An Analytical Reverse Engineering of IELTS Listening Tasks for a Construct Model

Abstract

The study reported here was concerned with the issue of reverse engineering of language test items as it relates to the identification of the language constructs underlying listening tasks of LELTS test. In this regard, the IELTS examination papers, from IELTS 1 to IELTS 10 were compiled as a corpus for the analysis. Tasks were analyzed using a taxonomic frame work adopted from Moore, Morton and price (2012), that was originally adapted from Weir and Urquhart (1998), with a focus on two dimensions of difference: level of engagement, referring to the level of text with which a listener required to engage in order to respond to a task (local vs. global); type of engagement referring to the way (or ways)  listeners expected to engage with a text in order to process the material to respond to a task (literal vs. interpretative). Overall, the analysis found evidences of bottom up processing underlying most IELTS listening tasks. The majority of tasks were identified to have a ‘local-literal’ configuration on their orientation, demanding primarily a basic understanding of relatively small textual units of the material. The results of the study were used to suggest the practical implications for the four groups of the people involved in the IELTS educational contexts: participants; teachers; material preparation experts, and curriculum designers.