The Structure of Method Section in Engineering Research Articles

Genre analysis is one of the common ways to analyse the rhetorical structure of a text. Previously, there are many genre analysis and rhetorical organisation studies on research articles (RAs) from different disciplines. Some of them focused on the whole RAs and some of them focused on certain sections of the RAs. This study attempts to identify the moves in the Method section of the Engineering RAs, and the most frequent verbs in the moves. The present study adopts a corpus-based analysis; 60 RAs from an Engineering discipline were retrieved from the Scopus websites. The Method section framework from Huang (2014) was adopted for the analysis. The findings suggest that there are crucial moves in the Method section of the Engineering RAs, which need to be adopted by the writers to enhance not only their discourse competency, but also to meet the readers’ expectation of the text readability.


Introduction
Research articles (RA) is one of the genres in academic writing; it is regarded as a medium to exchange and communicate new knowledge to the academic community members (Flowerdew, 2005;Kanoksilapatham, 2009). It involves the presentation of the research findings, and knowledge sharing among the academic community (Jusoff, Abdullah, & Samah, 2009). However, writing a presentable RA for publication requires not only the awareness of the language use, but also the rhetorical organisation and features of the texts (Amnuai & Wannaruk, 2013). These skills are referred as the discourse competency. Thus, it has been proposed that writers who wish to share their research findings with the academic community successfully should acquire the discourse competency (Bruce, 2009). They also need to be aware of differences in text structures to get their papers to be published in the international journals, and acquire international recognition (Zand-Vakili & Kashani, 2012).
There have been numerous studies on the structure of RAs after Swales ' (1990) introduction of the genre analysis of a text (Fazilatfar & Naseri, 2014). Some of these studies focused on a certain section of the RAs (Alhuqbani, 2013;Peacock, 2011;Swales & Najjar, 1987), and some on the whole structure of the RAs (Amnuai & Wannaruk, 2013;Kanoksilapatham, 2005;Nwogu, 1997;Posteguillo, 1999). decades is considerably low. Moreover, these few researches focused on only a small number of RAs for a discipline (Peacock, 2011).
Nevertheless, Berkenttor and Huckin (1995) assert that the Method section is an interesting section in which the writers need to ensure that the explanations given about the experiments are presented adequately to the readers. Kanoksilapatham (2008) claims that there is no clear model for the Method section in all the RAs of all disciplines, as many researchers may not pay much attention to the section. The model of the Method section is different from the other sections in the RAs; for example, the Introduction section of any disciplines applies the common moves sequence of the Swale's framework (Move 1-2-3), but the Method section requires different patterns of rhetorical structure for different disciplines (Kanoksilapatham, 2003).
For Sociology RAs, Brett (1994) found that the Method section appears to have the main tasks as (1) describe how the data were obtained, (2) explain the concepts and variables in the study, and (3) state the statistical techniques to obtain results. Peacock (2011) conducted a study of the structure of the Method section in RAs across multiple disciplines and discovered that there are variations in the moves across the disciplines, apart from the common ones. Nwogu (1997), in her study of the rhetorical moves of medical RAs, identified three main moves in the Method section: Data Collection -Experiment -Data Analysis. In addition, there are sub-moves (steps) for each of the moves. For example, for the Data Collection move, the sub-moves include data source, data size and data collection criteria. Similarly, the Data Analysis move comprises the procedures for data classification and analysis of the data as the sub-moves.
However, in the engineering RAs, it is crucial to have a detailed investigation of the Method section, as it is an integral section in this discipline (Lim, 2006). A study of the biomedical engineering RAs, which adopted Swales ' (1990) framework, reveals that the most crucial move in the Method section is Detailing methodological procedures in which the writers explain every step taken to obtain the results (Kanoksilapatham, 2005). However, Swales' framework was derived from the linguistics academic writing. Another framework, which is an extension of Swale's, was proposed by Nwogu (1997). However, Nwogu's framework was based on a scientific discipline i.e. medical RAs. Huang (2014), who adopted Nwogu's framework, identified four main moves in the medical RAs Method section (Table 1). It was also found that the writers of the medical RAs rarely include the explanation on the statistical methods used in their studies; they have the assumption that the readers would understand the terms and analysis method used in the Method section. From the same study, Huang also provides a new rhetorical structure for the Method section for the medical RAs (Table 1). Step 1: Announcing objectives Step 2: Specifying protocolised procedures Step 3: Detailing methodological procedures Step 4: Providing background of procedures Step 5: Justifying a procedural decision Step 6: Declaring ethical statements Move 2: Featuring methodological issues Step 1: Describing participants, instruments, materials Step 2: Setting apparatus Step 3: Identifying data sources Move 3: Reporting and consolidating findings Step 1: Announcing findings Step 2: Interpreting findings Step 3: Comparing findings Step 4 Step 1 : Describing type of data collection Step 2 : Describing the source for data collection www.ccsenet.org/ass Asian Social Science Vol. 11, No. 17;2015 Authors Disciplines Moves Proposed Move 2 : Provide inclusion criteria Step 1 : Describing the sample Step 2 : Describing the characteristics

