ANALISIS RESPON ITEM PADA INSTRUMEN WORLD VALUES SURVEY BERDASARKAN MODEL MULTIDIMENSIONAL GRADED RESPONSE

NOVIA SUCY ARISTAWIDYA, . (2023) ANALISIS RESPON ITEM PADA INSTRUMEN WORLD VALUES SURVEY BERDASARKAN MODEL MULTIDIMENSIONAL GRADED RESPONSE. Sarjana thesis, UNIVERSITAS NEGERI JAKARTA.

	Text COVER.pdf Download (1MB)
	Text BAB 1.pdf Download (318kB)
	Text BAB 2.pdf Restricted to Registered users only Download (1MB) | Request a copy
	Text BAB 3.pdf Restricted to Registered users only Download (426kB) | Request a copy
	Text BAB 4.pdf Restricted to Registered users only Download (865kB) | Request a copy
	Text BAB 5.pdf Restricted to Registered users only Download (229kB) | Request a copy
	Text DAFTAR PUSTAKA.pdf Download (217kB)
	Text LAMPIRAN.pdf Restricted to Registered users only Download (769kB) | Request a copy

Abstract

Multidimensional Graded Response Model (MGRM) merupakan salah satu pemodelan IRT multidimensional untuk data respon bertipe politomus ordinal yang mengasumsikan daya pembeda (item discrimination) bervariasi antar item. Pada penelitian ini, MGRM digunakan untuk menganalisis instrumen World Values Survey (WVS), yang merupakan program penelitian internasional untuk studi ilmiah dan akademik tentang nilai sosial, politik, ekonomi, agama, dan budaya masyarakat di dunia. Berdasarkan instrumen WVS, dilakukan pengukuran minat politik dan partisipasi politik responden asal Indonesia. Responden yang dianalisis pada penelitian ini berjumlah 1555 orang dan item yang diteliti adalah sebanyak 22 item. Hasil penelitian menunjukkan bahwa asumsi multidimensional IRT dengan dua sifat laten terpenuhi berdasarkan scree plot yang membentuk siku (elbow) pada faktor ketiga. Hubungan antar item dengan sifat laten ditentukan secara eksploratori karena tidak terdapat hipotesis terdahulu mengenai struktur tes. Pemodelan MGRM dibangun dengan hubungan kompensansi antar faktor dan struktur sederhana (between-item). Uji kesesuaian menunjukkan item fit lebih dari 70%, person fit sebesar 97,62%, dan model fit yang cukup baik dari statistik AIC, BIC, dan G2. Di antara 22 item yang dianalisis, sebanyak 6 item mengalami misfit. Besarnya informasi item yang diberikan oleh masing-masing item bervariasi dengan total informasi tes sebesar 6. Informasi tes paling besar diperoleh jika responden memiliki sifat laten minat politik dan partisipasi politik masing-masing dari -2,0 sampai 2,0. Hubungan antara minat dan partisipasi politik saling mengompensansi dengan korelasi sebesar 0,381. ******* Multidimensional Graded Response Model (MGRM) is one of the MIRT models for ordinal polytomous response data which assumes item discrimination varies between items. In this study, MGRM was used to analyze the World Values Survey (WVS) instrument, which is an international research program for scientific and academic studies on the social, political, economic, religious, and cultural values of people around the world. Based on the WVS instrument, the Indonesian respondent’s political interest and political participation were measured. Total respondents in this study were 1555 and the items analyzed were 22 items. The results showed that the assumption of a multidimensional IRT with two latent traits was fulfilled based on the scree plot forming an elbow on the third factor. The relationship between items and the latent traits was determined exploratory because there were no prior hypotheses regarding the structure of the test. In this study, MGRM is built with compensation relationships between factors and a simple structure (between-item). The fit test showed an item fit of more than 70%, a person fit of 97.62%, and a fairly good fit model from AIC, BIC, and G2 statistics. Among the 22 items analyzed, 6 items experienced a misfit. The amount of item information provided by each item varies with a total of 6. The largest test information is obtained if the respondent has political interest and political participation latent traits respectively from -2.0 to 2.0. The relationship between political interest and political participation compensate each other with a correlation of 0.381.

Item Type:	Thesis (Sarjana)
Additional Information:	1). Dr. Dian Handayani, M.Si. ; 2). Faroh Ladayya, M.Si.
Subjects:	Filsafat, Psikologi & Agama > Psikologi, Ilmu Jiwa Ilmu Sosial > Statistik Sains > Matematika > Software, Sistem Informasi Komputer
Divisions:	FMIPA > S1 Statistika
Depositing User:	Users 20014 not found.
Date Deposited:	29 Sep 2023 01:47
Last Modified:	29 Sep 2023 01:47
URI:	http://repository.unj.ac.id/id/eprint/43012

Actions (login required)

View Item