Engineering schools ranking methodology.

Selectivity

Goal : this criterion shows the average level of the school in the selection of its students, in proportion to these different access paths (post-prep, post-bac). It is based on the access path on the average for the French baccalaureate or the average rank for national competitive exams.

Sources : SCEI statistics, CGE survey and CTI survey.

Detailed calculation :

  • For students coming from Parcoursup or from post-baccalaureate applications, the grade is the average grade of the French baccalaureate;
  • For students who have passed the post-prep exams, the grade is the average integration grade;
  • In order to harmonize the marks obtained in the different exams, the average grade of the integrated students of a competitive examination is projected on the average obtained in the baccalaureate by the students of the schools of this examination;
  • On this projection, the method of least squares makes it possible to extract a mark (comparable to the baccalaureate average) for each of the access routes (different competitive examinations, different integration courses, etc.);
  • The marks are finally averaged in proportion to the number of integrated students according to the different access routes;
  • The average obtained reveals the average level of integrated students.

Note: integrations by university admission route are not taken into account because there is too little data available on these integrations.

Employability

Goal : this criterion shows the value of the diploma for companies in the sectors targeted by the school. It is based on the median salary of young graduates treated according to the output sectors.

Exemple : AgroParisTech, which places the majority of its students in agricultural industry and agriculture, has a significantly lower exit salary than ENSAE Paris, which is particularly targeted at the banking and insurance sector, even though these two schools are benchmarks in their respective fields. Thus, the method of calculation makes it possible to harmonize these differences due to the output sector.

Sources : CGE survey and CTI data.

Detailed calculations :

  • Firstly, the average wage per sector is determined by a least-squares method derived from the identification for automatic, allowing this method to be applied for a multi-dimensional system. The vector of wages per school is denoted b, the distribution matrix of output sectors is denoted A and its transpose At. Thus, the theoretical wage per sector is obtained by doing (At*b)/(At*A);
  • Then, the theoretical wage that a school should have at the end of the school leaving is determined according to its outlets and the average wages of these different sectors;
  • The final grade is obtained by the difference between the median wage actually found at school leaving and the theoretical wage calculated previously.

Research

Goal : this criterion shows the interest in research of students when they leave school. It is based on the proportion of young graduates continuing their thesis studies treated according to exit sectors.

Exemple : chemistry schools such as Chimie ParisTech, naturally send their students to research, however schools such as Centrale Nantes are references in research for other sectors such as mechanics. Thus, the calculation method allows to harmonize these differences due to the output sector.

Sources : CGE survey and CTI data.

Detailed calculations :

  • Firstly, the average proportion of thesis pursuit per sector is determined by a least squares method derived from the identification for automatic, allowing this method to be applied for a multi-dimensional system. The vector of average proportions per school is denoted b, the distribution matrix of output sectors is denoted A and its transposed At. Thus, the theoretical proportion per sector is obtained by doing (At*b)/(At*A);
  • Then, the average proportion that a school should have in terms of output is determined according to its outlets and the average proportions of these different sectors;
  • The final grade is obtained by the difference between the actual proportion of school leavers and the theoretical proportion calculated previously.

International

Goal : this criterion shows the interest developed in the school by foreign companies, universities and students. 5 rates are used to calculate this criterion, which makes it possible (i) to avoid valuing a particular school policy (tour on foreign recruitment, double diplomas or other) and (ii) to avoid valuing a school by its size.

Sources : CTI data.

Detailed calculation : the criterion results from the average of 5 rates

  • Proportion of students recruited on foreign courses;
  • Proportion of students entering the school with a foreign double degree;
  • Proportion of students obtaining a foreign double degree while attending the school;
  • Proportion of graduates with a foreign nationality;
  • Proportion of graduates who took up employment abroad.

Alumni

Goal : this criterion shows the stability of the school through the academic or professional success of its former students.

Sources : official websites of the different awards and CAC40 companies' websites.

Detailed calculation :

  • For each of the CAC40 CEOs since 2000, the diplomas are raised and confer a point to each school;
  • For each of the members of the CAC40 management committees (as of February 2020), the diplomas are raised and give one point to each school (thus the CEO is counted twice);
  • For all selected research prizes (Nobel Prize, Wolf Prize, Abel Prize, Turing Prize, Crafoord Prize, Vestlesen Prize, Fields Medal and CNRS Gold Medal) one point is awarded to each of the recipient's schools;
  • The Alumni grade is obtained by dividing the sum of the points by the number of engineers graduating per class.