Validating aspects of a model of academic reading

A thesis submitted for the degree of Doctor of Philosophy of the University of BedfordshireIn the past, the focus in language testing, teaching and research has largely been on careful reading while expeditious (quick, efficient and selective) reading has been largely ignored. However, some research suggests that careful reading ability alone is inadequate for students to meet the demands of undergraduate academic reading. In the main English for Academic Purposes (EAP), test instruments have been previously based on careful reading models which assume reading to be unicomponential. If this is not the case, the issue for language testing is whether the construct of academic reading can be validly measured by a focus on careful reading alone. The aims of this study were to investigate the types of academic reading required of firstyear undergraduates based on Urquhart and Weir's (1998) four-cell matrix of reading types which also forms an important part of Khalifa and Weir's (2009) reading model. Based on this, a valid academic reading test battery for undergraduate students was developed and used to examine the divisibility of the academic reading construct. The literature review on reading models suggested that current models were nearly all premised on careful reading and expeditious reading had in the main been ignored. The findings of a pilot and main questionnaire survey with undergraduates suggested that both careful and expeditious reading were important in accomplishing academic reading tasks at the undergraduate level. Accordingly, the empirical data generated by these surveys validated Urquhart and Weir's (2009) reading matrix and aspects of the reading model by Khalifa and Weir (2009). Based on this matrix and aspects of the model, a valid reading test was developed and administered to first-year undergraduate students. The performance of undergraduates across the different parts of the reading test confirmed that academic reading was a divisible construct. The findings of this study add to the literature on EAL academic reading by lending empirical support to a componential approach to the teaching and testing of reading. The componential model and the test design methodology employed should help test designers develop valid academic reading tests embracing both careful and expeditious reading types. The results from such tests might usefully inform pedagogical practice leading to more efficient reading practice at undergraduate level