Science! Part 1

12/16/2012

When Steve and I headed to dance to Perpetual e-Motion on Sunday night at Glen Echo, we decided we might do the initial forays into an experiment. Several weeks ago, one of my more vigorous partners (read: that guy who twirls the snot out of me every single time) made a comment about how he was tired, but "you're the one doing most of the work anyway."

Which got me (as someone who typically dances the follow's part) thinking: does the follow actually exert themselves more in a typical dance? There are people who object to my use of "lead" and "follow" on this blog because they believe it to be an entirely equal dance form for which those labels are (or should be) meaningless. Meanwhile, in a ladies' chain (which is so much more common than a men's chain that even a decent chunk of experienced dancers don't know how to pull a men's chain off smoothly, at least around these parts), someone dancing the lady's role and dancing without flourishes crosses the set and takes at least four medium-to-large steps around the outside of a circle, whereas someone dancing the gent's role will take maybe four small steps backwards to help the lady around the courtesy turn.

Admittedly, this doesn't seem like a large difference, but multiplied over the course of an entire dance where it's done maybe a dozen or so times through at least, it's a much larger difference, just for that one move. But does this difference even out over the course of a dance, or no? How much of a difference does being the inactive couple in a long set make?

So Steve and I gathered up some (cheap) pedometers and sought to calibrate them on Sunday.

We are making the following assumptions:

The number of steps someone takes is reflective of the amount of physical exertion someone expends in a dance;
The average of these numbers will give us a general sense of which "role" physically exerts itself more overall;
The pedometers are relatively accurate;*
Pedometers were reset to zero after every dance's walkthrough, but before the actual dance started so that we weren't counting mistakes made in the walkthrough in our step counts;
Dances take place over roughly the same amount of time;
We are using Glen Echo as our model; we are aware that most dance communities' local dances do not have 2 to 5 long sets of over 15 minor sets each, but that's our local sample so we're using it (acknowledging that this may make a difference too, especially with dances where your inactive/active position matters);
The differences amongst dancers, bands, and callers (fast/slow, chestnuts or more modern dances) will normalize over several weeks' worth of data and a varied list of partners (and combining Steve's partners with mine, as we tend to dance in different lines when not dancing together).

* some of the data points imply that perhaps this is not true, or that we should be more careful in matching up pedometers amongst partner pairs (or find a better way to calibrate them). But this is why we collect preliminary data ahead of time.

What follows is our raw data from Sunday:

Glen Echo, MD, Sunday, December 16, 2012
Band: Perpetual e-Motion
Caller: Bev Bernbaum

Raw data from initial calibration/data collection

We know that some of this is as a result of inaccurate collection methods, and we're planning to use that as a lesson learned when we go collect our "real" data. We need to be more careful about making sure that the pedometers are in the same spots on the dancers, and that they're identifying the same motion as a "step" to a reasonable degree of accuracy (i.e., calibrate them better). We're not sure that's happening here, so we need to find a way to mitigate it (preferably without dropping $100+ on better pedometers -- this is curiosity turned into an experiment, nothing more).

So we're opening up our process and our data to other minds as well. How else can we improve our experiment when we jump in for real? Anyone care to venture any hypotheses for when we get this off the ground?

Special thanks to Will, Jesse, Emma, Ron, Leslie, Abby, Miriam, Steve B., Eireann, Eric, Bethany, & Sophia for your help gathering preliminary data!

To be continued....

