Studi Simulasi Menggunakan Fuzzy C-Means dalam Mengklasifikasi Konstruk Tes  

Oleh:

Rukli

Jenis: Article from Journal - ilmiah nasional - terakreditasi DIKTI - non-atma jaya
Dalam koleksi: Jurnal Penelitian dan Evaluasi Pendidikan vol. 15 no. 1 (2011), page 115-138.
Topik: fuzzy c-means; factor analysis; unidimensional
Fulltext: 1090-4222-1-PB.pdf (412.41KB)

Isi artikel

This paper introduces the fuzzy c-means method for classifying the test constructs. To verify the unidimensional a test typically uses factor analysis as part of parametric statistics with some strict requirements, while fuzzy c-means methods including method heuristic that do not require strict requirements. Simulation comparison between the method of factor analysis using SPSS program and fuzzy c-means methods using Matlab. Simulation data using data type dichotomy and politomus generated through Microsoft Office Excel programs each with a number of 3 items using the number of participants 10 tests, while the number of 30 test items using the number as many as 100 participants. The simulation results show that the fuzzy c-means method provides a more descriptive and dynamic grouping of all the designs that have been made in verifying the unidimensional as a test.

Opini Anda

Klik untuk menuliskan opini Anda tentang koleksi ini!

Kembali