The Scientific Method

For the last few years I have been a judge at the Vermont State Mathematics and Science Fair. I've got a background in science-based disciplines, technical, physics-based stuff. It's quantifiable, measurable, and not governed by "consensus" but by physics. Repeatable, verifiable, quantifiable. ymmv

As my kids would say, "back in the day" I was a geek when it was not cool. Slide rule hanging from my belt, pocket protector with pens and fine tipped, mechanical pencils. Studying and using trig, designing and building circuits like flip-flops, decade counters, and other digital stuff back when no one really knew what digital was. Building digital clocks when they were still something unique and "modern".

So, I can relate. I did the science fair thing a couple of years in a row. Built a seismograph from an article in a magazine and ran it for awhile. Collected a few rolls of information, entered the seismograph in the school science fair and went from school to county science fair. The judges at the Los Angeles County Science Fair really liked that a seventh grader had done this and understood the science behind the device. But - I had not collected any data and then done the analysis. I had several sheets showing the output from the seismograph. Living in Southern California with tremors going on all the time that I had detected. Instances where a heavy truck had gone by and I explained how I could tell it from seismic activity. But no real data collection and analysis. Epic fail.

What do we have for our contestant Johnny?

So, I'm at the fair and there are a lot of good projects that I see that day. A couple where the kids involved said, "Gee, I wonder why this works this way?"  They did a decent job, but didn't really dig into the object of their study. Still, scientific method was used. Process good, no real curiosity, the initial experiment suggested no follow on questions to them.

C'est la vie.

Some of the others were middle schoolers and show really good potential. Good use of the scientific method, good rigor, nice analysis of the variables that could influence the results, data collected and analyzed (unlike some predecessors), and supportable conclusions drawn. Real science done at an early age.

A couple of them were just outstanding. Good understanding of the physics involved, excellent understanding of what variables and conditions affected their results, why they could have, and how they could have skewed results. In one case he also looked at return on investment (ROI) for anyone that would use his research and implement it. Nice touch, in high school and he already understands that for the average blue collar family, there has to be more in it than just being green. If even Kermit found it hard to be green, what about us non-amphibians?

And one student that really got her experiment. The more you spoke with her the more you were blown away by the fact that she got it. All of it. All of the physics that her experiment depended upon. How other factors, the physics of these factors, impacted her results. How viscosity could influence the results. Temperature. Bevels on the edges. Water flow and how it is affected by surface tension, which is in turn multiplied by the number of openings. In every case I asked the students, if you could do this experiment again, what would you change when you performed it again. If you were to perform a follow on experiment driven by the results of the first, what would it be? What follow on research was suggested by your first experiment? She had a number of interesting hypothesis suggested by her initial experiment.

One uniform characteristic exhibited by all of the students was that their conclusions were driven by their results and their conclusions were driven by the analysis of the results. So? Well, we see plenty of "science" nowadays that is supported by "scientific consensus" and is not repeatable. And where the "scientists" work to fit their data to the curve rather than fitting the curve to the data.

I'm no longer shocked when "scientists" and "researchers" announce what they are going to find before even examining the data. All of these kids plotted curves based upon their data. They weren't fixated on a pre-established conclusion where the hypothesis and data were only relevant in that they provided support and buttressed the conclusion. These kids are still budding scientists, unsoiled by what frequently passes for science nowadays.

But, there's still time. They still have many years ahead of them for the contemporary scientific community to teach them how to establish the "right" conclusion before going out to prove it. Creative statistics, curve fitting (and not the way you old timers learned it from your profs), and "socially responsible science" are still in their future.

Maybe that is why I enjoy it so much, I get to see them while they still think that science is about the truth of the physical sciences.

marcus erroneous

Addendum 7 April:


Here's a video about the fair. Five of the eight projects I judged got screen time, two of them were interviewed.