"Could This Be Answered Through Google?"

It would seem that John Paul and I have been assigned several of the northern Supercomputing teams and have been going to visit them. On Tuesday last week, we went to visit two teams in Aguilar, Co.
One team wanted to do a project about teen pregnancy, and the other wanted to work with Global Warming. Neither was exactly a computational problem. The main question we asked when we got there was "Could you find an answer to this question on Google?"

I'm sure most of you know it, but the scientific method goes like this:
1. Question
2. Hypotheses
3. Procedure
4. Data
5. Conclusions

Challenge projects follow the same thing, but the procedure should be written down in ink before anyone comes up with a hypothesis. The most common problem with Challenge projects is that they're not computational. John Paul and I had a long conversation about what made a problem computational and came up with the question "Could you find an answer on Google?" or more specifically "Could you find an answer somewhere else? and Would that somewhere else be too costly in time, energy, money, lives, or any other resources that it becomes impractical?"
This helps us state more specifically what the non-computational nature of most Challenge projects we come across is. I've always used the analogy of illustration. "You're illustrating your solution to a problem, not finding a solution." But this doesn't seem to get through to the students too often. They're not looking for an answer from their models, they've already decided they know what happens and have programmed their model to show it happening. That's why a hypothesis should be written after the procedure, or at the same time as the procedure, or not at all. (I've never been a fan of guessing prematurely, since I prefer to be surprised.)
The blame can't be placed entirely on the forming of a hypothesis though. When teams decided to program what they already know happens, they're not usually wrong in thinking that that's what happens in the real world. That's because they've already found their answers somewhere else. If you can find an answer to your question so easily that you don't actually have to model it, what's the point in modeling it? Hence the question "Could you find an answer to this question on Google?"
The group looking at teen pregnancies wanted to look at the influence of Media, drugs and education on teen pregnancy. They were going to have agents representing teens run around and bump into other agents representing these three influences. Media, drugs and education would have either positive or negative effects on the resistance of the agent to having a child.
At first glance this looks like a very clear idea of what they're doing and what they plan to do. It's a problem that would be too problematic and costly to manipulate and study in reality. It's also an inch away from being a step-by-step plan. All they need is to know how much media, drugs and education a normal teen runs into and how much they're affected by these influences.
All they need is to know how much media, drugs and education a normal teen runs into and how much they're affected by these influences. Isn't that they're question? If they need to answer their question through their model, but they need their question answered before they can build their model, isn't that an oxymoron? Couldn't they find that information on Google? They could and are expecting to. That's not a good Challenge project.
The team was originally looking at how social circles and the people around you affect your decisions as they apply to teen pregnancy and decided to go back to that. They only didn't know how to approach that problem, that's why they'd switched to the project they submitted.

The second team, looking at Global Warming, were looking at how throwing things away instead of recycling produces greenhouse gases and causes Global Warming. This is something that might have been alright, but was an ambitious choice.
There's a bit of a misconception that's very common. It's Trash = Global Warming. It should be something more like Trash = ? = Global Warming. Trash is trash. Global Warming is Global Warming. They're related, but they're not quite the same thing.
Their game plan was actually pretty good from the beginning. They wanted to model different efforts of a community such as Aguilar, and how those lessen their carbon footprint.
After talking to us, clarifying and simplifying the problem, they decided to look at what impact a recycling center in town would have on Aguilar's carbon footprint. Does having recycling pickup add more greenhouse gases from the trucks? What about from melting down and reworking those materials? What about Garbage? What toxins are let out of an incinerator or a landfill? What toxins are released when we mine or make new products?

John Paul and I got them started on programming while we were there. We'll be going back up next month to help them get ready to present in-progress projects to a panel of judges. We'll also ask them again, "Could this be answered through Google?"