Move 3 : Describe procedures
Step 1 : Taking the measurements Step 2 : Justifying the procedures Move 4 : Present the analysis of the experiment Step 1 : Statistical test techniques Huang's work seems to suggest that an investigation on the rhetorical structure of the Method section of the biomedical engineering RAs using the newly devised framework may promise interesting findings, in addition to Kanoksilapatham's (2005). This paper attempts to carry out an investigation on a biomedical engineering RAs corpus using Huang's framework of the Method section. In other words, this paper intends to identify the structure of the Method section by using the framework based on the scientific discipline instead. Therefore, the objective of this study is to find the most frequent moves in the Method section of the biomedical engineering RAs. In addition, the most frequent verbs in the identified moves are also highlighted. The findings from this study offer an insight into the organisation of the Method section for RAs writing to enhance the writers discourse competency.

Corpus-based Study
Corpus-based analysis is a commonly used tool in investigating the language use, characteristics and writing patterns in a text (Zhang, Pan, & Zhang, 2014;Panah, Yunus, & Embi, 2013). Corpus is a collection of texts, written or spoken stored in a computer application that is available for both qualitative and quantitative analysis (O'Keefe, McCarthy, & Carter, 2007). Studies involving corpus analysis are increasing as the free software and internet-based tools are increasingly available on the internet (Flowerdew, 2008). Recently, many corpus-based studies have been conducted in identifying the macrostructures and linguistic features of RAs (Noorli, 2011). The present study uses a corpus of articles as a sample to investigate the rhetorical structures and linguistic feature of engineering RAs. According to Supatranount (2012), the corpus based method is useful for obtaining a real language sample, thus serving the needs of the target groups.

Corpus
For this study, 60 RAs from the biomedical engineering were selected for the corpus. Three journals of the engineering discipline were selected based on its impact factor (IF) (Kanoksilapatham, 2005). The journals were Biomaterials, Biomechnics and Modelling, and Journal of Biomechanics. All the RAs were retrieved from the Scopus website. Only articles with the IMRD (Introduction -Method -Results -Discussion) structure were selected for the corpus. In order to obtain the most frequent verbs for the Method section, Wordsmith Tool 6.0 was used (Scott, 2012).

Genre Analysis
Studies on the identification of the rhetorical structure of texts have gained more attention from researchers since the introduction of the Swales' (1990) framework. A genre has been defined as "a type of text or discourse designed to achieve a set of communicative purposes" (Swales, 1990). The ESP scholars define genre as a class of communicative events, such as a university lecture or an academic essay (Paltridge, 2001). In the Systemic Functional Linguistics, a genre is more often defined as a kind of text with related form, function and context, such as description, procedure or exposition. Paltridge (2001), and Suharno, Badib, and Sutopo (2011) define genre as the macrostructure and rhetorical function in their study of the introduction section.
It has been a concern to explain how texts are organized in systematic ways to present findings and messages by the writers, especially in academic genres. This is also the main focus for this study, i.e to identify the rhetorical moves for the Method section in the engineering RAs. According to Kanoksilapatham (2005), Swales (1990), and Upton and Connor, (2001), genre analysis refers to an analysis of texts in identifying the regularities of a structure that differentiate one type of text to another. Many previous studies in genre analysis have successfully provided guidelines to enhance writing an impactful research articles (Amnuai & Wannaruk, 2013). Swales ' (1990) framework on the move analysis was intended to address the needs for the NNES writers to enhance their writing. At present, it is the commonly adopted framework to conduct investigations on the structural ognanisation of RAs. In this study too, the move analysis was adopted to identify the moves in the Method section of the biomedical engineering RAs. The analysis was based on the Method section framework proposed by Huang (2014). In Huang's (2014) framework, the Method section in the medical discipline consists of four main moves: (1) Describe study materials (2) Provide inclusion criteria (3) Describe procedures and (4) Present the analysis of the experiments.

