Abstarct: Abstract Due to numerous advancements in computing technology and software, various studies are conducted on a research problem from various perspectives today. It results in a large volume and excessive dispersion of scientific publications, even on a specific topic. To achieve a new aggregated result with less error, mexta-analyses combine the results of several studies. Measurement and sampling errors reduce the validity and reliability of single studies. mexta-analysis uses different statistical models to extract composite estimators for the unknown parameters in order to avoid this problem. Its main goal is to reduce the error of previous studies. Additionally, the inherent heterogeneity between studies must be modeled and considered. A variety of fixed, random, and mixed effect models have been developed as a result. By using R software and the mexta package, this thesis investigates the mexta-analysis approach with different statistical models and their application in real and practical examples.