We’re about a third of the way in The Pit and the Pendulum unit and I think most of the cherubs are getting a nice understanding the importance of sample size, controlled experiments, and representing data with histograms. After our initial experiments with pendulums, they couldn’t agree if the weight of the bob really affected1 the period of the pendulum or not. I told them they needed a new tool to make that determination and so began our journey to the land of statistical significance.
We began been collecting data and collecting data. We’ve collected data on pulse rate, stride length, estimating 5 seconds, and average amount of time spent online. Making histogram after histogram. Making them on graph paper and on butcher block paper. Posting them around the classroom and comparing. We haven’t moved to creating them on the graphing calculator yet as I think they need experience doing it by hand first. Once I know that they know how to create and interpret a histogram, we’ll move to using the calculator. We’ve briefly talked about the normal curve and will delve deeper into measures of central tendency this week.
I’ve been using this site from Shodor to create histograms2. I like that we can quickly change the class width or add more data. Here is the pulse data for my first period:
Initially I was hoping for a more normal distribution. As it obviously didn’t work out that way, we combined the data from two classes3:
Still not quite normal. So we looked at the data with a class width of 4:
Better. In each class one student suggested that if we had even more data it would look more like a normal distribution when the class widths were one. Okay, they actually said something like “More like that curve it’s supposed to be” but I’ll take it.
1 I’ll admit right now, I’m going to goof up affect/effect repeatedly in this blog – let alone in this post. I’ve always had a mental block on this and I expect I always will.
2 I think this counts as a happy marriage of my digital projector and laptop, no?
3 Obviously 5th period had a little problem with either data collection or data recording as our n is not a multiple of ten.