Inter-rater Reliability
Inter-rater reliability is used to assess the degree to which different raters or observers make consistent estimates of the same situation. For the inter-rater analysis, three coders from the Language Department were chosen. A coding system of the moves and sub-moves, referred as steps in this study, was first devised; then, a 3-hour training session was conducted to ensure the coders understood how to code the sample texts using the coding scheme given (Kanoksilapatham, 2005). After the training, the coders were given 3 sets of the Method section excerpts to be coded independently. From various categorical variables for the move analysis, Cohen Kappa was chosen to assess the inter-rater reliability of the Method section for the articles. The percentage of agreement among the coders was computed as shown in Table 2. The result shows that the there is a high consistency (98.4%) in the application of the coding system among the coders. This means that the devised coding system is a reliable tool for the tagging of the moves and steps in the study.

Results
After analysing the text and identifying the moves for all the texts, the whole structure of the Method section for the biomedical engineering is described throughout this section. All the four moves for the medical RAs, as proposed in Huang's (2014) framework, did appear in the biomedical engineering RAs. Nevertheless, there are new steps identified in the biomedical engineering RAs.

Results of Moves in the Method section Move 1: Describing Study Materials
Move 1 refers to the description of the materials used for the experiments of the study. This move basically consists of two steps: (1) Describing the type of data collection and (2) Describing the source for data collection. From the corpus, Move 1 appeared in only 16 RAs out of 60, i.e. 27% from the whole corpus. This suggests that though the move does present in some RAs in the corpus, most writers did not include the move. In other words, Move 1 is optional in the RAs writing of this discipline. The following sample sentence (a) is an example of Step 1 to describe the type of data collected and selected for the study; while (b) is the source for the data collection, i.e.
a) Brain tissue from post mortem human subjects (PMHS) was used in this study.
b) BALB/c, DBA/1J and C57/BL6 mice were purchased from SLC Japan and maintained under specific pathogen-free conditions.

Move 2: Providing Inclusion Criteria
Move 2 describes the samples and the characteristics required for the samples to be selected as a part of the experiments. This Move comprises two steps: (1) Describing the preparation of the sample and (2) Describing the characteristics required for the sample to be included in the experiment. This Move occurred in 52 RAs (87%) out of the whole corpus. Sample sentences (c) and (d) demonstrate the steps respectively.
However, an additional step seems to be prominent in this corpus; there is a preliminary action on preparing the sample for the experiments. It appeared in all the 60 RAs (100%) in the corpus. This step is not included in Huang's (2014) model. Sentence (e) is the sample of the step.
www.ccsenet.org/ass Asian Social Science Vol. 11, No. 17;2015 c) During specimen preparation, the tissue was kept frozen, and between two preparation steps, it was stored in a double plastic layer……… d) Subjects were excluded if they had a history of neck injury or pain, or known muscle, nerve, or balance problems. e) Cells were expanded using standard cell culture methods in a humidified atmosphere containing 5% CO2 at 37 C.

Move 3: Describing Procedures
This move describes the procedures taken in order to obtain the findings; it appeared in all the RAs (100%) of the corpus. Thus, this signifies the importance of this move in the section. The finding corresponds with Kanosilapatham's (2005), who adopted the Swales' framework, i.e. Detailing methodological procedure. This move was also revealed as the most crucial move in his study. There are three steps for this move: (1) Taking the measurements and (2) Justifying procedures. It is also found that in Justifying the procedures, the most frequent words used in this step is because and to explaining why the measures and steps are taken. (f) and (g) are the examples for each step respectively.
f) … Equilibrium stresses at 60% were determined by extrapolation of an exponential fit through the equilibrium stress strain data using a least squares optimization function (R2 = 0.99) g) … Local maxima were used to calculate the largest non-overlapping circles, with a minimum possible diameter of D2… It is also apparent that some of the writers cited previous studies in this step. The step was identified in 52 RAs of the corpus (87%). This step marks the distinguishing feature of the biomedical engineering RAs. (h) provides the sample of referring to previous studies.
h) ……. the commonly used approach in which the stress free reference configuration B0 remains fixed throughout the growth simulation

