WEIGHTS, STAFF NUMBERS AND CATEGORY LEADERS: HOW UZBEKISTAN’S NATIONAL RANKING SCORES RESEARCH OUTPUT

Abstract

The article evaluates the research output indicator of Uzbekistan’s national ranking of higher education institutions. The indicator combines Scopus papers in the Top 1% and Top 10% citation percentiles and in Q1-Q4 journals with weights of 12, 10, 8, 6, 4 and 1, divides the result by the number of academic staff and converts it into points relative to the category leader. Using official data for 105 institutions in 2025, the study shows that the overlapping weights give one Top 1% paper in a Q1 journal 30 weight units, while 18.6% of publications have no quartile. Alternative weights change institutional ranks little: the number of publications per staff member alone correlates with official scores at ρ = 0.935. The value of one additional Q1 paper differs by up to 89 times within a category, and in the Applied and Medical categories about nine out of ten institutions receive less than 10% of the maximum score. Under the logarithmic normalisation mentioned in the official report, the median score in the Applied category would be 0.730 instead of 0.072. The article proposes counting each paper only once, setting a minimum staff threshold, using normalisation that depends less on the leader, publishing the indicator components separately, and defining the numerator and denominator on the same basis.