An exploratory study of the hypothesis of divisible versus unitary competence in second language proficiency

Abstract

In this research Oiler's question 'Is language proficiency divisible into components?' was explored by determining which of three models best fit the experimental data: a model postulating numerous specific sources of variance (the extreme divisible model), a model postulating a single, large source of variance (the unitary model), or a model postulating a large general factor and several smaller specific factors. Following analysis of data gathered in a preliminary study, four tests which had clearly recognizable contrasts in content (grammar vs. vocabulary) and mode (listening vs. reading) were constructed to identify linguistic and method variance in a correlation matrix of language proficiency variables. These four measures were pilot tested, revised, and administered in conjunction with eight other language measures to a group of beginning-level ESL learners. The data were factor analyzed using image analysis to explore the relative congruency of the three models to the data. In addition, the relationships between the tests and the demographic variables age, sex, length of time in English Canada, and first language were also investigated. In the factor analysis, both of the methods used to determine the number of factors to be retained in the final solution indicated three. (The methods used were the Kaiser-Guttman criterion of selecting factors with eigen values greater than one in a principal components analysis and the inspection ol' a varirax rotation of a full image analysis to