How NOT to do an experiment

Over on Facebook I follow some groups who find provocative topics, and today’s “science fair” post was so over the top that I had to share it here.

science fair

Here’s the original post. Now the accompanying text about microwaves is whacky enough on its own (and well worth reading), but my primary interest is with the experiment. This exemplifies why there are basic rules for doing science.

This starts out okay – identical pots, the same type of media (I assume), similar sized plants – but then things go downhill:

1) Replicates are important. There is one treatment and one control, meaning that it’s impossible to run any kind of statistical analysis. Ideally between 10-20 replicates of the control and the experimental treatment are used in this kind of experiment. That’s 20-40 plants total.

2) Variable control is important. Plants in a windowsill are subject to light and temperature gradients. That makes analysis more complicated unless one has an extremely long windowsill so that all plants are treated uniformly. And then our researcher prunes the tops of the plants – yet another variable.

3) Consistency between treatments is important. It appears that the pot on the left is wetter than the one on the right – the media is darker. If it’s not draining well – for whatever reason – then you’ll have a hypoxic root environment. Plants don’t like that.

4) Objectivity is important. It’s difficult (impossible, really) for any researcher to be completely objective. Ideally, the pots would have been watered by another person and then labelled in such a way that the person recording the data would have no clue which was which.

I think it’s really important to get kids excited about science. But it’s just as important giving them guidelines about doing science in a way that advances their own understanding about how the world works. Otherwise, it’s just more fodder for the aluminum hat crowd.

4 thoughts on “How NOT to do an experiment”

  1. When judging a high school science fair one year, one of my favorite projects was someone who was growing small trees in pots (I forget what he was testing). He did everything right with the design and execution, but his dog dug up most of the plants before the project was over. The genius of his project was that he didn’t conclude anything. He explained what happened, showed what data he could salvage, but stressed that it wasn’t meaningful. It remains one of the most realistic science fair projects I’ve encountered.

  2. As someone who has worked in a garden center for more than two decades, I wish teachers would assign gardening science projects when the plants/materials/seeds are more readily available. When kids drag their parents in in October looking for quick growing annuals (when all of ours have already been marked down and sold) or when they only have a month to record growth on a plant in the kitchen window in January (when most plants are responding to the short days and weak sunlight of winter and not growing), it’s frustrating for them and me both.

    1. Excellent point. I think plant projects should only be assigned during spring semester. (Or maybe focus on winter vegetables in the fall?)

Leave a Reply