Since 2012 when first steps to putting together our meta-ranking were taken, the quite demanding process of unifying the different names of institutions and linking all three rankings has actually become a learning process, where most links elaborated in previous years serve for updating the Cross Ranking data set in the following year. We have linked score of individual higher educa¬tion institutions attained in all three rankings in order to get substantially more comprehensive data about each institution that correspond to aggregate data from all three rankings. For 2019, data on a total of 1 721 universities were identified and linked together.
After linking institutions from all three 2019 rankings and putting together the unified Cross Ranking data set, it has been necessary to define indicators that will express how successful an institution is in a given ranking. What has been used most often, is its final position in individual rankings. However, this approach has a number of serious imperfections and weaknesses, and we have decided to use all scores of each ranked institution instead, both the overall score and the partial scores in individual dimensions. Because the ARWU and the QS do not publish the overall score for all ranked institutions (only for the TOP 100 institutions), in some cases we have had to calculate our own overall score, basing it on the published dimensional scores, always in accordance with the methodology valid for a given ranking.
We have summed up the overall score of each of the three rankings and weighted it so that the weight of each of the three rankings – that is the sum of all distributed scores of the top 1 000 institutions in each ranking – would be the same (the number of l 000 institutions published in the ARWU 2019 and the QS 2019 is the lowest of rankings considered). The CR overall score of institutions has been computed as a weighted average of the score achieved in the rankings ARWU, QS a THE, with the theoretic maximum value equal to 100.
Each higher education institution, included in at least one of the three rankings considered, has been thus assigned a certain score composed of 1 to 3 overall scores and of 5 to 17 dimensional scores that constitute its profiles (the number of dimensions depending on the number of ratings). For example, Charles University ranked in all three rankings in 2019, and therefore its profiles composed of 3 overall and 17 dimensional scores. In contrast, Brno University of Technology ranked just in the QS 2019 and the THE 2019, and therefore its profile has only 2 overall and 11 dimensional scores, while University of Economics in Prague ranked in the THE 2019 only, and its profile has just 1 overall and 5 dimensional scores.
Employing this approach and the results of all three 2019 rankings (ARWU, QS, THE), it has been possible to assign the CR overall score to 1 721 best world higher education institutions and to rank them accordingly. They represent more than 9 % of the total number of about 20 thousand higher education institutions around the world (exactly 18 500 higher education institutions in 186 countries at the beginning of 2018 according to the world's largest World Higher Education Database, International Association of Universities). The ratio for 579 European universities included in the Cross Ranking is higher, about 14 %.
After linking institutions from all three 2019 rankings and putting together the unified Cross Ranking data set, it has been necessary to define indicators that will express how successful an institution is in a given ranking. What has been used most often, is its final position in individual rankings. However, this approach has a number of serious imperfections and weaknesses, and we have decided to use all scores of each ranked institution instead, both the overall score and the partial scores in individual dimensions. Because the ARWU and the QS do not publish the overall score for all ranked institutions (only for the TOP 100 institutions), in some cases we have had to calculate our own overall score, basing it on the published dimensional scores, always in accordance with the methodology valid for a given ranking.
We have summed up the overall score of each of the three rankings and weighted it so that the weight of each of the three rankings – that is the sum of all distributed scores of the top 1 000 institutions in each ranking – would be the same (the number of l 000 institutions published in the ARWU 2019 and the QS 2019 is the lowest of rankings considered). The CR overall score of institutions has been computed as a weighted average of the score achieved in the rankings ARWU, QS a THE, with the theoretic maximum value equal to 100.
Each higher education institution, included in at least one of the three rankings considered, has been thus assigned a certain score composed of 1 to 3 overall scores and of 5 to 17 dimensional scores that constitute its profiles (the number of dimensions depending on the number of ratings). For example, Charles University ranked in all three rankings in 2019, and therefore its profiles composed of 3 overall and 17 dimensional scores. In contrast, Brno University of Technology ranked just in the QS 2019 and the THE 2019, and therefore its profile has only 2 overall and 11 dimensional scores, while University of Economics in Prague ranked in the THE 2019 only, and its profile has just 1 overall and 5 dimensional scores.
Employing this approach and the results of all three 2019 rankings (ARWU, QS, THE), it has been possible to assign the CR overall score to 1 721 best world higher education institutions and to rank them accordingly. They represent more than 9 % of the total number of about 20 thousand higher education institutions around the world (exactly 18 500 higher education institutions in 186 countries at the beginning of 2018 according to the world's largest World Higher Education Database, International Association of Universities). The ratio for 579 European universities included in the Cross Ranking is higher, about 14 %.