Computational Thinking – What Is It?

Bonnie BraceyBy Bonnie Bracey Sutton
Editor, Policy Issues

I attended the first CS4HS High School Teacher Workshop: Computational Thinking and Computational Doing from June 25-27, 2010, at the ATLAS Institute on the University of Colorado at Boulder campus. (The conference was featured in the Boulder DailyCamera.)

First, a bit of background. I have been working in advocacy for STEM and related technologies, lurking around the edges of computational science for some time thinking about ways in which to incorporate new kinds of thinking for students in our schools. I have attended the leading Supercomputing Conferences and brought teams to the events to try to change teaching and learning so that computational thinking, with games and simulations, could find a prominent place in the forefront of those inserting STEM into the curriculum. However, I am not sure the conference is always happy with the outcomes of teacher participation since it’s difficult to gauge the longterm effects of what happens in the classroom and some presenters don’t have a very high regard for the ability of teachers to work with technology.

But this CS4HS workshop in Boulder was different. Its focus was on teachers, and it was par excellence!

So, what is computational thinking?

I hesitate to define it because I have been sitting through lots of workshops where people dance around the definition. But I will be bold and choose one of the definitions I like. In an ITEST small group discussion, of which I was a part, we came up with: Computational thinking is foundational. It is imperative that today’s youth learn to use computers for mathematical modeling and quantitative analysis to organize and make sense of the vast amounts of information they receive from multiple sources.

Jeanette Wing, the current director of CISE at the National Science Foundation, offers this definition: “Computational thinking is a way humans solve problems; it is not trying to get humans to think like computers. Computers are dull and boring; humans are clever and imaginative. We humans make computers exciting. Equipped with computing devices, we use our cleverness to tackle problems we would not dare take on before the age of computing and build systems with functionality limited only by our imaginations” (“Computational Thinking,Communications of the ACM, 2006).

At many of the conferences I’ve attended, one of the difficulties I’ve encountered is researchers and technology professionals, people working with K-12 teachers, telling me to bypass the teachers and go straight to the students. In other words, ignore the digital immigrants and work directly with the digital natives. They have also told me that teachers are impossible to teach and that we should not waste our time with them, that our time would be better spent with the students. Some were actually insulting when talking about teachers. Well, I understand that most teachers have not had in-depth professional development training in the computer sciences, but I also know that they’re dedicated to keeping up with the latest trends and are excellent learners.

Thus, it was a pleasure to see how easily Dr. Alexander Repenning had us, as teachers, collaborating and taking turns leading the lessons, teaching and learning from one another. We moved from creating games to simulations with little or no effort. (Click here to view workshop photos on my FaceBook page.) I probably would not have believed that I could do this based on the attitude that some technology professionals seem to have toward teachers.

Dr. Repenning, in an email message, said, “Unfortunately, a precise definition of computational thinking still eludes us.” However, he had this to offer: “We can actually measure [computational thinking] patterns in games and simulations using pattern recognition algorithms similar to Latent Semantic Analysis. When students submit their games/simulations we can instantly compute these pattern and provide feedback to students and teachers. This kind of feedback could recognize existing computational thinking skills, suggest more challenging projects, and potentially even document transfer of skills.”

In the same message, he asked for feedback on: Towards the Automatic Recognition of Computational Thinking for Adaptive Visual Language Learning by Kyu Han Koh, Ashok Basawapatna, Vicki Bennett, and Alexander Repenning (to appear in Proceedings of the 2010 Conference on Visual Languages and Human Centric Computing [VL/HCC 2010], IEEE Computer, Madrid, Spain).

Please review the paper and share your comments with me (as replies to this article), and I’ll share them with Dr. Repenning.

5 Responses

  1. I’ll begin by admitting that I have never before read the term “computational thinking.” Parsing the words doesn’t help me at all. Jeanette Wing’s definition helps a little, but seems more like just thinking to me. Kyu Han Koh et al. seem to agree with the fuzziness of this term.

    Regarding the paper with link provided, I immediately found another definition issue in the Abstract. What is “computer science?” Is there a science of computers? Setting aside my concerns that this commonly used term is a bit of an oxymoron, I’m still left wondering what it means to make “computer science” more accessible to students. Perhaps, the authors really intended making software engineering more accessible. This difference in terms may seem like nit-picking, but it illuminates a truly important distinction between science and engineering.

    Since the first assembly language and assembler were created, the trend has been toward ever-higher level computer languages, those arcane structures with difficult syntax that we use to instruct computers to do something useful to us.

    At each new level, the programmer moves farther away from the computer hardware, both a good and a bad thing. It’s good for two very important reasons: programs may be somewhat independent of the hardware and so survive the change to a new generation of hardware, and the number of lines of software “code” to be written for a given project becomes smaller and so is created faster while also becoming more manageable. It’s bad because programmers lose all contact with the underlying structures upon which their programs are built and within which they must run. Code “bloat” and other problems such as memory leaks can occur as a result of this disconnect.

