Milan Terek, The School of Management, Slovak Republic

Type of article: Informative article
Pages: 82 – 100

Abstract

With the spread of the Internet, nonprobability sampling has become popular again in sample surveys as a quick and inexpensive method of creating samples that can be quickly implemented online. The paper aims to provide a comprehensive assessment of statistical surveys based on nonprobability samples. The basis of the quality assessment of probability samples is the total survey error. The quality of a nonprobability sample can only be assessed by considering the extent of matching the final modeled sample with the population in terms of various characteristics. We will present the conclusions emerging from studies on empirical evidence of the accuracy of surveys based on probability and nonprobability samples. In general, it is advisable to prefer, as is the case now, surveys based on probability samples.