The Nature of Writing

13 Data which are collected from learners involve describing learner‟s interlanguages. One of types of the data is samples of learner language, which is expected as primary data in doing SLA research Ellis Barkhuizen, 2005, p. 21. In order to accomplish SLA goals, the researcher maintains the competence by analysing learner‟s performance. Speaking and writing are considered as natural language activities. Furthermore, there are also three ways of collecting samples of learner language. Clinical elicited sample is one of them Ellis Barkhuizen, 2005, p. 22. It involves a carefully task which is message conv eyance as learners‟ concern but language features as their utility are not neglected. In clinical elicited samples conducted, there are few types which differ from the researcher‟s goal. They are general samples and focused samples. Ellis and Barkhuizen 2005 note that focused samples provide the data which relates to specific language feature p. 34. Moreover, it is nearly similar to experimentally elicited focused samples. Otherwise, clinical elicited samples take up an essential part of SLA because it demonstrates low learners to construct message by using L2. Since samples can be taken in the form of writing, there are some advantages of collecting written samples. It is relatively permanent and easier to collect Ellis Barkhuizen, 2005, p. 28. Furthermore, Mackey and Gass 2012 state that clinical focused samples can be laid out in the form of open-ended elicited data, such as written compositions p. 8. When producing written samples, there are two conditions concerned. First, it is timed or untimed. Second, 14 whether there is access to open reference tools. Furthermore, the examination can produce natural written samples.

4. Error Analysis

Error is a natural phenomenon in learning language Hendrickson, 1981, p. 3. On the other hand, students used to feel that error is a failure and should be avoided Norrish, 1983, p. 1. Norrish also concludes that a fear of making mistakes is one of the suppressing factors in formal learning situation. However, Error A nalysis is a study of learners‟ errors which is helpful for the teacher, as a monitoring tool. It consists of some steps in identifying, describing, and explaining learner language Ellis Barkhuizen, 2005, p. 51. Learners are analysed as native speakers, and then the researcher treats the learner‟s interlanguage as a linguistic competence which describes the detail of interlanguage development. It can be analysed in the form of learners‟ speech and writing. In Error Analysis, the goal is t o determine learners‟ accuracy. Corder 1997 notes that learner errors has its significances in some purposes, which are in pedagogy by telling the teacher what the learner has mastered and not, in research purpose by showing information on how language is being mastered, and in learning by giving the learners the rules of target language as cited in Ellis Barkhuizen, 2005, p. 51. Moreover Error Analysis can give a description of learners‟ difficulty in learning language Norrish, 1983. Hendrickson 1981 adds that based on error produced, teaching techniques or instructional materials are questioned whether it is inadequate, ineffective or otherwise. Errors present PLAGIAT MERUPAKAN TINDAKAN TIDAK TERPUJI 15 feedback about the effectiveness of teacher‟s teaching materials and techniques, and also show the parts of the syllabus which is lack of attention. Therefore, Error Analysis is relevant to teaching and learning language issue. It can be a device to give teacher feedback in order to develop teaching technique and material in a classroom. Furthermore, if errors are generally exposed, Truscott 1996 also points out that error correction, especially in grammar causes dilemma as cited in Weigle, 2002, p. 184. Grammar correction rather than heads to improvement in grammar, it leads to discourage students. Otherwise, not all teachers are capable of explaining the errors. Moreover, error correction will create a diversion from assessing more important aspect in writing. As a result, error correction needs to prioritize error types, so it will be more effective Hendrickson, 1981.

a. The Steps of Conducting Error Analysis

By conducting Error Analysis, defining errors is becoming an essential part. According to Corder 1974 there are the steps to follow as cited in Ellis Barkhuizen, 2005, p. 57: 1 Collection of a Sample of Learner Language This step provides the data for the Error Analysis. The data is collected by the researcher by noticing some factors of producing it. The researcher should decide specific circumstances that will control the learners to result the data. PLAGIAT MERUPAKAN TINDAKAN TIDAK TERPUJI