swissUp 2006 CS Ranking::a Critique

Critique

• Critique of swissUp Ranking
  (11 page report — pdf)
• Analysis of the raw data (zipped excel)

swissUp Ranking 06

• swissUp Ranking of CS
• myRanking (pick indicators)
• IAM indicators
• Raw data for IAM (zipped excel)
• info/faq page
• Students' questionnaire
• Professors' questionnaire
• Response rates
• Methodology overview

Press

• In Medizin ist die Uni Bern Spitze — Der Bund, April 7, 2006
• Human- und Zahnmedizin sowie Mathematik top — Informatik flop — Berner Zeitung, April 7, 2006
• Rankings — mit Vorsicht zu geniessen &mdash NZZ, April 11, 2006
• Hochschul-Rankings dürfen nicht überinterpretiert werden — NZZ, April 11, 2006
• Vielleicht ist das Studium zu hart — Der Bund, April 25, 2006
• Uni Bern kann mithalten — Der Bund, May 5, 2006
• SwissUp-Ranking: Informatik wehrt sich — UniLink, May, 2006

Summary

In the swissUp Ranking study 2006, 21 out of 22 third semester students stated that they would recommend Computer Science at the University of Bern to beginners starting their studies, and 80% of Computer Science Alumni positively evaluate the impact of their studies on their career.

The swissUp Ranking of Computer Science, however, presents the Institute of Computer Science at the University of Bern (IAM) as obtaining four red scores out of five criteria selected for comparison. Bern clearly comes out worst out of the nine Computer Science departments evaluated. How can the scores be so bad?

There are essentially three key problems with the way that the swissUp Ranking has been carried out that have led to this curious result:

1. Selection of swissUp indicators is misleading. By selecting other criteria for ranking, such as Career Assessment of Graduates, Research Reputation, Study Program, and PhDs per professor, Bern comes out as one of the best Computer Science departments in Switzerland.
2. Reduction of raw data to ranked indicators is misleading. Regardless of how well schools are evaluated, they are classified as “top” (green), “middle” (yellow) or “bottom” (red). If all schools obtain good scores for selected criteria, then schools at the bottom appear to be doing badly, despite their good scores.
3. Questions biased towards large schools and departments. Many indicators are inherently biased towards large schools and departments, and against small Computer Science departments. The ranking is rigged so that the strengths of smaller departments are not in evidence.

Although many of the data collected in the swissUp study are certainly of interest, the way in which these data are summarized, processed and presented is extremely misleading and potentially highly damaging to the reputation of Computer Science at the University of Bern. At a period when the demand in industry for qualified Computer Scientists and IT specialists is constantly rising, yet inexplicably the numbers of new students starting studies in Computer Science are falling, rankings of this form can only have a catastrophic effect.

For more details, please see:

• Full report — A Critique of the swissUp Ranking 2006 of Computer Science (11 pages, pdf, in English)
• Followup articles in Der Bund:
• Vielleicht ist das Studium zu hart (April 25, 2006)
• Uni Bern kann mithalten (May 5, 2006)
• Followup article in UniLink — SwissUp-Ranking: Informatik wehrt sich (May, 2006)

Prof. Dr. O. Nierstrasz
Director, Institute of Computer Science
and Applied Mathematics,
University of Bern