Analiza klas ukrytych w badaniach sondażowych
Latent class analysis in survey research

Summary/Abstract: In survey research, in addition to manifest variables, we deal with latent variables. They are an abstract category usually used for the aggregation of properties contained in manifest variables. Latent class analysis allows to analyze discrete, as well continuous variables. We can also analyze non response datasets. Latent class analysis is based on two assumption. The first assumption is the condition of local independence, the second one is that the population is divided into homogeneous and exhaustive subpopulations which together form a latent class. Latent class analysis is in a sample survey, which goal is usually market segmentation. The main goal of this paper is to present latent class analysis in the analysis of social behavior. All calculations will be conducted in R software with the use of poLCA{poLCA} function.

• Details
• Contents