Random Effects Meta Regression on the Effectiveness of Acceptance and Commitment Therapy for Depression
Felinda Arumningtyas (a), Bambang Widjanarko Otok (a*), Santi Wulan Purnami (a)

a) Faculty of Science and Data Analytics, Institut Teknologi Sepuluh Nopember, Kampus ITS-Sukolilo, 60111 Surabaya
*dr.otok.bw[at]gmail.com

Abstract

Meta-analysis is a statistical technique for summarising the results of two or more similar studies into a blend of quantitative data. Heterogeneity between studies often occurs in meta-analysis. Meta-regression is an extension of meta-analysis that can explain heterogeneity among the results of multiple studies can be attributed to one or more study characteristics. The model in meta-regression that involves variance between studies is the random effects model. The purpose of this study is to apply the random effects meta-regression model. Estimation of model parameters in random effects meta regression uses the Weighted Least Square (WLS) method and for variance estimation is done using the DerSimonian & Laird approach. The data used as research units were 33 published studies that discussed the effectiveness of Acceptance and Commitment Therapy (ACT) in reducing depression levels collected from PubMed, Google Scholar, and Science Direct databases. A combined effect size of -0.321 was obtained, indicating that ACT can reduce depression levels as seen from the decrease in depression levels 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 results of meta regression show that the variables of average patient age and length of therapy sessions can explain the heterogeneity between effect size.

Keywords: ACT- Depression- Meta Analysis- Meta Regression- Weighted Least Squares

Topic: Mathematics and Statistics