Research Instruments A study on the textual features of the narrative texts written by the students of SMU N 2 Klaten.

\ selected the sample. As this study investigated the students’ narrative texts, the English teachers said that class X B was the most appropriate class to be the sample of this study as 1 the class had average-achievement in English, 2 the class was just learning how to write a narrative text and 3 the class was accessible. It was the perfect timing for the students in class X B to produce narrative texts. Thus, class X B was selected as the sample of this study. There were 40 students in class X B. There were 26 female students and 14 male students. The students of class B had been taught about narrative text for about 4 times in 4 meetings. In teaching the students, the teacher used a handbook and student work book. The teacher also had explicitly taught narrative text’s textual features.

C. Research Instruments

The research instruments were the students’ narrative texts, questionnaires, and interview. The first instrument was the student’s narrative text. The students’ narrative texts were collected from the English teacher after the students of class B were given homework to write a 150-word length narrative text by the teacher. The students had their freedom to choose the topic of their text and to use the dictionary. The students’ narrative texts were the documents of this research as this research uses document analysis. Referring to Ary et al. 2002: 442, document analysis was applied to analyze types of errors, which was, in this research, related to the students’ errors in applying textual features, which were the generic structure and linguistic features of a narrative text. 27 PLAGIAT MERUPAKAN TINDAKAN TIDAK TERPUJI \ The second instrument was the questionnaire. There were two kinds of questionnaires, one of which was for the students and the other was for the teachers. The writer distributed the questionnaires to the teachers and the students to investigate the possible causes of the most frequent errors occurred in the students’ textual features of narrative texts. The writer considered these several important things suggested by Wiersma 1995: 180 – 181 in constructing questionnaire:  Having the items directly relate to the research problem  The items are to be clear and unambiguous  Including only one concept in a single item  Avoiding the use of leading questions  Not asking questions so that certain responses make respondents disapprove of themselves  Avoiding questions demanding personal information including specific income information and age of the respondents  Requesting only information that the respondent is able to provide  Making the reading level of the items appropriate for the respondents  Shorter items are preferred to longer items, and simpler items are preferred to complex items  When requesting quantitative information, ask for specific number rather than average  For some items, it is necessary to provide a neutral response to avoid forcing the respondent to make an undesirable response 28 PLAGIAT MERUPAKAN TINDAKAN TIDAK TERPUJI \  Avoiding negative items and not using double negatives. The questionnaire was in the form of a checklist and an essay. There were 25 closed questions and an open question in the questionnaire for teachers. Meanwhile, there were 26 closed questions and an open question in the questionnaire for students. The third instrument was interview. The writer interviewed 10 students having erroneous textual features of narrative texts. Those students having the most errors were interviewed to verify and substantiate the result of the questionnaire. Thus, the possible causes of making errors in the textual features were revealed.

D. Data Gathering Technique