Investigating the construct of academic writing in TOEFL-iBT: implications for the explanation inference in the validity argument

    Research output: Chapter in Book/Report/Conference proceedingChapterpeer-review

    Abstract

    The qualitative study reported in this chapter investigated the construct of academic writing as it relates to the explanation inference in the Test of English as Foreign Language (TOEFL) interpretive argument. Based on an interactionalist model of construct definition, data were collected through face-to-face interview sessions with 19 postgraduate students, who reported on their use of metacognitive strategies. During the interview the participants were prompted to reflect on the strategies they used when completing the writing component of the Test of English as Foreign Language-Internet-based Test (TOEFL-iBT). In order to examine the participant’s consistency in using metacognitive strategies across writing tasks in test vs. real-life academic situations, data were also collected from the same participants after the completion of one of their academic assignments. The data showed that time constraints restricted participants’ full employment of metacognitive strategies in test situations. Results are presented and discussed, and the implications for the explanation inference in the validity argument for TOEFL are considered.
    Original languageEnglish
    Title of host publicationStudies and essays on the learning, teaching, and assessing L2 writing in honour of Alister Cumming
    EditorsA. Mehdi Riazi, Ling Shi, Khaled Barkaoui
    Place of PublicationLondon
    PublisherCambridge Scholars Publishing
    Chapter12
    Pages309-335
    Number of pages27
    ISBN (Electronic)9781527549517
    ISBN (Print)1527549518, 9781527548145
    Publication statusPublished - 2020

    Fingerprint Dive into the research topics of 'Investigating the construct of academic writing in TOEFL-iBT: implications for the explanation inference in the validity argument'. Together they form a unique fingerprint.

    Cite this