Information reliability assessment and filtering for decision making. Case study of Lithuanian and China students’ life goals survey data

Abstract

This article presents the methodology and tools to evaluate the reliability of quantitative sociological research data. The problem of filtering unreliable data is usually solved by statistical methods. This article proposes an improved method for filtering unreliable data. In this case, the statistical methods are not applied to the initial data but the value of the distance function between the two preferences. This allows for the disclosure of conflicting or erroneous data. Calculation of the distance between two preferences and prioritisation of life goals are based on binary relation theory, where the properties of symmetry (antisymmetry) are very important. The article presents a case study on 11 life goals evaluation and ranking by Lithuanian and China students. The study revealed that the China student data filtered at least twice as much as the Lithuanian student data, i.e., they are less reliable. The filtered data show that students of both countries ranked the most and the least important life goals in a very similar way with minimum deviations detected in the ranking results.