University of Leicester
Browse
- No file added yet -

Assessing Secondary School EFL Teachers’ Language Assessment Literacy: The LAL Test Development

Download (5.84 MB)
thesis
posted on 2024-04-22, 15:15 authored by Yohana Gratiana

Although language assessment literacy (LAL) is fundamental to language testing and assessment, we know little about the extent to which teachers are well-prepared for assessment-related activities in the language classroom. In that regard, LAL measurements have the potential to be a powerful tool to get knowledge of teachers’ LAL. The study attempts to develop and validate a measurement tool for EFL teachers’ LAL in Indonesia. The theoretical definition of teachers’ LAL for the test development in the study consisted of three major constructs: (1) understanding the English language skills, (2) understanding language testing and assessment concepts, (3) understanding classroom-based assessment. The LAL test was administered to 130 EFL teachers in Jakarta in two stages: pilot and main studies. The item responses were analysed using the Classical Test Theory (CTT) perspective in the pilot study, and the Cronbach Alpha coefficients were computed to measure the test reliabilities. The pretesting results indicated good test reliabilities and psychometric qualities. The content validity indices obtained for the item-construct alignment were also satisfactory. In the main study, I investigated the structural validity of the test using the first order Confirmatory Factor Analysis (CFA). The findings in the main study showed considerably lower test reliabilities than in the pilot study. The factor loadings of the items were mainly very low across all the constructs, suggesting a poor internal structure of the LAL test based on the sample data. The most probable explanation for the items’ poor performance in the main study was that the respondents taking the tasks were largely ignorant of language assessment, leading to random responses to the items. The findings suggested that the LAL test may represent a sound model of LAL. However, it could not be justified based on a sample drawn from the population, which had very little, or no knowledge about language assessment.

History

Supervisor(s)

Glenn Fulcher; Jim Askham

Date of award

2024-03-22

Author affiliation

School of Education

Awarding institution

University of Leicester

Qualification level

  • Doctoral

Qualification name

  • PhD

Language

en

Usage metrics

    University of Leicester Theses

    Categories

    No categories selected

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC