Use this URL to cite or link to this record in EThOS:
Title: Teachers designing low-stakes summative language assessment
Author: Fang, Lin
ISNI:       0000 0004 6056 8685
Awarding Body: University of Bristol
Current Institution: University of Bristol
Date of Award: 2015
Availability of Full Text:
Access from EThOS:
Teacher-made summative tests have been accused of being of poor quality due to teachers not using a systematic test development process, not being properly trained in designing tests and having insufficient resources in the local context. Previous research studies have often recommended that language teachers follow systematic processes widely used in standardized test development. A few models of language assessment literacy (LAL) for teachers have outlined the knowledge and skills needed in assessment practices. Yet, neither systematic processes nor assessment training have been entirely effective in helping teachers to design quality test papers. Systematic processes might not be applicable to the local context and the components of the LAL knowledge base have not conclusively been determined. Few studies explored how teachers design low-stakes summative language tests, with minor consideration given to the influence of the local context. Moreover, the data in previous studies have often suffered from the limitations of the memory recall effect .. To gain a more comprehensive understanding of teachers' test development process, I conducted a case study in a naturalistic setting in China to investigate how 12 language teachers cooperated to design a midterm test paper. There were two main research questions: what do teachers do when designing a midterm test paper and what affects teachers' test development process? I video recorded and observed teachers' development process using software. Data on teachers' cognitive processes were collected through retrospective interviews and focus groups. I also had field notes on the teaching process, curriculum documents, testing materials, daily journals and email correspondence as data. The data revealed that the teachers' test design process transformed. Teachers intended to use the systematic test development process recommended by textbook. However, influenced by contextual elements, gaps in teachers' assessment literacy and the absence of validation plans, the systematic process was accommodated by teachers. Fundamentally, teacher-made summative tests are related to the curriculum and teaching in nature, which distinguishes them from standardized assessment. Systematic processes are not appropriate for teachers designing summative tests. The data also revealed the components in the teachers' LAL knowledge base. Assessment knowledge and skills comprise the core component. This is integrated with pedagogical knowledge, knowledge of second language acquisition (SLA), linguistic knowledge, contextual knowledge and teachers' beliefs/attitudes, forming an evolving context-specific knowledge base. These findings could help researchers to establish a test development process for teachers and aid teacher educators in designing assessment training programmes.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (D.Ed.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available