Testing the factor structure of the beliefs about language learning inventory

Area of Science:

• Educational Psychology
• Applied Linguistics
• Psychometrics

Background:

• The Beliefs About Language Learning Inventory (BALLI) is a key instrument for understanding student perspectives.
• Horwitz's five-factor model provides a theoretical framework for these beliefs.
• Understanding student beliefs is crucial for effective language pedagogy.

Purpose of the Study:

• To examine the factor structure of the BALLI in a specific learner population.
• To identify a potentially more robust and empirically supported model of language learning beliefs.
• To assess the theoretical and empirical validity of alternative factor structures.

Main Methods:

• Confirmatory Factor Analysis (CFA) was used to test Horwitz's five-factor model.
• Exploratory Factor Analysis (EFA) and subsequent CFA were employed to evaluate alternative models.
• The study involved two independent samples of college students learning English as a foreign language in Taiwan (N ≈ 750 each).

Main Results:

• Confirmatory factor analysis provided only partial support for the theoretical five-factor model.
• A revised four-factor model, comprising 12 items, demonstrated superior theoretical and empirical performance.
• The four factors identified were: Difficulty of Language Learning, Foreign Language Aptitude, Importance of Spoken Language, and Analytical Approaches to Language Learning.
• Replication of the four-factor model in an independent sample was successful, but factor reliability was limited.

Conclusions:

• The findings suggest that a four-factor structure may better represent language learning beliefs in this context.
• While empirically supported, the identified factors demonstrated insufficient reliability, indicating limitations in the current item set.
• Further research is needed to develop a more representative inventory for accurately measuring specific aspects of language learning beliefs.