the upper group, it can be said that the distracter did not function as it
should be.
One of the objectives of item analysis is to know about the answer distribution to a subject in alternative answers. Through distracter
efficiency, teacher may know the number of students who answered correctly, which distractor is too showy and make it easier for students
not to vote, the misleading distracter and the distractor who managed to attract lower group students.
29
Distractor is considered as an effective functioning if it is selected at least by 5 of examinees, and if it is chosen by less than
5 of examinees, it means that the distracter is categorized as an ineffective non-functioning distracter. This principle is based on
Marrie Tarrant’s theory,
30
and Suharsimi Arikunto’s theory.
31
C. Relevant Studies
The research concerning in item analysis have numerously been carried out by. The first study comes from Sunan Kalijaga State Islamic University
written by Rakhmaddiniah Sulistyawati 2012. The title of her study is “An
Analysis of items on Arabic Final Examination on the eighth grade of State Islamic Junior High School Piyungan, Bantul Academic Year 20122013. This
study focused on the validity, reliability, discriminating power, level of difficulty and the conformity between item test, standard competence and basic competence.
This study use program from Microcat Model ITEMAN version 3.0 and Anates version 4.0.5 to study the data. Although this study use document based analysis,
the researcher here also uses computer analysis program to make her analysis of data easier to process. The result of her study is 55 of the items are valid, the
reliability score passed 0.83 which means good, the difficulty level of the items is
29
Sumarna Surapranata, Analisis, Validitas, Realibilitas, dan Interpretasi Hasil Tes, Bandung: Remaja Rosdakarya Offset, 2006, pp. 43-44.
30
Marrie Tarrant, et.all, “An Assessment of Functioning and Non-functioning Distractors in Multiple-
Choice Questions”, BMC Medical Education, 9, 2009, p. 2.
31
Arikunto, op. cit., p. 220.
55 good, but for the discriminating power it has 47.5 bad and most of the distracters did not work as it should. Based on the feedback from this study, the
researcher concludes that the items of a summative test may be categorized as good items, but since the ability and comprehensive of students
’ skill in receiving the material is still low; therefore, the items that have good quality will turn out
into lousy. So it would be better if the teacher gives more concern in the mastery and the development to teaching material in the Arabic.
32
The second study is conducted by Marrie Tarrant, James Ware and Ahmed M Mohammed. This study was taken at Department of Nursing Studies, Li Ka
Shing Faculty of Medicine at 2009. The title of the study is “An Assesment of
functioning and non-functioning distractors in multiple-choice questions. The purpose of this study was to examine non-functioning distractors in teacher-
developed tests to provide additional guidance to teachers in health-science disciplines regarding the optimal number of distractors to include in MCQs. This
study uses item-analysis data, they assessed the proportion of non-functioning distractors on seven test papers administered to undergraduate nursing students.
tThey reviewed 514 items, 1542 distractors and 514 correct responses. Non- functioning options were defined as ones that were chosen by fewer than 5 of
examinees. The study shows a result that the proportion of items containing 0,1,2 and 3 functioning distractors was 12.3, 34.8, 39.1 and 13.8 respectively.
Overall, items contained an average of 1.54 SD = 0.88 functioning distractors. Only 52.2 n=805 of all distractors were functioning efficiently and 10.2
n=158 had a choice frequency of 0. The result showed that the majority of items developed by teacher will not have more than two functioning distractors, and
including more distractors may not be a good investment in item development. Three option-items have many advantages for both item writers and examiners,
while further non-functioning distractors are not likely increase the item.
33
32
Sulistyawati, Rakhmaddiniah, “An Analysis of items on Arabic Final Examination on 8th grade of State Islamic Junior High School Piyungan, Bantul
Academic Year 20122013”, Skripsi of UIN Sunan Kalijaga, Yogyakarta, 2013, pp. 50-93, not published.
33
Tarrant. loc. cit.
The last is the study titled “Relationship between Types of Distractor and
Difficulty of Multiple-Choice Vocabulary Tests in Sentential Context. This study was conducted by Yuko Hoshino. This study uses 372 students from seven
universities in Japan. Hoshino divided the participants randomly into a those who took the test with paradigmatically related distractors, b those who took the
test with syntagmatically related distractors, and c those who took the test with distractors with no related to a or b.This study focused on the change of test
difficulty by different types of distractors. The result showed that the test with syntagmatically related distractors were the most difficult and the test with
unrelated distractors, the easiest; the paradigmatically related distractors remained in the middle.
34
The three research above focused on analyzing the test items by doing item analysis with document based method. The first study focused on finding the
quality of the whole test items based on the validity, reliability, difficulty and discrimination index, also distractors efficiency in general. The second and the
third study focused on the non-functioning distractors. The second study focused on investigating the NF-D in teacher-made test while the third study focused on
the changing of difficulty index by the difference of distractors on English test. All of the research above had both strength and weakness. The strength of
these studies is the researcher doing it in order to make a good item through difficulty and discrimination index, also the relation between the ideal of item and
non-functioning distractor. The weakness of these studies is all of the researcher of those previous studies did not rewrite the items based on their analysis result.
Therefore, the writer interested in conducting a study in item analysis to overcome the weakness occurs in those previous studies. This study will hopefully,
turn out quite different from those studies above, because the writer intended to include the strength side from the research above also intended to revising the
items based on analysis result.
34
Yuko Hoshino, “Relationship between Types of Distractor and Difficulty of Multiple- choice Vocabulary Tests in Sentencial Co
ntext”, Springer; Language Testing in Asia, 3, 2013, pp. 1-12.
