The method of increasing the reliability of online survey

• Martyshenko S.N.

  Vladivostok State University of Economics and Service
  Vladivostok. Russia

In recent years, the number of online surveys has increased many times, which are used as research tools for economic processes. This was facilitated by the distribution of special services for the design of questionnaires for further presentation in the network. Because of the lack of direct contact with respondents, the problem of data reliability during online surveys only worsened compared to traditional paper-based surveys. The paper presents a method for assessing the reliability of respondents’ answers based on an analysis of the pairwise comparison of the values nominal characteristics that are most common in online surveys. The method is based on the procedure of “sliding examination”. The main hypotheses that form the basis of the developed method of reliability of questionnaire data are formulated. On the basis of multidimensional data from the sample formed as a result of the questionnaire, a single integral criterion is calculated that makes it possible to identify “suspicious” questionnaires that contradict the general logic of respondents’ perception of the socioeconomic process studied in the questionnaire. The procedure for calculating the criterion is considered using an example of some model data. The advantage of using the criterion for assessing the reliability of data is the speed of data analysis in the process of collecting them in the network, which makes it possible to make decisions to improve the quality of the data collected. The considered method can be used for the analysis of various types of data after preliminary transformation of characteristics to a nominal scale of measurement.

Keywords: questionnaire, data quality, online survey, multidimensional statistical methods, the criterion for assessing data quality, computer technology, Internet service.