Metode Kuadrat Terkecil Dua Tahap Dalam Structural Equation Models (SEM)

Sulistiyawan, Edy (2008) Metode Kuadrat Terkecil Dua Tahap Dalam Structural Equation Models (SEM). Masters thesis, Institut Teknologi Sepuluh November.

Text 1306201012-Master_thesis.pdf Restricted to Repository staff only Download (35MB)

Abstract

Ada beberapa metode estimasi dalam Structural Equation Models (SEM), diantaranya adalah Maximum Likelihood (ML) dan Metode Kuadrat Terkecil Dua Tahap (Two Stage Least Squares I TSLS). Metode estimasi ini mempunyai perbedaan yang mendasar, dimana ML memerlukan asumsi sampel besar dan variabel-variabel indikator berdistribusi normal multivariat, sedangkan TSLS tidak memerlukan asumsi distribusi apapun serta dapat digunakan dalam sampel kecil. Taksiran TSLS dapat diperoleh dengan cepat tanpa iterasi untuk beberapa model persamaart struktural yang berguna pada tahap awal penelitian. Penelitian ini mengkaji penaksir TSLS pada model persamaan struktural dalam Structural Equation Modeling baik dari segi teoritis maupun aplikasi. Hasil kajian empiris pada studi kasus data Kinerja Pengajar SMA Kristen Petra pada P3K Petra Surabaya menunjukkan bahwa TSLS mempunyai perilaku yang lebih baik dalam sampel kecil dibanding ML.
=====================================================================================================================================
There are some methods estimate in Structural Equation Models (SEM), they are Maximum Likelihood (ML) and Two Stage Least Squares (TSLS). This method estimate has the elementary difference, where ML need the big sanple of assumption and variable of indicator has a normal multivariate distribution, while TSLS doesn't need the any distribution assumption and also can be used in small sample. The obtainable valuation TSLS can be found swiftly without iterasi for a few Structural Equation Models which is good for the beginner research. This research study the appraiser TSLS of Structural Equation models in good Structural Equation Models in theoretical and application. The result of empirical case study of data performance the teacher of Christian Petra Senior High School at P3K Petra Surabaya indicate that the TSLS have the better behavior then ML in compared to by small sample.

Item Type:	Thesis (Masters)
Additional Information:	RTSt 519.42 Sul m
Uncontrolled Keywords:	Structural Equation Models, LISREL, Maximum Likelihood, Two Stage Least Squares (I'SLS).
Subjects:	Q Science > QA Mathematics > QA275 Theory of errors. Least squares. Including statistical inference. Error analysis (Mathematics)
Divisions:	Faculty of Mathematics and Science > Statistics > 49101-(S2) Master Thesis
Depositing User:	magang .
Date Deposited:	01 Oct 2026 04:18
Last Modified:	01 Oct 2026 04:18
URI:	http://repository.its.ac.id/id/eprint/145135

Actions (login required)

View Item