Sten Score Method And Cluster Analysis: Identifying Respondents Vulnerable To Drug Abuse

Abstract

In this article, the authors assess the methodological reliability of big data processing in sociological research. The authors compare sten score method and cluster analysis as methods of processing the results of socio-psychological tests aimed at identifying groups of young people potentially vulnerable to drug addiction. The survey was conducted in eight universities in a city in Siberia with a large student population where 22884 students aged from 18 to 25 were questioned. First, the obtained results were processed by using the sten score method. Then, cluster analysis was conducted to define a high-risk group of students having a propensity for drug consumption. Advantages and disadvantages of the two methods for processing a large sample of data are compared. The results of this comparison demonstrate that the cluster analysis method is the most appropriate method for this type of research as it produces statistically correct data. The use of cluster analysis makes it possible to work with any type of information, both qualitative and qualitative data. On the other hand, the sten scores method can only be applied in certain conditions, i.e. where the original distribution resembles a normal distribution; where some theoretical basis to expect normal distribution exists, and where there is certainty that the normalization group is sufficiently large and representative to be a true reflection of the population.