The Language Testing Research Centre has worked on several projects relating to the Aptis test, the online language proficiency test designed by the British Council.
Investigating the discourse produced at score levels B2.2 to C2 on the Aptis Advanced Writing Test
This project aims to examine the discourse produced by students scoring between B2.2 and C2 on the Aptis Advanced Writing test with the aim of validating the rating scales for the three writing tasks and gaining a deeper understanding of the discourse produced by advanced L2 learners. Detailed discourse analyses drawing on both qualitative and quantitative methodologies will be employed, as well as focus groups with experts in L2 writing and Aptis raters.
Aptis for Teachers
Using interviews and a verbal reporting methodology, we explore the ways test takers respond to the speaking tasks of Aptis for Teachers, a domain-specific variant of Aptis, a general proficiency English language test developed by the British Council. We compare test taker interview and verbal report data with teacher educator perceptions of the oral communication demands faced by teachers in different classroom settings to investigate if the Aptis for Teachers speaking test provides an appropriate measure of the speaking skills relevant to effective communication in Australian classrooms.
Review of the Aptis' tests control item system
In this test, speaking and writing performances are not routinely double-rated and for this reason, raters encounter control items at regular intervals which are designed to ensure that raters are continuously re-standardised and rating to standard. The system therefore has a double function, one of quality control and one of ongoing rater training.
The findings are presented in three areas, namely (a) selection, presentation and retention of control items, (b) standardisation and (c) quality control. The results show that the Aptis team have put in place a number of well-thought out strategies for the selection and presentation of the control items. Possible efficiencies in administration could be gained by automatising the suspension and restandardisation process for the raters, and we have made recommendations along those lines. We also feel that it is important to empirically establish whether the control item system is indeed successful in ensuring continuous rater standardisation and quality control for the Aptis test.