Assessing university students' epistemological beliefs about foreign language learning

The purpose of this thesis was to develop a self-report survey that measures foreign language students' epistemological beliefs about language learning. Epistemological beliefs are those concerned with the nature of knowledge and learning. A secondary purpose was to test the unique use and combination of a variety of innovative research methods, including focus groups, specification-driven criterion-referenced language test development, concept mapping, and web-based survey development and administration. The theoretical assumption underlying survey development is found in the Theory of Reasoned Action, which suggests that beliefs are partial determinates of attitudes; attitudes are partial determinates of intentions; and intentions are partial determinates of behaviors or learning strategies. The resulting Foreign Language Beliefs Survey---Epistemological (FLBS-E), clearly measures four primary beliefs constructs: (a) Development in Language Learning; (b) Student Initiative; (c) Ability; and (d) Ease/Difficulty of language learning. Although the FLBS-E displays sufficient internal consistency to be measuring an overall construct of beliefs about language learning, defining that construct requires additional research.;Pearson correlations between the FLBS-E and Schommer's Epistemological Questionnaire were modest, suggesting that the FLBS-E is measuring somewhat different constructs than those traditionally found in research on epistemological beliefs. Results further suggest that beliefs about language learning may be domain specific (i.e., different from beliefs about other learning domains). A possible model for better understanding beliefs about language learning is discussed. This model proposes a relationship between students' beliefs and their motivations and strategies that is bridged by their beliefs about development in language learning and initiative in language learning.;Concept mapping, criterion-referenced language test development, and web-based data collection were successfully used in this research design as methods of increasing the reliability and validity of the FLBS-E. Suggestions are given for future use of these and other combinations of interdisciplinary and mixed methods (i.e., quantitative and qualitative) in second language acquisition.