فهیمه معرفت - Professor, Allameh Tabataba’i University, Department of English Language and Literature, Faculty of Persian Literature and Foreign Languages, Tehran, Iran
مجتبی حیدری - Ph.D. Candidate, Allameh Tabataba’i University. Department of English Language and Literature, Faculty of Persian Literature and Foreign Languages, Tehran, Iran

With the growing access to new types of reference tools, today’s L۲ writers have a plethora of choices when completing an academic writing assignment. Such resources are absent in most high-stakes academic writing exams, making the two situations dissimilar. Aimed to compare the performances of ESL writers in Exam and Non-exam (real-life) academic writing situations, the present study recruited seven ESL university students who had previously taken an IELTS test. The students completed two analogous writing tasks: an exam-setting and a Non-exam writing test which aimed to simulate the real-life setting. Coh-Metrix analysis of the linguistic features of syntactic complexity, lexical sophistication, and text cohesion of the writings suggested that the students improved the textual quality of their writings in real-life academic writing situation. In addition, FACETS analysis of the quality of the writings, as assessed by the human raters, showed that the students did not benefit equally from the merits of the real-life settings compared to the Exam settings. The findings suggest that the students spent different amounts of time and used different types of queries to consult with external resources. Students’ background training and writing strategies can highly affect their performance in real-life academic writing compared to the writing exams, warning against the validity of such tests.

academic writing, Coh-Metrix, Exam/ Non-exam Settings, FACETS, writing exams validity