Self-Assessment, French Immersion, and Locus of Control

Abstract

This article compares the self-assessments of French proficiency made by approximately 500 Grade 8 students in two different French immersion programs (‘early’ and ‘middle’) in Toronto, Canada. Two self-assessment benchmarks are used: the perceived language proficiency of francophone peers and the difficulty represented by specific everyday tasks in French. The study investigates: (1) the extent to which self-assessment is a valid and reliable indicator of tested proficiency in French immersion programs; (2) how benchmarks influence correlations of self-assessment with tested proficiency; (3) whether self-assessment research can inform or support current theories of second language learning and assessment. The results indicate that: (1) selfassessments of language proficiency correlate only weakly with objective measures of language proficiency; (2) self-assessment measures on specific tasks are more highly correlated with tested proficiency than are global selfassessment measures; (3) irrespective of program, students agree on the relative difficulty of oral and literacy tasks in French under specific conditions of reception and production. These findings are explained with reference to current research on self-assessment, Spolsky’s Conditions for Second Language Learning (1989), and the authors’ construct of ‘locus of control’ in a communicative event. It is argued that the locus of control operates at the ‘interface’ (Bachman 1989) between language assessment and second language acquisition research.