sabato 1 settembre 2018

World University Rankings 2018 methodology

With more data from more institutions, our 2018 rankings give a bigger picture than ever before. Here we explain the methodology that underpins the tables





The Times Higher Education World University Rankings are the only global performance tables that judge research-intensive universities across all their core missions: teaching, research, knowledge transfer and international outlook. We use 13 carefully calibrated performance indicators to provide the most comprehensive and balanced comparisons, trusted by students, academics, university leaders, industry and governments.

The performance indicators are grouped into five areas:
Teaching (the learning environment)
Research (volume, income and reputation)
Citations (research influence);
International outlook (staff, students and research)
Industry income (knowledge transfer)

Independent audit
The calculation of the Times Higher Education World University Rankings has been subject to independent audit by professional services firm PricewaterhouseCoopers (PwC).

Exclusions
Universities are excluded from the World University Rankings if they do not teach undergraduates or if their research output amounted to fewer than 1,000 articles between 2012 and 2016 (and a minimum of 150 a year). Universities can also be excluded if 80 per cent or more of their activity is exclusively in one of our 11 subject areas.

Data collection
Institutions provide and sign off their institutional data for use in the rankings. On the rare occasions when a particular data point is not provided, we enter a conservative estimate for the affected metric. By doing this, we avoid penalising an institution too harshly with a “zero” value for data that it overlooks or does not provide, but we do not reward it for withholding them.

Getting to the final result
Moving from a series of specific data points to indicators, and finally to a total score for an institution, requires us to match values that represent fundamentally different data. To do this we use a standardisation approach for each indicator, and then combine the indicators in the proportions indicated to the right.

The standardisation approach we use is based on the distribution of data within a particular indicator, where we calculate a cumulative probability function, and evaluate where a particular institution’s indicator sits within that function. A cumulative probability score of X in essence tells us that a university with random values for that indicator would fall below that score X per cent of the time.

For all indicators except for the Academic Reputation Survey we calculate the cumulative distribution function of a normal distribution using Z-scoring. For the data in the Academic Reputation Survey we use the cumulative distribution function of an exponential distribution in our calculations.

Nessun commento:

Posta un commento