Perry

12/18/2012 07:29:39 pm

My hypothesis is that it'll probably come out roughly equal in the wash. Most callers try to make their evening of dances so that no one gender is doing more "work" than the other. At least that's what I do. For every gents allemande 1 1/2 there is another dance with a ladies allemande. Generally. Of course there are more ladies chains than men's chains, but I figure there are usually enough gents-only moves that make up for it. Especially if you are going for a long-term experiment. And I guess that depends on who you dance with too - looking at dance 9, which should have been the same dance for each couple, but one couple had the gent with more steps and the other had the lady with more steps. Perhaps one couple flourished more than the other (and most flourishes have men spinning ladies?). Maybe there was role-switching going on? I don't know if you're going for strict science here, but doing things like extra flourishes, role switching, and things like that will definitely skew the results. You're going to have to control your conditions if you're going to get an accurate reading, otherwise you might get results that don't tell you much. Though you do assume normalization amongst dancers - one extra twirly dancer will cancel out some of us fat lazy dancers who don't twirl as much. ;)

Ryan (blogger) link

12/18/2012 11:16:40 pm

Some of it does depend on who you're dancing with, I agree -- but I also figure that enough data points should normalize it.

Also -- for the purposes of this experiment, no role switching. That *would* skew the results. :-)

Perry

12/19/2012 05:56:32 am

Could you help define the chart? For each dance you have two numbers for lead and two numbers for follow. So that tells me that you have four different people wearing pedometers for each dance. I am quite fascinated by dance 9, where one couple has the man with 100 more steps and the other couple has the woman with almost 200 more steps. Presuming it was the exact same dance, I am wondering how that could be? To be sure you normalize, you're going to need to balance out those who flourish a lot with those who don't flourish very much, or use only non-flourishers, or use only flourishers. Seems to be a lot of degrees of freedom.

Ryan (blogger) link

12/27/2012 12:29:44 am

I should have fixed this and hope to do so this week:

The two rows are Steve and I...we each had on a pedometer and put one on our partner, so we had two couples per dance with pedometers on (except when one of us sat out, and therefore did not have a partner, nor a count of our own steps for that dance).

For the ones we danced together, we each had on two, which is why we think the pedometers need better calibration before we can use them in a proper experiment.

Maynard

12/19/2012 01:41:15 am

You might consider using the Weight Watcher activity measures. Also you could have more than one pedometer on the same person to see if there are differences and average them out. Some dancers have a weight watcher activity measurement devise or pedometer, and would bring it to the dance.
The other measure you could get is a heart rate monitor. A lot of people have this. It would be great if you could get one that record the heart rate chronologically.
The other thing is you could contact the http://www.cooperaerobics.com/Cooper-Institute.aspx
group and see if you could work with them to figure a method and devises that you would need to measure it. They may be interested in dance as it is a social activity that may be more appealing to some than traditional exercises. They are a non-profit like you.
Some callers alternative between fast and slow dances and between different types of dances. It would also be interesting to note differences in dances.
It may be useful to measure heart rate at a certain level and also calories burned based on heart rate and movement.
Some moves like ladies chain may have a lot to do with how the person is turned when the reach the other side. It takes more energy to turn the person into place than to do an overhead twirl.
The research may lead to a better understanding of contra dance as a form of exercise and how to make the most of the workout.

It would also be useful to interview people on how they were injured to see if there are methods to reduce the accident rate. You could have an online data form to record injuries. And the best methods of dealing with them.

Ralph Sturgen

12/20/2012 11:38:06 pm

1. I believe that pedometers measure the number of "steps" only, neither their length nor power.

2. I think you shall find that the # of steps is more dependent on the dancer than the role although the role may play a part. I'm an "older man" who dances the "lead" almost exclusively. Thus, some thoughts:

a. While you're chaining to me, I'm probably standing still = 3 or 4 more steps each repetition of the dance.

b. If you use a buzz step for a swing and I do a walking swing = 8 more steps per 8 beat swing.

c. Whoever clogs or step dances while waiting out at the end = 128 more steps (4 beats X 32 measures).

If you can correct for all of these variables, I expect you shall find that both dancers take the same numbers of steps because they're dancing to the same music. So, I doubt that pedometers are the necessary measuring tool.

I'd still like to see the data, though.

Bob

12/21/2012 06:37:25 am

I wore a pedometer to several dances in a row during a month of counting steps for a work contest. Each time at our local Thursday nite dance (8-11pm) I clocked 10,000 steps, +/- 300 or so.

Comments are closed.

Science! Part 1

Author

Archives

Categories