The quality of education: using data-mining of the centralized testing results

Abstract

The paper presents results of research of entrants and students quality by means of date-mining. The authors introduce a new quantitative indicator of a university entrant quality. The indicator is derived from results of the centralized testing by using the principal components and classification analysis. The quality indicator is considered as a systematic knowledge indicator. The results were used in marketing of the university. It is shown how an effect of the direct mailing was exposed and how the mail efficiency was estimated. The target group was defined a priory by using a systematic knowledge indicator. Bootstrapping was used to reveal the direct mailing effect from a superposition of other marketing efforts towards the university entrants and to validate the effect. The efficiency of the applied method to enroll the target group of entrants was equal. Specific cost to attract the target entrant by the direct mail was calculated.