TY - JOUR AU - Ua-umakul, Anon AU - Vittayapirak, Jirapa PY - 2016/08/31 Y2 - 2024/03/28 TI - CORPUS-BASED ERROR ANALYSIS OF THAI STUDENTS’ LABORATORY SCIENTIFIC ABSTRACT WRITING IN ENGLISH JF - Journal of Industrial Education JA - J. IND. EDUC. VL - 15 IS - 2 SE - Research Articles DO - UR - https://ph01.tci-thaijo.org/index.php/JIE/article/view/122716 SP - 183-190 AB - <p>The aims of this study were to analyze and compare the main causes of common errors of Thai students using the corpus-based approach. The samples were two groups of upper secondary school students (Grade 11) at Watsuthiwararam School, Bangkok who enrolled Physics Laboratory Technique course in 2014. They were divided into two groups by simple random sampling: (1) the experimental group and (2) the control group. The instruments were (1) a corpus of 540 Laboratory Scientific Abstracts (LSA) written by the participants, (2) an error classification recording scheme, (3) lesson plans for teaching scientific abstract writing in English, and (4) a guided interview for investigating the causes of errors. After the corpus was compiled, the errors and causes of errors were categorized. The frequency and percentage of errors were counted and calculated using Markin 4.2.2 and AntConc 3.2.4 software. Chi–square test (c<sup>2</sup>-test) was used to compare the errors committed by the two groups of students. The results were as follows: (1) The top three morphological errors were the omission of article “the” (9.62%), the omission of article “a” or “an” (8.35%), and the subject-verb agreement (4.65%), respectively. (2) The top three syntactic errors were the misuse of active voice (5.79%), the misplacement of verb (4.51%), and the incorrect form of parallel structure (3.87%). (3) The top three mechanical errors were the use of commas (9.85%), periods (5.04%), and capitalization of words (2.75%). (4) For lexical errors, misspelling (7.82%) was found more than homonym words (0.15%). (5) The top three semantic errors included specific term usage (0.98%), wrong collocation usage (0.97%), and the more general term usage (0.71%). The results indicated that there was statistical difference of errors between the two groups at .01 level of significance. &nbsp;The main causes of error were the L1 interference (33.33%), and the misconception in language rules (31.93%).</p> ER -