Improving a Multifactor Regression Model for Assessing the Work Performance of Faculty Members in Higher Education

This article proposes an improved multifactor regression model for assessing the work performance of faculty members in higher education. The rapid expansion of higher education in Uzbekistan during 2023–2025 is analysed, and the main specification problems of conventional linear performance models are demonstrated using official regional data for the 2025/2026 academic year: the dominance of large units, multicollinearity among scale-related variables, and non-proportional relationships. The elasticity of teaching staff numbers with respect to student enrolment across regions is estimated at 0.75, indicating that the teaching load per faculty member increases with institutional scale. On this basis, an improved model is developed that combines within-discipline standardisation of indicators, a log-linear specification, multicollinearity diagnostics, robust standard errors, and department-level random effects, together with a procedure for deriving indicator weights from estimated coefficients.