and the United States, are often faced with the task of teaching their pupils via a weaker or second language. In such situations it seems natural to consider the adoption of some form of bilingual instruction; and, in fact, both Canada and the United States have recently allocated Federal funds in an attempt to stimulate the development of model bilingual education and second language teaching programs. A very interesting and comprehensive review of bilingual schooling in North America has recently been completed by Andersson and Boyer (1970). The adherents of bilingual education believe that there may be certain cognitive and attitudinal advantages to the early introduction and maintenance of bilingual education programs. The programs have as their goal the development of bilingual-bicultural individuals who are able to interact successfully at various levels of North American society. Given this goal, an evaluation of the successes, failures, or limitations of these early programs assumes paramount importance. We realize how difficult it is to reliably assess the relative language skills