Pendekatan Weighted Least Square pada Meta Regresi (Studi Kasus: Efektivitas Acceptance and Commitment Therapy dalam Menurunkan Tingkat Depresi)

Arumningtyas, Felinda (2023) Pendekatan Weighted Least Square pada Meta Regresi (Studi Kasus: Efektivitas Acceptance and Commitment Therapy dalam Menurunkan Tingkat Depresi). Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6003212007-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2025. Download (2MB) | Request a copy

Abstract

Metode statistika terus berkembang, generalisasi terhadap populasi tidak hanya dilakukan berdasarkan hasil satu temuan saja, akan tetapi didasarkan pada sintesis beberapa hasil temuan dan sering disebut sebagai meta analisis. Selain itu, meta analisis juga dapat dilakukan untuk mengkaji kekonsistenan atau ketidakkonsistenan hasil penelitian, yang sering kali justru memperbesar terjadinya variasi antar hasil penelitian. Variasi atau heterogenitas dari seluruh penelitian tersebut disebabkan oleh perbedaan karakteristik-karakteristik penelitian. Pendekatan yang dapat digunakan dalam kaitannya dengan penjelasan heterogenitas hasil dari meta analisis dikenal dengan istilah meta regresi. Meta regresi merupakan perluasan untuk meta analisis yang menyelidiki sejauh mana heterogenitas antar hasil beberapa penelitian dapat dikaitkan dengan satu atau lebih dari karakteristik penelitian tersebut. Tujuan dari penelitian ini adalah untuk mendapatkan estimasi parameter pada model meta regresi. Estimasi parameter model pada meta regresi diselesaikan dengan pendekatan Weighted Least Square (WLS) dan untuk estimasi varian dilakukan menggunakan pendekatan DerSimonian&Laird. Data yang dijadikan unit penelitian adalah 33 penelitian terpublikasi yang membahas mengenai efektivitas Acceptance and Commitment Therapy (ACT) dalam menurunkan tingkat depresi yang dikumpulkan dari database PubMed, Google Scholar dan Science Direct. Hasil uji heterogenitas pada meta analisis menunjukkan varian antar penelitian adalah heterogen, sehingga model random effect meta analisis lebih tepat digunakan. Didapatkan gabungan effect size sebesar -0,321 yang menunjukkan bahwa ACT dapat menurunkan tingkat depresi dilihat dari menurunnya tingkat depresi kelompok eksperimen jika dibandingkan dengan kelompok kontrol. Heterogenitas yang diperoleh sebesar 92,58% yang menunjukkan heterogenitas yang tinggi dan harus ditelusuri penyebabnya, hasil meta regresi menunjukkan bahwa variabel rata-rata usia pasien, lama sesi terapi dan continent dapat menjelaskan heterogenitas di antara effect size pengamatan.
=======================================================================================================
In the development of statistics, generalisation to the population is not only done based on the results of one finding, but is based on the synthesis of several findings and is often referred to as meta-analysis. In addition, meta-analysis can also be conducted to assess the consistency or inconsistency of research results, which often actually increases the variation between research results. The variation or heterogeneity of all studies is caused by differences in research characteristics. An approach that can be used in relation to explaining the heterogeneity of results from meta-analysis is known as meta regression. Meta regression is an extension to meta-analysis that investigates the extent to which heterogeneity among the results of multiple studies can be attributed to one or more of the study characteristics. The purpose of this study is to obtain parameter estimates on the meta regression model. Estimation of model parameters in meta regression is completed with the Weighted Least Square (WLS) approach and for variance estimation is done using the DerSimonian & Laird approach. The data used as research units are 33 published studies that discuss the effectiveness of Acceptance and Commitment Therapy (ACT) in reducing depression levels collected from PubMed, Google Scholar and Science Direct databases. The results of the homogeneity test in the meta-analysis showed that the variance between studies was heterogeneous, so the random effect meta-analysis model was more appropriate to use. A combined effect size of -0.321 was obtained, indicating that ACT can reduce the level of depression as seen from the decrease in the level of depression of the experimental group when compared to the control group. The heterogeneity obtained is 92.58% which indicates high heterogeneity and must be traced the cause, the results of meta regression show that the variables of average patient age, length of therapy sessions and continent can explain the heterogeneity among the effect size observations.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	ACT, Depression, Meta analysis, Meta Regression, Weighted Least Square,Depresi, Meta Analisis, Meta Regresi, Weighted Least Square.
Subjects:	Q Science > QA Mathematics > QA275 Theory of errors. Least squares. Including statistical inference. Error analysis (Mathematics)
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49101-(S2) Master Thesis
Depositing User:	Felinda Arumningtyas
Date Deposited:	10 Aug 2023 08:59
Last Modified:	10 Aug 2023 08:59
URI:	http://repository.its.ac.id/id/eprint/104538

Actions (login required)

View Item