The role of motivational and attitudinal variables in second language acquisition

Includes bibliographical references (pages [159]-168).This study assesses Gardner’s socioeducational model of second language acquisition in elementary school students. A sample of 120 elementary school participants from 4th, 5th, and 6th grades responded to items on a Spanish version of the Attitude/Motivation Test Battery adapted for this study, and the Modern Language Aptitude Test. The scores of the ACCESS for ELLs were used as a measure of second language acquisition. Confirmatory factor analysis (CFA) was used to assess the adequacy of six measurement models: language aptitude, language-learning strategies, language attitudes, motivation, confidence, and second language achievement. The confirmatory factor analyses confirmed that the measurement models adequately represented the theoretical constructs. The relations among the factors were assessed in a full structural model using structural equation modeling. The first assessment of the model showed a large discrepancy and an overall lack of fit. A modification of this model improved fit somewhat, but still failed to attain established goodness of fit criteria. A multiple regression analysis revealed that contrary to the predictions of the model for adult learners, aptitudinal variables are better predictors of second language achievement than motivational variables when the learners are elementary school children. Additional analyses revealed very limited effects of gender and age on language achievement, the use of language strategies, and motivational variables. Possible adaptations to improve the fit of the model in elementary school students include clarifying the role of language-learning strategies in a second language academic setting and further theoretical elaboration of the variables in the motivation factor. Future studies would benefit from a better specification of the factor confidence, maintaining the measures of language aptitude, and attitudinal variables. Additional variables for inclusion in the model are cultural influences, measures of proficiency in the first language, and willingness to communicate as a language outcome. The inadequate fit of the socioeducational model in the present study is interpreted in the context of the paucity of studies with elementary school participants.Ed.D. (Doctor of Education