2013 IJRSET – Volume 2 Issue 1
Author/s:
Tavakoli, Mansoor*
University of Isfahan, Iran (Mr.tavakoli14@gmail.com)
Farshid, Majid
University of Isfahan, Iran (Farshidmajid@yahoo.com)
Rezazadeh, Mohsen
University of Isfahan, Iran (M1rezazadeh@gmail.com)
Abstract:
This study attempted to explore the notion of construct validity of Graduate Record Exam (GRE) tests (especially when it is used in some Iranian universities as a high-stakes test) taking a qualitative perspective. High-stakes assessment is any assessment whose outcome has life-changing implications for the test taker (Rover, 2001). To do so, five university instructors of English, all having post-graduate degree, were selected to take a GRE test. From different parts of the GRE test the analogy and the quantitative as well as the vocabulary parts of the GRE test were chosen for the experiment. It is because these parts, it seems, render GRE tests inappropriate in a situation like that of Iran. The data were then gathered via think-aloud technique, an introspection method. The study revealed some insights about the construct validity of GRT tests from the test takers’ perspectives. The subjects unanimously reported that the mathematics questions in the quantitative part had little or nothing to do with the notion of English proficiency. In other words, they believed that GRE tests were not a good indicator of one’s English knowledge. The participants also reported they had problems answering the analogy part of the test for a good number of reasons. This was partly had to do with the fact that the analogy questions needed more time to be solved and partly with the fact that they were more of intelligence questions than English proficiency items. In the end, it was concluded that GRE tests suffer from both construct under-representation and construct-irrelevant problems. Consequently, GRE tests should not be used as a high-stakes test in Iranian universities.
Keywords: construct validity; graduate record exam; introspection techniques; qualitative paradigm; validation; think-aloud technique
DOI: https://doi.org/10.5861/ijrset.2013.351
*Corresponding Author