Hathaichanok Anghirun.. A comparative study of local errors in scientific abstracts of predatory and professional journals. Master's Degree(Applied Linguistics). Mahidol University. Mahidol University Library and Knowledge Center. : Mahidol University, 2016.
A comparative study of local errors in scientific abstracts of predatory and professional journals
Abstract:
The objective of the present study was to investigate local errors in research abstracts published in predatory and professional journals during 2013 to 2015. Two corpora were used in the study: the corpus of fifty predatory journal abstracts (8,649 running words) and fifty professional journal abstracts (8,601 running words). The predatory journal, Indian Journal of Scientific Research, was randomly selected from Beall's list of standalone journals (2015), selection criteria included: 1) be in the field of science, 2) have been published at least three years, and 3) be online accessible. For the professional journal, Science Journal was chosen as the second source of data from the Journal Citation Report (2014) with two criteria:1) be in the field of science and 2) had the highest Eienfactor score. In regards to sample abstract selection, stratified random sampling and simple random sampling techniques were employed. Errors were identified and classified based on Hendrickson's (1979) four local error categories including lexis, syntax, morphology, and orthography. Three research instruments were used to validate the study: spreadsheet programs, error analysis forms, and two grammar books as the coding sources of references. To assure the reliability of the analysis, two inter-raters and the researcher identified errors in sample abstracts independently. The value of inter-rater reliability was relatively high (K = 0.77 and 0.67). Three hundred and thirty errors were found from the sample abstracts of the two journals. Among these two sources, 309 errors (or 94% of the total local errors) occurred in the predatory journal, while only 21 errors (6%) were found from the professional journal. Moreover, the findings showed that orthographical errors dominantly occurred in the sample abstracts of the two journals, while lexical errors occurred the least. In addition, the researcher discovered the occurrence of three new error types (no space between words, the use of Latin abbreviations in text, and omission of an infinitive -to) in the predatory corpus (n = 30, or 10%). These results serve as empirical evidence asserting that occurrence of language errors is the significant characteristic of predatory journals.