Kasin Janjaroongpak
Email: kasin@tni.ac.th
Theodore Maria School of Arts, Assumption University, Thailand
ISSN 1905-7725 NET 15.1 January 2021 14
Received: 2020-04-06 Revised: 2020-07-13 Accepted: 2020-09-08
Abstract
With relative variations, formal paper-and-pencil tests in Thailand, especially nation-
wide entrance examinations–O-NET, CU–TEP, TU–GET, NIDA TEAP, NT– were generally
concerned with 4 test types: vocabulary, grammar, indirect conversation, and reading. While a
large proportion of both students and teachers has been grappling with test specifications–how
to beat the test–, to the best of my knowledge, little attention has been paid to examine the
extent of relations amongst the four pillars. (Whittome, 2019) This paper attempted to shed
light on the relations among test sections in a formal examination whether there was an
indicative section to the total score. This exploratory study was conducted with the test results
of more than 700 second-year students in a course that every student needed to pass. The results
were exported and processed with simple regression statistics tools of Microsoft Excel by
testing the relative strength between two test sections with the total score. The preliminary
analysis suggested that the section designed with TOEIC as a model was a bellwether,
statistically significant at the 0.05 level, for the overall achievement compared with the other
section. A brief expository account was provided as well as possible pedagogical implications
for developing an achievement examination.
Keywords:
TOEIC, simple regression, test, quantitative research, ESP
