Abstract:

The presentation considers the problem of the academic heterogeneity among universities freshmen. This problem is widely discussed in circumstances of demographic decline when higher education becomes widespread. As a result differentiation of universities based on the quality of students' freshmen arises: some universities admit predominately high-scored students, while the others have a large variation in students’ preparation level. Universities have to work with non-uniform contingent of students.
In Russia the Unified State Exam (USE) is used to admit students to the universities. There are two obligatory exams (Mathematics and Russian language) and different optional exams which students can choose depending on their future direction of study. Scores for USE are the only basis for students’ admission to universities.
The points of interest are simultaneously the quality of admission and its heterogeneity. Heterogeneity implies differences in the student’s USE scores in one university. The data base includes USE scores of all students enrolled in all universities in 2011-2012 in the context of different majors. The procedure of academic heterogeneity investigation has been developed with the 5-points scale of heterogeneity levels.
There are two types of issues in this research. The first one includes the analysis of academic heterogeneity indicators over years and the classification of universities based on their heterogeneity. The second issue concerns intra-university level and includes the following points:

• Identification of factors that influence academic heterogeneity of enrolled students
• Analysis of heterogeneity in the context of different majors
• Analysis of the dynamics of academic heterogeneity over the years
• The university contribution to the dynamics of heterogeneity during the period of education.
• Recommendations for dealing with heterogeneous contingent of students.

For assessing student’s heterogeneity traditional statistical methods were applied as well as mathematical modeling based on Item Response Theory (IRT). The use of IRT methods allows us to place all the data on the unique metric scale, construct classification of universities on the quality of admission and assess heterogeneity.
Despite the differences in the systems of state exams and the academic selectivity in various countries the methods developed can be used to assess the heterogeneity of university freshmen in each country.

Keywords:

Academic heterogeneity, freshmen, education research.