A DESCRIPTIVE STUDY ON THE QUALITY OF ENGLISH FINAL TEST AT THE FIRST SEMESTER OF 12th GRADE STUDENTS OF SMAN 1 KEDUNGWARU IN ACADEMIC YEAR 2014/2015

shofiyah, atik laelatus (2015) A DESCRIPTIVE STUDY ON THE QUALITY OF ENGLISH FINAL TEST AT THE FIRST SEMESTER OF 12th GRADE STUDENTS OF SMAN 1 KEDUNGWARU IN ACADEMIC YEAR 2014/2015. [ Skripsi ]

Text ATIK LAELATUS SHOFIYAH.pdf Download (1MB)

Abstract

Abstract:One of the essential parts of teaching and learning process is evaluation because conducting an evaluation can give any information about the students, and also the effectiveness of teaching and learning process itself; and the information taken, later, can be used to the improvement of teaching and learning program. One of the instruments in doing evaluation in teaching and learning program is a test. The result of the test will represent the students’ language proficiency of learning language, thus it is necessary to create a good test. A test is considered to be good if it fulfill the characteristics of a good test; validity, reliability, difficulty level, discrimination power, and distractor efficiency if the test is in the form of multiple-choice test. The formulation of the research problem was how is the quality of English final test of the 12th grade students at the first semester made by SMAN 1 Kedungwaru in term of its validity, reliability, level of difficulty, discrimination power, and distractor efficiency? The purpose of this study was to present the quality of the English final test of the 12th grade students made by SMAN 1 Kedungwaru in term of its validity, reliability, difficulty level, discrimination power, distractors efficiency. The research method applied in this research were: 1) the research design in this study was descriptive with quantitative approach, 2) the population of this study was the English final test; test-pack A and B; and students’ answer sheets of the 12th grade students at the first semester, 3) the sample was the English final tests and 40 students’ answer sheets of the 12th grade students which was taken randomly, 4) the research instrument was documentation, and 5) the data analysis method was test item analysis. The findings of this study showed that both test-packages were lack of content and construct validity. In term of the content validity, both tests - packages did not fully test all materials stated in the syllabus, furthermore, one of the skills of language was not tested at all, listening. Related to the construct validity, some of the techniques of testing used to test language skills were not relevant to the language testing theory, especially writing and speaking because the test was in the form of multiple-choice while these skills need practicing in order to evaluate them. Then, one of the test-packages categorized to have low reliability with the coefficient reliability of 0.48, and another one was high with the coefficient reliability of 0, 72. The analysis on the level difficulty of both test-packs showed that the percentage of the easy items of test-package A was 72.5%, and 60% for test-package B; fair items was 17.5% of test-package A and 27.5% of test-package B; and difficult items was 10% of test-package A and 7.5% of test-package B. It means that both test-packages were too easy for the students. Next, for the discrimination power of both test-packages were 20% of excellent test items for test-package A and 12.5% for test-package B; 5% of good test items for test-package A and 17.5% for test-package B; 70% of poor test items for test-package A and 62.5% for test-package B; and 5% of very poor items for test-package A and 7.5% for test-package B. It means that both-test-packages could not really discriminate the students. In line with the discrimination power, the effective of the distractor analysis for both test-packs also showed bad result in which the distractors were dominated by the omit distractors with the percentage of 83.125% for test-package A and 65.385% for test-package B. Omit distractor means that the distractors must be removed or revised totally. Keywords: final test, validity, reliability, difficulty level, discrimination power, distractor efficiency

Actions (login required)

View Item