PEMODELAN ITEM RESPON POLITOMUS UNTUK MENGUKUR TINGKATAN STRES AKADEMIK MAHASISWA

RAHFA QUR'ANIYATIN DHUHA, . (2025) PEMODELAN ITEM RESPON POLITOMUS UNTUK MENGUKUR TINGKATAN STRES AKADEMIK MAHASISWA. Sarjana thesis, UNIVERSITAS NEGERI JAKARTA.

	Text COVER.pdf Download (1MB)
	Text BAB I.pdf Download (513kB)
	Text BAB II.pdf Restricted to Registered users only Download (859kB) | Request a copy
	Text BAB III.pdf Restricted to Registered users only Download (685kB) | Request a copy
	Text BAB IV.pdf Restricted to Registered users only Download (1MB) | Request a copy
	Text BAB V.pdf Restricted to Registered users only Download (490kB) | Request a copy
	Text DAFTAR PUSTAKA.pdf Download (378kB)
	Text LAMPIRAN.pdf Restricted to Registered users only Download (994kB) | Request a copy

Abstract

Stres dikalangan mahasiswa yang disebabkan oleh faktor akademik dinamakan stres akademik. Keberadaan tingkatan stres akademik yang dialami seorang mahasiswa merupakan karakteristik/peubah laten yang dapat diduga melalui pengukuran terhadap beberapa peubah indikator yang diperoleh dari jawaban responden pada suatu instrumen pengukuran. Salah satu pendekatan yang dapat memodelkan hubungan antara jawaban responden dengan karakteristik laten responden adalah Item Response Theory (IRT). Penelitian ini menggunakan instrumen hasil modifikasi dari Perceptions of Academic Stress (PAS) oleh Bedewy & Gabriel (2015). Setiap item memiliki kategori respons bertipe politomus dengan skala ordinal, sehingga digunakan model IRT politomus ordinal, yaitu Partial Credit Model (PCM), Generalized Partial Credit Model (GPCM), dan Graded Response Model (GRM). Responden yang dianalisis merupakan mahasiswa program studi non-kependidikan Universitas Negeri Jakarta sebanyak 300 responden. Jumlah item yang diteliti sebanyak 23 item dengan 4 kategori respon ("sangat tidak setuju" hingga "sangat setuju"). Hasil penelitian menunjukkan bahwa GRM merupakan model yang paling sesuai (fit) berdasarkan statistik AIC dan BIC. Pengujian asumsi dasar IRT terpenuhi, yaitu unidimensi, independensi lokal, dan invariansi parameter butir. Setiap item memiliki nilai dugaan daya pembeda yang positif dan dugaan tingkat kesulitan yang terurut secara menaik. Hasil dugaan karakteristik laten responden sangat bervariasi tergantung pada sumber stresnya yaitu harapan akademik, beban tugas dan ujian, serta persepsi diri akademik. Uji kesesuaian item menunjukkan seluruh item (100%) pada dimensi harapan akademik serta beban tugas dan ujian, dan 85,71% item pada dimensi persepsi diri akademik tergolong fit. Sementara itu, uji kesesuaian responden menunjukkan lebih dari 90% responden fit pada setiap dimensi, sehingga instrumen dikatakan cocok (fit) dimodelkan dengan GRM.*****Stress among students caused by academic factors is known as academic stress. The level of academic stress experienced by a student is a latent characteristic/variable that can be estimated through measurements of several indicator variables obtained from respondents' answers to a measurement instrument. One approach that can model the relationship between respondents' answers and their latent characteristics is Item Response Theory (IRT). This study used a modified version of the Perceptions of Academic Stress (PAS) instrument by Bedewy and Gabriel (2015). Each item has a polytomous response category with an ordinal scale; therefore, the polytomous ordinal IRT models applied are the Partial Credit Model (PCM), Generalized Partial Credit Model (GPCM), and Graded Response Model (GRM). The respondents consisted of 300 students from non-education major at Universitas Negeri Jakarta. A total of 23 items were analyzed, each with four response categories from "strongly disagree" to "strongly agree". The results indicate that the GRM is the best model based on AIC and BIC. The basic assumptions of IRT were fulfilled, including unidimensionality, local independence, and item parameter invariance. Each item had a positive discrimination parameter and a monotonically increasing difficulty parameter. The latent trait estimates of the respondents vary depending on the sources of stress, including academic expectations, workload and examinations, as well as students’ academic self-perceptions. Item fit tests indicate that all items (100%) in the academic expectations and workload and examinations dimensions and 85.71% of items in the students' academic self-perceptions dimension are fit. Meanwhile, respondent fit tests show that more than 90% of respondents are fit in each dimension, indicating that the instrument demonstrates good fit according to the GRM.

Item Type:	Thesis (Sarjana)
Additional Information:	1). Dr. Dian Handayani, M.Si. ; 2). Dra. Widyanti Rahayu, M.Si.
Subjects:	Pendidikan > Psikologi Pendidikan Sains > Statistika
Divisions:	FMIPA > S1 Statistika
Depositing User:	Rahfa Qur'aniyatin Dhuha .
Date Deposited:	15 Aug 2025 03:35
Last Modified:	15 Aug 2025 03:35
URI:	http://repository.unj.ac.id/id/eprint/61310

Actions (login required)

View Item