Open AccessOn the Effect of Perceived Test Impact on Test Anxiety and Attainment of PhD Candidates: Studying English Proficiency Test in Islamic Azad University
Author(s) -
Hamid Ashraf,
Mahmood Khosravani
Publication year - 2021
Publication title -
international journal of linguistics, literature and translation
Language(s) - English
Resource type - Journals
eISSN - 2708-0099
pISSN - 2617-0299
DOI - 10.32996/ijllt.2021.4.1.2
Subject(s) - test anxiety , test (biology) , likert scale , psychology , anxiety , context (archaeology) , graduation (instrument) , perception , scale (ratio) , social psychology , clinical psychology , applied psychology , developmental psychology , mathematics , paleontology , physics , geometry , quantum mechanics , neuroscience , psychiatry , biology
Research findings has shown that high stakes test bring about unfortunate consequences inside and beyond educational context which are known as washback and impact respectively. English proficiency test (EPT) of Islamic Azad University (IAU) in Iran, as an obligatory part of the PhD graduation program, is not an exception. Thus, considering possible adverse consequences of the test, the current research was an attempt to investigate the effect of perceived test impacts on test anxiety and attainment of the examinees. To this aim, a mixed method design was employed in which quantitative and qualitative data were collected from 81 PhD candidates in different disciplines of the humanities. In the first phase, adopting a framework for test impact perception, a structured interview was conducted on the basis of which participants were assigned to NPTI (non-perceived test impact), MPTI (moderately perceived test impact) and SPTI (severely perceived test impact). Then performance of these was compared in an adopted test anxiety scale (five-point Likert scale) and EPT of IAU. Kruskal-Wallis and one-way ANOVA were used to analyze the data. The findings on test anxiety scale indicated a significant difference between groups (χ2 (2) = 52.618, p = 0.000.). However, the group means were not significantly different in EPT (F (2, 77) = 1.975, p = 0.146). Findings obtained from test anxiety supported previous studies carried out in the literature while those from EPT stood in sharp contrast. An experimental design is required to control other variables that interact with these findings. Yet, these findings have implications for applied linguistics, researchers and educational decision makers.