    Visual programming tools represent the latest in the progression of ever higher level languages. For software developers, they have been represented for many years by the IDE (integrated development environment), a generally drag-and-drop visual interface for creating software. In my experience, the output of the IDE tends to be inefficient and unusable except within the IDE. That is, the resultant software program looks unstructured and presents serious maintenance problems should anyone attempt to work outside of the IDE. Of course, the IDE vendors like that aspect.

    As to the question of how effective using visual programming tools is in creating some sort of higher-order thinking, I propose that the answer is as expected. The greater engagement, in my opinion, stems from creating something and even more so from creating it using fun, technology-based tools and creating something “neat” or “cool.”

    The lack of transferable skills (computational thinking or higher-order thinking) cannot be a surprise if the visual programming tool provides a simple enough interface to let any middle school use it successfully.

    Unfortunately, software engineering skills require years to develop well. For some people, the field shows total opacity; they just cannot do it. I believe that you’re asking unfairly that some visual programming tool be used by middle school students and that it develops transferable skills of a higher order.

    It’s like learning to drive with an automatic shift and then being expected to drive a manual shift car. This approach works well as long as you’re not expected to drive a manual shift car.

    if you make the design process sufficiently realistic, then many, if not all, middle schools students will be unable to use the tool. Indeed, most non-programmers will have the same problem. They can create usable games or other programs with the simple tool but still cannot design and build more complex software products with any tool.

    I really must ask the obvious question. How important is learning “computational thinking,” especially when the term has such an unclear definition? If you can find a decent definition and justify the value, then why should playing with a visual development tool be the best way to promote this learning? As far as anyone seems to know, it could even be the worst.

    All of this discussion ignores the purpose of the paper: evaluating CTP graphs in educational settings. Until someone answers the more basic questions, the CTP question appears to me to be moot, or perhaps the conclusions are a bit too abstract for me to understand fully.

    The use of visual programming tools appears to be a great way to introduce students to a few of the basic concepts of programming or software engineering. In particular, students should learn that computers are really, really stupid but are also very, very picky. A computer is a very literal slave that understands only certain commands.

    It also introduces students to another very important idea: creating things, even electronically, can be loads of fun. The high percentages of students interested in continuing the course illustrates that outcome. Just imagine if the tool allowed for the creation of World of Warcraft. On the other hand, the number of students who answered negatively suggests that yet another important lesson may have been imparted: writing software takes lots of time and attention to detail.

  2. To get students involved with thinking about computers, probably the best starting points are Scratch (for elementary and middle school students) and Python for high school students. My friend Janet Lathan has encountered a lot of success teaching Python in high school here in Washington DC. She uses the free book Snake Wrangling for Kids: Learning to Program With Python.

    To help students develop a great sensibility to the power and uses of computers in society as a whole, have them monitor the news. The supercomputer prediction of the BP oil spill reaching the Atlantic Ocean is a prime example.

    To my mind, computational thinking sounds like something that can be reached only when students use computers beyond the very narrow ways computers are often used in schools. We must explore ways of using computers in outside of school settings so that the power of informal learning can steer the kinds of discoveries we want students to be experiencing. Your mind can’t be making discoveries with the rigid time structure of the modern day school. Discoveries don’t happen on schedule.

  3. You asked this question

    I really must ask the obvious question. How important is learning “computational thinking,” especially when the term has such an unclear definition?

    I loved this comment. I should have started at the base level of thinking in STEM which I call .. problem based learning and innovation.

    THINKING if you will.

    REFLECTION if you will.

    UNDERSTANDING if you please.

    I cannot steal Alexander’s slope idea.. but I love that too,
    Learning that lead to the next steps with facility.
    Not mindless memorization without understanding.

    Here is what I am replying to.

    In the elementary, and middle schools we have still the old math and reading of science.. I mean the memorization of facts and ideas that have been around since my mother started teaching. She taught like in NCLB. Teach and test, teach and test, but what kind of test? What is the tranceference of learning ?

    There was a game in which we constructed bicycles, cars, Ariel balloons and other devices and we had to use reasoning, math and problem solving to make the device work, but then not just to work but to win in a race with other devices of the same kind.

    The kids who were good at memorization and usually the ones who answer question were stumped. We found new mind muscles in the group. We learned to collaborate in our learning community. We learned to have conversations about this information we needed to know. I was always learning along with the kids, most of the time but that was fine. They liked that. Harris Libergot at NIST was furnishing examples and ideas to me.

