Communications of the ACM

Enhancing student learning and understanding by combining theories of learning with the computer's unique attributes.

Alexander Repenning

Computational Thinking in Public Schools

I share some of the frustration Cooper et al. express in their article with the lack of clear definition of computational thinking that is actionable for public schools teachers especially at the middle school level. In the context of the NSF ITEST funded Scalable Game Design project we have explored notions of computational thinking, i.e., what is it, how can we teach it, how can we assess it, how does it transfer between domains such as game design and science modeling... One reaction common between the teachers from several states is that the term computational thinking feels right and is much less threatening than previous notions such as programming. Some teachers indicate that they got burned by programming. By burned they typically referred to elective courses using traditional programming approaches - some of the teachers had Logo experience - that were considered hard and boring by students and had extremely low participation of girls and minorities. Moreover, the collateral benefits that many had hoped for of programming never materialized. One teacher commented on the removal of programming from the curriculum at the time we did not see the benefits of programming to STEM and programming simply was not on the test.

The positive response of teachers to the term computational thinking makes me at least question the need to come up with yet another term. The main problem appears to be not the term and possible positive/negative connotations it may carry but the need to come up with some concrete frameworks that help teachers to do their daily work in public schools and that have some bearing on the collateral impact of computation onto public school curriculum. While many of us are working towards, or at least waiting for, a more tangible definition of computational thinking we have found that district leaders may not have a concrete sense of how computational thinking is defined but they do have a very concrete expectation of computational thinking. This concrete expectation is important in order to avoid making the same mistake twice. One instructional technology district director summarizes the situation succinctly by stating a concrete, testable expectation: at one point I want to be able to walk up to a student participating in game design and ask now that you can make Space Invaders, can you make a science simulation? We have been taking this idea of transfer very seriously and have looked at what some of the abstractions of computational thinking may be and if we could find a way to measure them. The early results are quite exciting by providing early indicators that we can not only teach game design based on these abstractions but that the same abstractions indeed do cary over to building computational science models in scientifically measurable ways.

In summary I think we should stay away from the idea to side step the problem we already have by introducing yet another term. It would appear to be more promising to stick the term computational thinking due to the level of comfort that some teachers have already reached with it. However, while we are increasing efforts to define what the term really means we should perhaps think even more about the pragmatic expectations for public school education. How do we need to structure curriculum so that computer science education becomes more appealing to a larger audience and, at the same time, how can we make sure that this time there is some collateral benefit, perhaps in the form of transfer, from computer science courses to other STEM related classes?

Information about the Scalable Game Design project including papers on measuring transfer of computational thinking: http://scalablegamedesign.cs.colorado.edu