METODE PENDETEKSI PERILAKU ANOMALI DALAM TES BERBASIS KOMPUTER BERSKALA BESAR BERDASARKAN RESPONS JAWABAN DAN WAKTU RESPONS ANOMALOUS BEHAVIOR DETECTION METHOD IN LARGE SCALE COMPUTER-BASED TEST BASED ON RESPONSES AND RESPONSE TIME

DENI HADIANA, . (2020) METODE PENDETEKSI PERILAKU ANOMALI DALAM TES BERBASIS KOMPUTER BERSKALA BESAR BERDASARKAN RESPONS JAWABAN DAN WAKTU RESPONS ANOMALOUS BEHAVIOR DETECTION METHOD IN LARGE SCALE COMPUTER-BASED TEST BASED ON RESPONSES AND RESPONSE TIME. Doktor thesis, UNIVERSITAS NEGERI JAKARTA.

	Text COVER.pdf Download (786kB)
	Text BAB 1.pdf Restricted to Registered users only Download (465kB) | Request a copy
	Text BAB 2.pdf Restricted to Registered users only Download (1MB) | Request a copy
	Text BAB 3.pdf Restricted to Registered users only Download (800kB) | Request a copy
	Text BAB 4.pdf Restricted to Registered users only Download (3MB) | Request a copy
	Text BAB 5.pdf Restricted to Registered users only Download (463kB) | Request a copy
	Text DAFTAR PUSTAKA.pdf Restricted to Registered users only Download (693kB) | Request a copy
	Text LAMPIRAN.pdf Restricted to Registered users only Download (3MB) | Request a copy

Abstract

Penelitian kuantitatif ini bertujuan menganalisis dan membandingkan metode untuk mendeteksi perilaku anomali dalam tes berbasis komputer berskala besar berdasarkan respons jawaban dan waktu respons secara komprehensif. Data respons jawaban dan waktu respons yang digunakan dalam penelitian ini merupakan data sekunder yang diperoleh dari Pusat Penilaian Pendidikan, Badan Penelitian dan Pengembangan, Kementerian Pendidikan dan Kebudayaan. Data respons jawaban dan waktu respons diolah dan dianalisis secara statistik untuk digunakan dalam penentuan dan perbandingan metode untuk mendeteksi perilaku anomali peserta tes dengan mempertimbangkan waktu ambang batas, waktu respons, respons jawaban, parameter butir, dan parameter kemampuan peserta tes. Berdasarkan hasil analisis dan interpretasi terhadap metode pendeteksi perilaku anomali diperoleh hasil: (1) metode penentuan waktu ambang batas berdasarkan TS yang dikembangkan pada penelitian ini lebih sensitif daripada metode VI, 2,5SD, dan 2,75SD dalam membedakan perilaku anomali dan perilaku wajar peserta tes berdasarkan waktu respons; dan (2) metode outfit Model Rasch lebih efektif dalam mendeteksi perilaku anomali random responding berdasarkan respons jawaban dibandingkan metode IJ; (3) metode TC-TS yang dikembangkan pada penelitian merupakan metode yang paling efektif dalam mendeteksi perilaku anomali peserta tes berdasarkan waktu respons; (4) Metode BC-TS, SC-TS, PBC-TS, PSC-TS, BBC-TS, dan BSC-TS yang dikembangkan pada penelitian merupakan metode yang paling efektif dalam mendeteksi perilaku anomali peserta tes berdasarkan gabungan waktu respons dan respons jawaban. Kata Kunci : waktu respons, ambang batas, respons jawaban, metode pendeteksi anomali, perilaku anomali. **** This quantitative study aims to analyze and compare methods for detecting anomalous behavior in large-scale computer-based tests based on responses and response times comprehensively. The responses data and response time used in this study is secondary data obtained from the Center for Educational Assessment, Research and Development Agency, Ministry of Education and Culture. The responses and response time are processed and analyzed statistically for use in determining and comparing methods to detect test taker anomalous behavior by considering the threshold time, response time, responses, item parameters, and ability or person parameter. Based on the results of the analysis and interpretation of the anomalous behavior detection method, it is found that (1) the method of determining the threshold time based on TS developed in this study is more sensitive than the VI, 2.5SD, and 2.75SD methods in distinguishing anomalous behavior and normal behavior of test taker based on response time; and (2) the Rasch Model outfit method is more effective in detecting random responding anomalous behavior based on responses compared to the IJ method; (3) the TC-TS method developed in this study is the most effective method in detecting the anomalous behavior of test-takers based on response time; (4) The BC-TS, SC-TS, PBC-TS, PSC-TS, BBC-TS, and BSC-TS methods developed in this study are the most effective methods in detecting anomalous behavior of test-takers based on response time and responses. Keywords: response time, threshold, responses, anomalous detection method, anomalous behavior.

Item Type:	Thesis (Doktor)
Additional Information:	1). Bahrul Hayat, Ph.D. ; 2). Prof. Dr. Burhanuddin Tola, M.A.
Subjects:	Pendidikan > Pendidikan Pendidikan > Teori, Penelitian Pendidikan Pendidikan > Psikologi Pendidikan
Divisions:	PASCASARJANA > S3 Penelitian Dan Evaluasi Pendidikan
Depositing User:	Users 5444 not found.
Date Deposited:	05 Nov 2020 06:09
Last Modified:	05 Nov 2020 06:09
URI:	http://repository.unj.ac.id/id/eprint/11574

Actions (login required)

View Item