Learner variables as predictors of ESL proficiency

Abstract

While all humans exhibit inherently human traits of learning, every individual approaches a problem or learns a set of facts or organizes a combination of fee lings from a unique perspective. Therefore, one of the major issues in the Second Language Acquisition (SLA) field is the question of differential success among language learners. Research on SLA has identified a variety of variables hypothesized to account for some of the variance in the level of proficiency attained by individuals learning a second language. However, more research is needed before statements can be made about which combination of learner variables is ultimately crucial to Second Language Acquisition in a particular setting. The purpose of this study was to establish the relationship between the field independence/dependence (FI/D), language learning strategies (LLSs), personality types/traits (PT) and ESL proficiency of Afrikaans first year university students studying English as a second language. The methodology employed in this study was discussed under four main headings: subjects, instrumentation, data collection procedure, and design and analysis. A total number of 305 Afrikaans first year students at the Potchefstroom University taking English were included in this study. Five paper-and-pencil instruments were used: * The Gottschaldt Figures Test (GFT), * The Strategy Inventory for Language Learning (SILL), * The Jung Personality Questionnaire (JPQ), * The High School Personality Questionnaire (HSPQ), and * The Test of English as a Foreign Language (TOEFL). The tests for the predictor variables were group-administered during scheduled afternoon tutorial periods in April 1991, while the criterion test was group administered towards the end of June 1991. The data were analysed by using the "Statistical Analyses System" (SAS) programmes (1988). The following analyses were used: Pearson product-moment correlations, canonical correlations and stepwise multiple regression. The results indicated: a statistically significant, but not practically significant, relationship between FI/D and ESL proficiency, a statistically significant as well as a practically significant relationship between LLSs and ESL proficiency, a statistically significant, but not practically significant, relationship between only two personality traits and ESL proficiency, and that LLSs accounted for approximately 45% of the total variance on the TOEFL test. The results indicated the importance and significance of LLSs in predicting ESL proficiency. As a result various factors which could influence LLS use were also assessed. The results indicated: that ENG 111 students used LLSs far more frequently than ENG 112 students and that the ENG 111 students were more proficient language learners than the ENG 112 students, that females differed significantly from males in their LLS use; females using LLSs more frequently, that course status had a significant influence on LLS use, and that major field of study significantly influenced LLS use. The results indicated that LLSs were the most significant predictors of ESL proficiency. The contribution of the other variables that were investigated was small, however, it is not possible to discount their influence. It would rather seem as if a combination of different variables is necessary to predict ESL proficiency successfully. Teachers/researchers are well advised to bear in mind that "the good language learner" has not yet been defined. An overriding and all-pervading variable that classifies learners neatly into categories of "successful" and "unsuccessful" has not yet been identified.