I've been thinking about science experiments. Work has me in a bit of an academic mode (I'm [co-]writing a paper!), but I don't think that's it so much.
But I've never been happy with the school science experiments I did. I only remember two (although I'm pretty sure I also did the volcano baking soda "experiment" in 3rd grade).
The first was a bicycle with a generator on it, making a light bulb go. I don't think there was really any pretense at experimentation, more of a "this is power generation, cool!". Looking back, some pretense on science could have been...how much power is lost in this process? Measuring lumens and effort some how.... putting a generator to run the bike, to run the generator, to light the lamp, ... dunno. That's not where my thinking is right now.
The second was growing yeast, measuring output by balloons on the ends of bottles. I don't remember what all I grew, or what the results were...but it was a little closer to science. More interesting might have been trying to estimate the amount of sugar in various products...admittedly it would be wrong due to other factors inhibiting growth...hmm, these days it would have been interesting to throw in all the sugar substitutes as well....still not really where my head is.
So I've had two ideas, maybe, for science experiments. (I know, there are books and books of these!)
1) tin can conduction. use a speaker and a mic and test sound propagation over a cross section of wires and tautness.
2) experimental error. do the old "gravity is how much" test, with a handful of objects (ideally with little to no air resistance as that's a different, more complicated? experiment). Repeat each drop 20 times, from different heights (such that the objects have much higher velocities, such that their exact drop times and velocities are more sensitive to measurement error), and see how those errors propagate to calculations of gravity.
What's the coolest science project you did growing up? How would you do it differently "now"?