EXAMINATION OF PRECISION EVOLUTION OF POPULATION PARAMETER USING A SIMPLE AND STRATIFIED RANDOM SAMPLENemanja Budimir |
Published 11/05/2021 |
ABSTRACT
In order for research to be conducted, it is very important to choose the appropriate sample. Namely, it is important to choose a sample that will show the basic characteristics of the population. Simple random sampling is a sampling procedure or plan in which each sample of size n has the same probability of selection. This form of sampling is practically performed by selecting one member in the sample from the total number of N population units on the list or list of population members in each step, but each population member is given an equal chance to be sampled. On the other hand, stratified sampling is common when taking samples from very heterogeneous populations, i.e. populations that can be divided into strata, layers or subpopulations according to some characteristics. The stratified sample is one of the most commonly used sample plans. This paper examines the accuracy of estimating a simple and stratified random sample on the market of domestic scientific journals from Bosnia and Herzegovina, Croatia and Serbia. The SJR indicator(Scimago Journal & Country Rank) was chosen as a corresponding feature value. The SJR indicator is used for the evaluation and analysis of journals in the field of science. The conducted analyzes showed that the stratified sample provides more precise evaluation than the simple random sample. It follows that the stratified sample has less variance and its middle average is closer to the middle average of the population. It can be concluded that this sample has smaller deviations from the parameters of population.
Keywords: systemic risk, financial system, systemic risk indicators, regulatory framework.