    We worked with the Challenger Center in solving problems in the Challenger Space Center. One of our school based projects was Marsville, and other classes wanted to do it. Marsville was the solving of problems based on an imaginary colony on Mars and we had to build systems , solve problems and create ways to make the colony habitable. My students loved this one so much we also did Mars CIty Alpha, designed for Middle Schools.

    I had a secret weapon.

    I was allowed to have a Wednesday and a Saturday Computer Class at the Arlington Career Center.. We did lots of things. We build small computers, we learned to understand simple robotics ( we were working in a high school that had robotic arms, we had industrial robotics that we could program , and I wrote some grants and got some Lego robotics.. but the most important part of the grant was getting Lego’s to teach physics ( that was for the third grade) That was so awesome to be able to do. The hardest thing was writing the grants to support such a program. The National Geographic had a video on robotics and we programmed , wrote programs, this was before the easy programs came for students. We had basic.

    Problem solving to me is very different. Math to me when thinking is involved is very different. Say a kid of any age being able to understand the concepts instead of merely prattling off the numbers. I remember a fourth grade class I had when we were able to do new math, and when I had Cuisinaire rods, a prime number game and other ways of teaching math but to do the games , they had to know the concepts to get the right answers. It was not me, the games were from MECC and from MIT. I was working with Seymour Papert and turtle math through the NEA NFIE.

    The kids would want to skip lunch upstairs to play the games. I don’t remember when they became adept at

    Those children made such high scores the school made them take the tests over but we did not cheat. The games and exercises in problem solving allowed them to transfer the ideas of multiplication, division, and they were careful in their execution of problem solving because they understood that if the process being followed to solve the problem was not accurate they would
    not get the right answer.

    Memorization did not work because the games asked about the conceptual framework, How many ways can you multiply to get this number. We did not call it computational thinking, What the games did was to allow them to do it step by step or, vault to the next level of understanding by demonstration of the knowledge.

    The site” Interactivate “comes to mind. Here is the link

    In my early teaching years they gave us beans to use to try to solve the problem. I don’t have a name for this except understanding math.

    Once a child in my room, exploring base 12, said to me, I get it and he started to show me all of the bases ( ok I will admit that I was not able to see or understand it in the way in which he did, until later on that night. I was using the book and I didn’t have to know anything) So I went back to school to do math that summer, but REAL MATH not math for teachers.
    I suppose for me it was thinking, being able to understand comceptual frameworks or ideational scaffolding as furnished here.;

    Also look at this. Modeling And Simulation Tools for Education Reform
    MASTER Tools, developed by The Shodor Education Foundation, Inc. are the result of on-going collaborations with the National Center for Supercomputing Applications (NCSA), George Mason University, and other education organizations. They are designed to be interactive tools and simulation environments that enable and encourage exploration and discovery through observation, conjecture, and modeling activities. This is what I am talking about.

    MASTER tools will soon be fully integrated with new collaboration tools and online research facilities to create an authentic scientific experience.

    This information is from Dr. Robert Panoff , and there is more because he teaches NCSI courses in education for teachers. I have some examples of the workshops he does as well on my facebook page.

    In Alien Contact!, students use GPS-enable Dell Axim handheld computers that correlate their real world locations to their virtual locations in the game’s digital world. As the students move around a physical location, such as their school playground or sports fields, a map on their handheld displays digital objects and virtual people who exist in an augmented reality world superimposed on real space. This capability parallels the new means of information gathering, communication, and expression made possible by emerging interactive media (such as web-enabled, GPS-equipped cell phones with text messaging, video, and camera features).

    Alien Contact’s mission is to challenge teams of students to solve problems through proportional reasoning, justify their thinking with peers, and use technology as a means of representation and interpretation. In addition, the simulation also puts math into a more comprehensible format than writing an equation on a chalkboard might.

    “I felt excited about the possibility of taking technology that students are intrigued by and fluent in and using it to teach mathematics in a way that simulates and enhances reality,” says doctoral student Rebecca Mitchell, a former math teacher who works on developing the game’s curriculum.

    “Students can really act as though they are CIA agents trying to figure out why aliens have been found on earth through augmented reality. At the same time, I wanted to make sure the curriculum is created carefully, as too often I have encountered curriculum that was showy or exciting but with little quality in terms of mathematics instruction.”

    Dr. Chris Dede’s old project is Alien Contact which is another example of what I am talking about.

    Alien Contact! has been piloted in two Massachusetts schools so far. Based upon preliminary data, AR is a highly engaging environment for students who struggle within the limited instructional approach of a traditional classroom. The issue of engagement isn’t the biggest concern, Dunleavy says. “Alien Contact! is driven by pop culture…. We thought about what kids were interested in and designed around these topics,” he says. “The challenge is figuring out how to leverage the technology using good pedagogy that will translate into meaningful learning.”

