Doğrusal Olmayan Kanonik Korelasyon Analizi ve Türkiye Sağlık Araştırması Verilerine Uygulanması

Abstract

In this study, it was aimed to examine "Nonlinear Canonical Correlation Analysis (NCCA)", which is defined as a multivariate statistical method that analyzes the relationships between two or more variable sets, including categorical variables. In the study, the 2019 Turkey Health Survey (THS) microdata set, which was carried out by the Turkish Statistical Institute, was used as application data. Thus, within the line of THS data, it was intended to reveal the relationship between the socio-demographic characteristics, health problems, functional competencies, and health attitudes of the individuals and to provide an easier interpretation of the mentioned data with a holistic approach. In this way, bringing a new perspective to the evaluation of categorical research data, which is generally analyzed with basic descriptive statistics, was the main viewpoint. In the analysis, selected population groups were examined, which were formed by considering variables such as age, employment status, and disability. As a result, the determinants of general health status for the selected population were generally low back, neck, hypertension, and arthrosis diseases; and it was determined that the body mass index of the people who do not have the aforementioned disease are normal and these people do not have sleep problems and do not use medicines. In addition, widowed, lonely, uneducated, and elderly people experience delays in getting health care due to distance to hospitals or transportation problems. Besides, it was determined that the divorced female population in the 25-54 age group with low household income and low education level could not meet their medical care, dental care, mental treatment, and medication needs due to difficulty in paying.