Square Dance Experimentation

[bjarvis]

I started wondering a little while ago: who has to dance more steps, the folks dancing boy or girl? Does this vary from level to level? Do some callers make one role work harder than the other (at least in terms of dance steps)?

Tonight, the experimentation began.

Tonight was a rare 5th Thursday so DC Lambda Squares hosted an A2/C1 club night with Ett McAtee calling. I put pedometers on at least each dancer of a particular couple and recorded their step count per tip, noting which were A2 and which were C1.

While the pedometers are of the same manufacturer and lot, they have different sensitivity. I'm trusting that by randomly switching these between dancers across the course of the experiment, the variation will average out in a sufficiently large sample space.

The raw numbers aren't meaningful yet. I'll write more on this when I have a much larger sample. Who knows... it may be good filler in a slow news quarter for the GCA journal. :-)

Comments

[bjarvis.livejournal.com]

Not quite that dramatic, but yup. That's why they're being randomly assigned in each test couple... in a sufficiently large sample, the odds of each having the more sensitive pedometer approaches 50%.

Alternatively, I could wear both of them myself, take multiple measures, determine the skew of the more sensitive one and use that factor to normalize the numbers. This however would require additional record keeping to mark which dancer received which pedometer.

It was easier to declare Pluto wasn't a planet.

[trawnapanda.livejournal.com]

Alternatively, I could wear both of them myself, take multiple measures, determine the skew of the more sensitive one and use that factor to normalize the numbers.

This looks like the more reliable correction, though. You should normalise by having the same dancer (which could be you), because the bouncing around that a dancer does is significantly different from flat-level walking, and I suspect that the differences between the two may be exacerbated by that.

This however would require additional record keeping to mark which dancer received which pedometer.

However, to eliminate this increased record keeping, I further suggest that once you've normalised the two (two or at most three full evenings of dancing, both worn by the same dancer, maybe one night lead one night follow), designate one as the "boy" pedometer and the other as the "girl" pedometer. Then you'll know from your own records which pedometer was which, and the correction factor is constantly applied to just one set of data.

It's been a while since I danced regularly, and I always danced follow, but I got the distinct impression we were having more fun and bouncing around a bit more than the leads. (Teacup chain particularly springs to mind). You might find stats skewed by people who style more than others.