Analisis Item Dalam Pembuatan Tes

Sukoco - Universitas Surakarta

Abstract


Abstract

Men accessories can not be measured directly. To measure one’s competence it is needed an instrument, namely, test. Test consists of some items which each item describe every indicator. The indicators will make one’s competence. To make good test there must be item analysis that constructs it. There are two theories to analyze item, they are Classical Test Theory (CTT) and Item Response Theory (IRT). Item statistic on CTT are mostly used in selecting items, while item parameters on IRT are used in judging respondent scores, calibration, and equating test. The benefit of the parameters on IRT can be used on the adaptive test that is conducted by computer, bias study, and test equation.

 

Keywords: Item analysis, test construction,  clasical tes teory, modern test theory


Full Text:

PDF

References


Arikunto, S.1999. Dasar-dasar Evaluasi Pendidikan. Jakarta: Bumi Aksara

Allen, M.J. & W.M Yen. 1979. Introduction to measurement theory. Montere: Brooks/Cole Publising Company.

Azwar, S. 1997. Reliabilitas dan validitas. Yogyakarta: Pustaka Pelajar.

Hayat, B. 1997. Pedoman Penggunaan Iteman. Jakarta: Pusisjian Balitbang Depdiknas.

Cronbach, L. J. 1970. Essentials of Psychological Testing. New York: Harper and Row.

Ebel, R.L. & D. A. Frisbie. 1991. Essentials of Educational Measure-ment. New Jersey: Prentice Hall.

Camilli, G., & L.A. Shepard. 1994. MMSS Methods for Identifying Biased Test Items. Thousand Oaks, CA: Sage Publication.

Hambleton, R.K., & Swaminathan, H., & Rogers, H.J. 1991. Fundamental of Item Response Theory. Newbury Park, CA: Sage.

Kerlinger, F. N. 2000. Asas – asas Penelitian Behavioral. Edisi ketiga. (Terjemahan Landrung R Simatupang). Yogjakarta: Gajah Mada University Press. Buku asli diterbitkan tahun 1986.

Naga, D. S. 1992. Pengantar Teori Sekor pada Pengukuran Pendidikan. Jakarta: Besbats.

Sudjana. 1995. Desain dan Analisis Eksperimen. Bandung: Tarsito.

Wu, M. L. 1997. The Development and Applicatian of a fit Test for UseWith Marginal Maximum Likelihood Estimation and Genera-lised Item Response Models. Tesis tidak diterbitkan. Melbourne: University of Melbourne.


Refbacks

  • There are currently no refbacks.