This paper presents basic methods and theoretical background of panel data analysis as a very important aspect of numerous research topics in natural and social sciences. Panel data contain and analyze the observations of phenomena obtained over multiple time periods and explain influencing variables. First, the paper presents the linear regression that forms the basis of statistical modeling, which links the dependent variable to the function of independent variables. The next chapter provides an overview of data type and explains specific structure of panel data as combination of cross-section and time series data. Afterwards, text explains various linear models of the panel data: pooled OLS model, fixed effects model and random effects model, as well as tests to evaluate how statistical model corresponds to the data: \(F-\) test, the Breusch-Pagan test and Hausman specification test. The theoretical concept of the panel data analysis presented in this paper is outlined through data aggregated by the European Statistical Office (Eurostat) and computed in the statistical analysis software SAS. Presented panel analysis is based on annual collections of statistics on international migration flows as dependent variable, and statistics on monthly wages, unemployment, educational attainment level and divorce rate as independent variables. Besides Croatia, four other European countries have been selected as observation units over the period of 2008 to 2016.