Cluster analysis from multivariate statistical methods; An application of chronic smokers

IV SUMMARY Cluster Analysis from Multivariate Statistical Methods; An Application of Chronic Smokers This study's aim is to analyze multivariate construction about smoking using clustering methods. Clustering methods form successful cluster about this kind of multivariate construction. This study was carried out on six groups including students, teachers, health officers (doctors, nurses), policemen, tradesmen and unemployed for the purpose of analyzing the variables about smoking. 64 variables belonging to 1133 individuals were investigated. According to our purpose, the ward cluster analysis which reveals proportions of interesting observations related to these variables together with cluster constructions was carried out. Taking into consideration the data related to these six groups, we got dendograms about these variables. From the variables about smoking through the dendograms, we got clusters that reveal the sexualities, different habits, social and psychological situations of individuals. Because of being visual, we had more concrete conclusion while interpreting. As conclusion using the ward cluster analysis which is carried out on the variables related to smoking. We have successful and meaningful conclusions from the clusters about the subject. Key Words : Cluster Analysis, Hierarchical Cluster Analysis, Nonhierarchical Cluster Analysis, Smoking