Move 4: Present Analysis of the Experiments
Move 4 appeared in almost every RA of the corpus -52 RAs out of 60 or 87%. This move serves as a description of the statistical tests techniques and calculation for the data. In this move also, the software used and the justifications for their uses are provided. Move 4 is noted to be appearing in the last part of the Method section. i) Statistical differences among the experimental groups were evaluated by one way ANNOVA and Tukey's multiple comparison test.
j) Statistical tests were performed using SPSS IBM SPSS Statistics 19. Table 3 summarizes the structure of the Method section identified from the corpus, with the addition of the two new steps: Describing the preliminary action and Referring to previous studies. Step 1 : Describing the types of data Step 2 : Describing source of the data Move 2 : Provide inclusion criteria Step 1 : Describing the sample Step 2 : Describing the characteristics Step 3 : Describing the preliminary actions Move 3 : Describe procedures Step 1 : Measurement taken Step 2 : Justifying the procedures Step 3 : Referring to previous studies Move 4 : Present the analysis of the experiment Step 1 : Statistical test techniques

Frequent Verbs in the Moves of the Method Section
The basis of a language is its lexis (Noorli, 2011). Lewis (1993) argues that meanings are constructed from fixed words rather than fixed structures. Therefore, grammatical expressions are closely linked with the nature of the rhetorical structure of a text. The fixed grammatical expressions provide better understanding of the rhetorical intent of a structure; they reflect the rhetorical purpose of a structure (Bloch, 2010). Hence, this paper also identifies the most frequent verbs in the moves of the engineering RAs Method section. Table 4 shows the top 10 most frequent verbs in the Method section. Kanoksilapatham (2005Kanoksilapatham ( , 2012 asserts that the procedure in the Method section is described using past tense verbs (were obtained from, was added to, were removed). Though the list does not reveal the tenses of the verbs, the purpose of this list is to present the most frequently used verbs in the corpus regardless their forms and tenses. In other words, the verbs were selected based on their prototypical category. Table 5 displays the three most frequent verbs in each of the moves in the Method Section. The identified verbs reflect the rhetorical purposes they serve, for example, purchased, obtained and used to show describing the materials of the experiments or study. Further investigations on the linguistic features of the verbs can be conducted by observing the word cluster of each verb in the corpus ( Table  6). The word cluster provides the most frequent structure the verbs assume in the corpus. From the 3-word clusters, it can be observed that most of the structures are in past tenses: was used to, were used to, was assumed to, were performed in, etc., and of passive constructions: assumed to be, was determined by, was calculated by.  Vol. 11, No. 17;2015

Discussion
Previously, Kanoksilapatham (2012) investigated the moves in the Method section of the biomedical engineering RAs by adopting Swales's moves, devised from the linguistics discipline. This study, provides a different perspective to the move analysis of the Method section of the biomedical engineering RAs by adopting Huang's (2014) model, which is derived from a technical/medical discipline.
The findings from the study reveals that though there are common moves in both the medical and biomedical engineering RAs, two new steps are identified in Move 2 (Step 3 -Describing the preliminary actions) and Move 3 (Step 3 -Referring to previous studies) of the biomedical engineering corpus respectively. It is found that one of the new steps identified, Step 3 -Describing the preliminary actions, forms an obligatory submove in Move 2. Another obligatory step for the biomedical engineering is Move 3 (Describing procedures), which is similar to the medical discipline, as well as Kanosilapatham's study. It is an important step in which the writers describe the procedures taken to allow replications of the procedure for future studies. These two newly identified obligatory steps distinguish the RAs of this discipline. Nevertheless, the finding also reveals that Move 1 is an optional move in the biomedical engineering RAs writing. Thus, the model is summarised in Table 3.
In addition, the analysis on the most frequent verbs in the Method section of the corpus reveals that the structures are mostly in past tenses and of passive construction. The 3-word clusters of the verbs provide an insight into the most common fixed words in the Method section of the corpus. The fixed words reflect the rhetorical intent of the structure. These findings, hence, highlight the potential of empirical investigations into the structure and language use of RAs, especially of different disciplines, even sub-disciplines, for the purpose of discovering and enhancing discourse competency of the RAs writers (Panah, Yunus, & Embi, 2013).

Conclusion
The RA writing consists of scientific communication elements, which depend heavily on the ability of the writer to use the appropriate structure and language in the fields. The main goal of writing an RA is to help the readers understand the concept and ideas that the writer intends to propose. Therefore, appropriate writing skills need to be acquired by writers to ensure the new knowledge that has been explored can be shared with the readers.
This study demonstrates that the rhetorical structure of an RA of a discipline can be analysed and understood from different frameworks. The findings provide another possible way of approaching the writing up of a biomedical engineering RA. In addition, the analysis of the verbs in the moves promises interesting insights into the most frequent fixed words to express the rhetorical intentions. The moves and fixed expressions in the moves identified in this empirical study can be highlighted in training NNES writers to prepare for their RA submission to prestigious journals.