Print Page   |   Contact Us   |   Sign In   |   Join or Create a Guest Account
2017 Conf. - Operation. an expanded construct of writing proficiency for an occupation-specific test
Share |

The importance of input from occupational experts in defining the language proficiency construct and developing valid criteria to assess performance on ESP tests is widely acknowledged. However the process of converting the insights gathered from occupational experts into a workable test rating scale has received scant attention.

The paper reports on this process with specific reference to the writing sub-test of the Occupational English Test (OET) for overseas-trained health professionals. The OET writing task requires candidates, in the role of health professional, to draw on a set of written prompts to write a letter of referral to a colleague on an aspect of patient care.

With the aim of expanding the test construct as embodied in the existing, linguistically-based, assessment criteria to better reflect health professional values, nurses and doctors from rural and metropolitan hospitals were asked to judge the adequacy of patient records obtained from health settings, specifically documents involving the handover of the patient’s care to another health professional. Based on their feedback, an elaborate checklist was developed to reflect the many qualities of the written documents that they considered critical for effective patient care. Because some qualities related to clinical rather than language competence, the checklist was narrowed and collapsed into a smaller set of occupationally-relevant communication categories that could be linked to the more limited domain of the language test. A group of experienced OET raters was then convened to examine a set of OET writing scripts at various levels of proficiency and develop descriptors for each communication category representing different levels of performance.

The paper discusses the challenges involved in the process including the inevitable construct shrinkage evident in the final version of the scale due to the constraints of the testing situation. Implications are drawn for LSP assessment more generally.

Membership Software Powered by YourMembership  ::  Legal