Improving Student Writing with Paper-Based and Computer-Based Text Analysis

Abstract

In this study, mixed proficiency level classes of science and engineering university students performed both paper-based and computer-based text analysis using corpora to improve specific language forms such as noun, verb, and prepositional phrases, statements of intent, and results reporting in the production of a 2000-word academic research paper. Preand post-language samples for noun phrases and student feedback were analyzed to determine the effectiveness of this approach. Results were grouped into advanced and intermediate level students. Despite 5% and 8% overall gains respectively, more individual gains were seen with the advanced level students. All student responses to questionnaires regarding the usefulness of corpora and the various corpus tasks were positive. While advanced level students seemed to benefit the most from the corpus tasks, 90% of students reported they would use corpus analysis in the future, and 83% felt their writing had improved; thus, intermediate level students also benefited. Corpus-Based Exercises in the L2 Classroom Corpus-based text analysis has been shown to benefit L2 students in many ways. The learner controls the learning process (Braun, 2005; Huang, 2008), inductive thinking is encouraged (Johns, 1991), and there is a virtually limitless supply of data (Conrad, 2000). Since corpusbased text analysis is text-oriented and uses lexical patterns, it naturally lends itself to reading and writing (Flowerdew, 2002). Because grammar and vocabulary are interrelated (Sinclair, 1991), it is possible to clearly see common patterns and frequency of language use (Biber & Conrad, 2001). Yoon and Hirvela (2004) report that corpus analysis is increasing for English for Specific Purposes (ESP) courses in particular, since authentic texts provide specialized word patterns. Learners are able to see technical words in context, commonly occurring phrases and language chunks, and, as Yoon and Hirvela (2004, citing Odlin, 2001) noted, “where to put words into sentences.” In addition, various studies using corpus analysis in the L2 classroom report on particular language objectives. A sampling includes academic English vocabulary (Thurston & Candlin, 1 Language Education in Asia, 2012, 3(1), 60-70. http://dx.doi.org/10.5746/LEiA/12/V3/I1/A06/Oghigian_Chujo Language Education in Asia, Volume 3, Issue 1, 2012 Oghigian and Chujo Page 61 1998), the overuse of logical connectors (Milton & Tsang, 1993), basic grammatical structures such as noun and verb phrases (Chujo & Oghigian, 2008) and ESL university-level writing (Yoon & Hirvela, 2004). With the exception of Chujo and Oghigian, these studies have been conducted with intermediate or advanced level learners. In fact, there are very few studies at the beginner level (Boulton, 2008) or studies that incorporate data driven learning (DDL) in a class comprised of a range of levels. (For an excellent literature review focused on writing and student attitudes toward corpus use, see Yoon & Hirvela, 2004.) The purpose of this study is to investigate the use of text analysis as an aid to EFL technical science and engineering research paper writing in a class of mixed proficiency level university students. The approach is task-based and employs paper-based and computer-based concordancing as well as text analyses of sample journal articles. To determine if the corpus activities had an impact on learning, an analysis of noun phrases was done on preand postcourse writing samples. Feedback from students on end-of-term questionnaires was also collected and analyzed. Case Study Technical Writing Technical Writing 1 (TW1) and Technical Writing 2 (TW2) are one-semester elective English classes in a science and engineering university faculty. The goal of TW1 is for students to produce a 2,000-word research paper on a topic related to their fields, which is IEEE (Institute of Electrical and Electronics Engineers) -cited and referenced. The goal of TW2 is to write a research paper based on primary-sourced data on a topic in their fields that is formatted to a relevant journal identified by each student. These are the first writing-focused courses offered in the English program, although students in the first year write lecture summaries, and second year students do a collaborative written research project. TW1 is not a prerequisite for TW2, but it is highly recommended. Participants Twenty-four students enrolled in TW2 in the fall of 2011 participated in this study. In addition to varied test scores (self-reported scores of 375-975 on the Test of English for International Communication [TOEIC]), students’ listening, speaking, reading, and writing abilities also greatly varied, based on teacher observation and homework assessments. All were thirdor fourth-year undergraduate students. Student majors included applied mathematics, civil and environmental engineering, medical bioscience, electrical engineering, chemistry, and applied physics. Weekly classes met for 90 minutes for 15 weeks, which comprised one semester. Ten classes were held in a regular classroom and five classes were held in a computer classroom, as dictated by each weekly objective. Corpora and Corpus Tools In a computer room in Week 2, students were shown how to access and use three online corpora: the Corpus of Contemporary American English (COCA) (http://corpus.byu.edu/coca/), Springer Exemplar (http://www.springerexemplar.com/), and the Professional English Research Consortium (PERC) corpus (http://scn.jkn21.com/~percinfo/). All three are both corpora and corpus tools. Although COCA is slightly more complex to use, sample concordances are easier to understand for lower proficiency level students. It is also possible to choose only academic sources for concordance lines. The user interfaces for Exemplar and PERC are very simple; however, the corpora used are taken from journals and professional books and are therefore at an advanced level.