    Instructional materials are currently available for GalaxSee, SimSurface, and the Fractal Microscope. We also have a beginning collection of models and materials in medicine and biosciences, and environmental science.
    These are thinking tools.

    Other thinking tools are being introduced by Dr . Chris Dede who came to my 4th grade class( and I to his college teachers and technology class)

    Dr. Dede has gone on to Harvard. His ideas are being developed with tools of imbedded assessment.

    At a suburban Boston middle school, aliens have landed. A team of seventh graders armed with Global Positioning Systems (GPS) and handheld computers wander the school field trying to figure out why aliens are there. However, the path to finding the right answer is getting a bit challenging.

    “It’s asking us to do algebra,” says one boy to another.

    Solving math problems, translating Latin and Greek, interviewing virtual characters, and working with other students are all aspects of Alien Contact! — the augmented reality (AR) game that the students are playing. The game is the brainchild of the Handheld Augmented Reality Project (HARP) — developed by the Ed School and the Teacher Education Program at M.I.T. with funding from the U.S. Department of Education’s Star Schools Program grant — an effort to improve middle school mathematics and literacy by leveraging emerging technologies. HARP is one of the first university projects to study whether augmented reality — a simulated real world environment through a handheld computer — can enhance students’ learning.

    At Berkeley I took Astronomy. We also did emerging technology work, not just the same old same old astronomy. It was a symposium that the NSTA offered. The part was that we had the lesson in the morning and the discussion in the afternoon and then we would get tasked with teaching the lesson to our peers.
    I think in technology today, we trust the vendors to tell us that what we are teaching works.

    You said

    I really must ask the obvious question. How important is learning “computational thinking,” especially when the term has such an unclear definition?

    I enjoyed sitting around at a two day Computational Science workshop to hear the Pi’s stumble around with how to define” computational thinking”
    but i think if we wait for the term to be defined well to all, or to everyone’s satisfaction it will be years before we decide as politics, and pundits, and personalities as well as pyschometricians actually made the definitions after awhile.

    The use of visual programming tools appears to be a great way to introduce students to a few of the basic concepts of programming or software engineering. In particular, students should learn that computers are really, really stupid but are also very, very picky. A computer is a very literal slave that understands only certain commands.

    Hsving just come from the ISTE conference, with games, gadgets and gizmos
    I prefer the computer slave to the clickers and the white board , but can even use them with proper teacher preparation and development of the use of how to transform education.

    Here is where Chris’s work is today.

    VR’s Frames of Reference: A Visualization Technique for Mastering Abstract Information Spaces

    Marilyn C. Salzman, Chris Dede, R. Bowen Loftin


    This paper describes a research study that investigated how designers can use frames of reference (egocentric, exocentric, and a combination of the two) to support the mastery of complex multidimensional information. The primary focus of this study was the relationship between FORs and mastery; the secondary focus was on other factors (individual characteristics and interaction experience) that were likely to influence the relationship between FORs and mastery. This study’s outcomes (1) clarify how FORs work in conjunction with other factors in shaping mastery, (2) highlight strengths and weaknesses of different FORs, (3) demonstrate the benefits of providing multiple FORs, and (4) provide the basis for our recommendations to HCI researchers and designers.

    Chris was at the NSF and I will admit that I took the workshops there and that I have followed him and his ideas since I first met him. He, like Alexander, gives me hope that teaching and the professional development of teachers will be changed to create transformational education.

    Ok, I am just a teacher. I have asked my network of friends to help us with this discussion. I think I am a learner in this journey. I do know that I have been able to broaden engagement and to bring in students who hated school through the use of innovation, personalization and real teaching and learning.

    Bonnie Bracey Sutton

  4. There is nothing particularly mysterious about computational thinking and if you read NSF program announcements you get some idea of what is meant. It basically means the skills to apply computation to solve scientific problems – so it includes old topics such as numerical solutions to PDEs, new topics such as data driven science, and the application of standard computer science/mathematical problem solving strategies such as divide and conquer (recursion), iteration, the use of finite automata modeling and more exotic strategies such as model checking, tree construction, and substring matching applied to biological sciences. It’s more important now than it was 20 years ago because there is a lot of application specific programming environments that scientists can directly interact with and because of the recognition that abstractions of computational and data structures cross disciplines – e.g. every discipline is computational in some sense so paradigms studied by computer scientists may apply to other disciplines.

    • Perhaps we can summarize the situation as follows:

      1. Computer scientists and programmers seem to be in consensus about what computational thinking is and why it is important.

      2. Educators are very unclear about what computational thinking is, although they do believe it’s an important skill for the 21st century.

      3. There needs to be more training and communication between the scientists and educators on the core concepts and practical applications of computational thinking in K-12 schools.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: