The real problem with computer science education

A Popular Myth in Computer Science Education

Informatics educators expend more energy tackling education issues than educators in almost every other field. I believe much of this effort is unnecessary and could be eliminated if informatics educators could agree on a base set of computing principles.

In other fields, researchers are not so busy in discussing curricula, roles, skills and professional profiles. Informatics appears as a rapidly changing discipline and even with all the attention to education the results are not up to expectations. For instance, the Royal Society talks about the “vicious circle” of modern computer science education. It highlights how inadequate courses teaching computing in schools may result in negative perceptions of computing, in turn putting young people off the further study and resulting in fewer computing graduates and thus fewer specialists teaching the subject in schools.

Several commentators ascribe the responsibility of the difficulties encountered in computer science education to the subject matter. The following—extracted from the ACM-IEEE Computing Curricula 2001—summarizes an amply shared viewpoint: “Computing has changed dramatically over that time in ways that have a profound effect on curriculum design and pedagogy.”

I will summarize what I have heard from many educators in Information and Communications Technology (ICT):

“Informatics is a novel and fast evolving discipline. Hence, the pedagogy of informatics meets severe obstacles.” (1)

This common belief seems a dogma rather than a statement sustained by logical reasoning. The first part (i.e. “Informatics is a novel and fast evolving discipline”) is established on the basis of facts and is true beyond any doubt. However the conclusion (i.e. “hence the pedagogy of informatics meets severe obstacles”) is not so evident.

Let us look at the electronics  sector, which is also fast evolving but is not experiencing pedagogy problems.

The advances in the development of the electronic market appears remarkable even to common people. It may be said that electronics drag the advancement of computer science. But despite its rapid growth and evolution, the study of electrical devices does not result in pedagogical issues similar to those occurring in computing. Usually, the electronics courses begin with fundamental principles–say Faraday’s law of induction, Kirchhoff’s equations, the Ohm’s law, the Maxwell’s formulas etc.—and proceed toward specialized topics.

Electronics do not give rise to severe educational problems regarding the subject contents. Teachers are not so stressed to “find a way to present” a certain topic. Students do not undergo disorientation and a sense of groundlessness so frequent in computing. They progressively enrich their culture and arrive at precise professional competencies.

The Association for Computing Machinery (ACM), the Institute of Electrical and Electronics Engineers (IEEE), and other organizations published several reports for experts who prepare courses on computing. This vivid activity is absent in the electronic domain. No professional organization felt the need to promote the publication of curricula in electronics similar to those cited above. Teachers arrange and optimize the lessons in electronics as technology advances without the need of special guidelines.

The reader perhaps objects that informatics has developed so much that now we face different sub-disciplines. The ACM/IEEE Curricula in 2001 holds: “The scope of what we call computing has broadened to the point that it is difficult to define it as a single discipline.”

One could reply that electronic experts have inaugurated several new areas such as nanoelectronics, photonics, robotics, power electronics, quantum electronics, and spintronics, which do not subvert the didactical curricula. New specialist subject matters substitute the obsolete ones or are appended at the bottom of the pedagogical pathways without great discussion. The new topics give substance to new professional figures and do not raise lively debates comparable to the discussion occurring in informatics.

I think the reason behind this is that computing has not yet established an agreement on the base principles of the field. We lack agreements on what our versions of Ohm’s law and Kirchoff’s laws are. I speculate the formation and execution of computing curricula would be much easier if we discussed computing principles and came to some agreements on the laws that affect computation from algorithms to architectures.

Concluding, the case of electronics makes evident how advanced and rapidly evolving technology does not oppose necessarily dramatic pedagogical difficulties. Statement (1) sounds like a generic and untrustworthy myth. Statement (1) is false and results in repeated and prolonged misunderstandings.

The case of electronics illustrates the facilities offered by the principles of a discipline to the education of that discipline. Are we aware of the importance and role of the principles of computing could take? And is the scientific community undertaking sufficient efforts to develop and apply the principles of computing?


  • alfredthompson

    I think there is a difference between pedagogy issues and curriculum topics. Topics do change. For example the CS 2013 curriculum guidelines are very different from the ones in 2001. However how we teach those topics is largely a different question. We as a community spend a lot of time on what to teach but not much on how to teach. The latter is where we need a lot of work. Ideally research backed work but most education departments feel that is a problem for CS while most CS departments are more concerned with other research questions. It’s a problem.

  • TGL

    Who says computer science is a science? The closest I can come to defining CS is to compare it with economics vs physics. Economics isn’t a science either, although it uses math. Both are collections of techniques found to be useful in other disciplines. Hence, the difficulty with teaching it. Not only do topics change, the underlying ideas change.

    • Paolo Rocchi

      Economics is not a science like
      physics or chemistry for sure but experts use math and make efforts to establish
      the scientific aspects of economics. They have fixed some fundamentals such as the
      law of supply and demand. Also CS should be defined a science and not a mere
      collection of techniques in that researchers mean to establish the basic properties
      of systems which are typical of systems and do not vary with time, fashions and
      social needs. For instance, Shannon’s entropy qualifies the properties of transmission
      channels that are independent of human habits and will.

      Obviously a discipline – say CS – is called to support a variety of activity sectors
      but the scientific profile of CS does not vary with applications. The stability
      of CS is the virtue that aids people to tackle ever new situations with success.