# SIGCSE Keynote #1 – Computational Thinking For All, Robert M. Panoff, Shodor Education Foundation

**Posted:**March 7, 2014

**Filed under:**Education |

**Tags:**bob panoff, education, higher education, keynote, learning, robert panoff, sigcse, SIGCSE2014, teaching, teaching approaches, thinking 2 Comments

Bob Panoff is the wonder of the 2014 SIGCSE Award for Outstanding Contribution to Computer Science Education and so he gets to give a keynote, which is a really good way to do it rather than delaying the award winners to the next year.

Bob kicked off with good humour, most of which I won’t be able to capture, but the subtext of hits talk is “The Power and the Peril”, which is a good start to the tricky problem of Comp thinking for all. What do we mean by computational thinking? Well, it’s not teaching programming, we can teach programming to enhance computational thinking but thinking is the key word here. (You can find his slides here: http://shodor.org/talks/cta/)

Bob has faced the same problem we all have: that of being able to work on education when your institution’s focus is research. So he went to start an independent foundation where CS Ed where such activities could be supported. Bob then started talking about expectation management, noting that satisfaction is reality divided by expectations – so if you lower your expectations. (I like that and will steal it.)

Where did the name Shodor come from? Bob asked us if we knew and then moved to put us through a story, which would answer this question. As it turns out, he name came from a student’s ungenerous pattern characterisation of Bob, whose name he couldn’t remember, as “short and kinda dorky looking”.

I need to go and look at the Shodor program in detail because they have a lawyered apprenticeship model, teaching useful thinking and applied skills, to high schoolers, which fills in the missing math and 21st century skills that might prevent them from going further in education. Many of the Shodor apprentices end up going on as first-in-family to college, which is a great achievement.

Now, when we say Computational Science Education, is it Computational (Science Education) or (Computational Science) Education? (This is the second slide in the pack). The latter talks about solving the right problem, getting the problem solved in the right way and actually being right.

Right Answer = Wrong Answer + Corrections

This is one of the key issues in modelling over finite resources, because we have to take shortcuts in most systems to produce a model that will fit. Computationally, if we have a slightly wrong answer (because of digital approximations or so on), then many iterations will make it more and more wrong. If we remember to adjust for the corrections, we can still be right. How helpful is it to have an exact integral that you can’t evaluate, especially when approximations make that exact integral exceedingly unreliable? (The size of the Universe is not 17cm, for example.)

Elegant view of science: Expectation, Observation and Reflection. What do you expect to see? What do you see? What does it actually mean Programming is a useful thought amplifier because we can get a computer to do something BUT before you get to the computer, what do you expect the code to work and how will you now what it’s doing? Verification and validation are important job skills, along with testing, QA and being able to read design documents. Why? Because then you have to be able to Expect, Observe and Reflect. Keyboard skills do not teach you any of this and some programming ‘tests’ are more keyboard skills than anything else.

(If you ever have a chance to see Bob talk, get there. He’s a great speaker and very clever and funny at the same time.)

Can we reformable the scientific method and change the way that we explain science to people? What CAN I observe? What DO I observe? How do I know that it’s right? How am I sure? Why should I care? A lot of early work was driven by wonder (Hey, that’s cool) rather than hypothesis driven (which is generally what we’re supposed to be doing.) (As a very bad grounded theorist, this appeals.)

How do we produce and evaluate models? Well, we can have an exact solution to an exact model, an exact solution to an approximate model (not real but assessable), an approximate solution to an exact model and an approximate solution to an approximate model. Some of the approximation in the model *is the computing **itself*, with human frailty thrown into the mix.

What does Computational Thinking allow you to? To build and explore a new world where new things are true and other things are false, because this new universe is interesting to us. “The purpose of computing is insight, not numbers” — R. Hamming, “If you can’t trust the numbers, you won’t get much insight” — R. Panoff. Because the computer is dumb, we have to do more work and more thinking to make up for the fast and accurate moron that does what we order it to do.

“Killing off the big lie” – every Math class you have, you see something on page 17 showing a graph and an equation which has “as you can see from the graph” starting it. Bob’s lament is that he CAN’T see from the graph and not many other people can either. We just say that but, many times, it’s a big lie. Pattern recognition and characterisation are more important than purely manipulating numbers. (All of this is on the Shodor website) Make something dynamic and interactive and student can explore, which allows them to think about what happens when they change things – set an expectation, observe and reflect, change conditions and do it again.

