Puspita, Ika (2025) Pemodelan Partial Least Square (PLS) dengen Pendekatan PATHMOX Menggunakan Variabel Mediasi (Studi Kasus: Jumlah Pasien Tuberkulosis). Masters thesis, Institut Teknologi Sepuluh Nopember.
![]() |
Text
6003232006-Master_Tesis.pdf - Accepted Version Restricted to Repository staff only Download (3MB) | Request a copy |
Abstract
Salah satu teknik analisis statistik multivariat yang dapat menjelaskan pola hubungan kompleks yang melibatkan banyak variabel adalah Structural Equation Modeling (SEM). SEM berbasis varians (Variance-Based SEM atau VB-SEM), yang juga dikenal sebagai Partial Least Squares (PLS), merupakan metode yang lebih fleksibel dalam SEM karena mampu mengatasi berbagai keterbatasan, seperti skala indikator variabel yang berbeda, jumlah sampel yang terbatas, dan asumsi distribusi data. PATHMOX merupakan strategi khusus dalam PLS-SEM yang digunakan untuk melakukan segmentasi hasil model struktural guna mengatasi heterogenitas yang teramati. Penelitian ini bertujuan untuk membangun model persamaan struktural (SEM-PLS) serta melakukan segmentasi hasil model persamaan struktural menggunakan pendekatan PATHMOX. Penelitian ini menggunakan data sekunder dari 1019 pasien TB yang di ambil dari Dinas Kesehatan Kabupaten Paser pada tahun 2022. Dengan pendekatan PATHMOX, menghasilkan empat segmentasi utama yaitu, segmen pertama terdiri dari laki-laki dewasa sebanyak 313 pasien. Segmen kedua terdiri dari perempuan dewasa sebanyak 258 pasien. Segmen ketiga mencakup anak-anak dan remaja sebanyak 199 pasien serta segmen keempat mencakup bayi, balita, dan lansia, dengan jumlah 249 pasien. Setiap segmen memiliki karakteristik yang berbeda. Secara umum, proporsi pasien dengan riwayat diabetes di setiap segmen relatif rendah, yaitu di bawah 9%. Sebaliknya, setiap segmen menunjukkan proporsi riwayat pengobatan yang cukup tinggi, yaitu di atas 70%. Pemeriksaan diabetes melitus pada setiap segmen sebagian besar tidak diketahui, dengan rata-rata ketidaktahuan di atas 87% pada seluruh segmen.
Kata kunci: Heterogenitas, Partial Least Square (PLS), PATHMOX, Segmentasi, Structural Equation Modelling (SEM), Tuberkulosis.
==========================================================================================================================================
One of the multivariate statistical analysis techniques that can explain complex patterns of relationships involving multiple variables is Structural Equation Modeling (SEM). Variance-based SEM (VB-SEM), also known as Partial Least Squares (PLS), is a more flexible method within SEM because it can address various limitations, such as differing variable indicator scales, small sample sizes, and distributional assumptions. PATHMOX is a specific strategy within PLS-SEM used to segment the structural model results in order to address observed heterogeneity. This study aims to construct a structural equation model (PLS-SEM) and perform segmentation of the structural model results using the PATHMOX approach. The study uses secondary data from 1,019 TB patients obtained from the Paser District Health Office in 2022. Using the PATHMOX approach, four main segments were identified: the first segment consisted of 313 adult male patients, the second of 258 adult female patients, the third included 199 children and adolescents, and the fourth consisted of 249 infants, toddlers, and elderly patients. Each segment has distinct characteristics. In general, the proportion of patients with a history of diabetes in each segment is relatively low, under 9%. Conversely, all segments show a relatively high proportion of patients with a history of previous TB treatment, exceeding 70%. Diabetes mellitus testing in each segment is mostly unknown, with an average rate of unawareness above 87% across all segments.
Keywords: Heterogeneity, Partial Least Square (PLS), PATHMOX, Segmentation, Structural Equation Modelling (SEM), Tuberculosis
Item Type: | Thesis (Masters) |
---|---|
Uncontrolled Keywords: | Kata kunci: Heterogenitas, Partial Least Square (PLS), PATHMOX, Segmentasi, Structural Equation Modelling (SEM), Tuberkulosis. Keywords: Heterogeneity, Partial Least Square (PLS), PATHMOX, Segmentation, Structural Equation Modelling (SEM), Tuberculosis |
Subjects: | H Social Sciences > HA Statistics Q Science > QA Mathematics Q Science > QA Mathematics > QA278.3 Structural equation modeling. Q Science > QA Mathematics > QA278 Cluster Analysis. Multivariate analysis. Correspondence analysis (Statistics) |
Divisions: | Faculty of Mathematics and Science > Statistics > 49101-(S2) Master Thesis |
Depositing User: | Ika Puspita |
Date Deposited: | 06 Aug 2025 03:59 |
Last Modified: | 06 Aug 2025 03:59 |
URI: | http://repository.its.ac.id/id/eprint/127730 |
Actions (login required)
![]() |
View Item |