Going to teachers, they know that teaching mathematics is frequently teaching information repetitively with false rules so that simple assessment can be carried out. (Every histogram must have 10 bars and so many around the mean, etc) Using computing to host these sorts of problems allows us to change the world and then see what happens. Rather than worry about how long it takes students to produce one histogram on paper, they can make one in an on-line environment and play with it. There are better and worse ways to represent data so let’s use computational resources to allow everyone to do this, even when they’re learning. This all comes down to different models as well as different representations. (There is value to making kids work up a histogram by hand but there are many ways to do this and we can change the question and the support and remove the tedium of having to use paper and pen to do one, when we could use computing to do the dull stuff.)

Bob emphasised the importance of drawing pictures and telling stories, they hand-waving that communicates site complicated concepts to people. “What’s this?” “I don’t know but here comes a whole herd of them!”

The four things we need for computational thinking are: Quantitative Reasoning, Algorithm Thinking, Analogic Thinking, and Multi-scale Modelling. Bob showed an interesting example of calculating a known result when you don’t know the elements by calculating the relative masses of the Earth and Pluto using Google and just typing “mass of the earth / mass of pluto” Is this right? What is our reason for believing it? You would EXPECT things to be well-know but what do you OBSERVE? Hmm, time to REFLECT. (As the example, the earth mass value varies dramatically between sources – Google tells you where it gets the information but a little digging reveals that things don’t align AND the values may change over time. The answer varies depends upon the model you use and how you measure it. All of the small differences add up.)

The next example is the boiling point of Radium, given as 1,140C by Google, but the matching source doesn’t even agree with this! If you can’t trust the numbers then this is yet another source of uncertainty and error in our equations.

Even “=” has different interpretations – F = ma is the statement that force occurs as mass accelerates. In nRT = PV, we are saying that energy is conserved in these reactions. dR/dT = bR – the number of rabbits having bunnies will affect the rate of change of rabbits. No wonder students have trouble with what “s=3” means, on occasion. Speaking of meaning, Bob played this as an audio clip, but I attach the text here:

The missile knows where it is at all times. It knows this because it knows where it isn’t. By subtracting where it is from where it isn’t, or where it isn’t from where it is (whichever is greater), it obtains a difference, or deviation. The guidance subsystem uses deviations to generate corrective commands to drive the missile from a position where it is to a position where it isn’t, and arriving at a position where it wasn’t, it now is. Consequently, the position where it is, is now the position that it wasn’t, and it follows that the position that it was, is now the position that it isn’t.

In the event that the position that it is in is not the position that it wasn’t, the system has acquired a variation, the variation being the difference between where the missile is, and where it wasn’t. If variation is considered to be a significant factor, it too may be corrected by the GEA. However, the missile must also know where it was.

The missile guidance computer scenario works as follows. Because a variation has modified some of the information the missile has obtained, it is not sure just where it is. However, it is sure where it isn’t, within reason, and it knows where it was. It now subtracts where it should be from where it wasn’t, or vice-versa, and by differentiating this from the algebraic sum of where it shouldn’t be, and where it was, it is able to obtain the deviation and its variation, which is called error.

Try reading that out loud! Bob then went on to show us some more models to see how we can experiment with factors (parameters) in a dynamic visualisations in a way that allows us to problem solve. So schoolkids can reduce differential equations to simple statements relating change and then experiment – without having to know HOW to solve differential equations (what you have now is what you had then, modified by change). This is model building without starting with programming, it’s starting with modelling, showing what they can do and then exposing how this approach can be limited – which provides a motivation to learn how to program so you can fix the problems in this model.

Overall, an excellent talk about an interesting project attacking the core issue of getting students to think in the right way, instead of just getting them to conform to some dry mechanistic programming approaches. The National Computer Science Institute is doing work across the US (if they come and do a workshop, you have to give them a mug and they have a lot of mugs). NCSI are looking for summer workshop hosts so, if you’re interested, you should contact them (not me!) Here’s one of the quotes from the end:

“It was once conjectured that a million monkeys typing on a million typewriters could eventually produce all of the works of Shakespeare. Now, thanks to the Internet, we know that this is not true” (Bob Willinsky (possible attribution, spelling may be wrong))

What would happen if the Internet went away? That’s a big question and, sadly, Bob started to run out of time. Our world runs in parallel so we need to have be able to think in parallel as well. Distributed computation requires us to think in different ways and that gets hard, quickly.

Bob wrapped it up by saying that Shodor was a village, a lot of fun and was built upon a lot of funding. Great talk!