Tian Zheng implemented my candies demo using Legos:
Also lots of details on the results. The point here is not exactly what happened (but, yes, the demo did work) but rather the idea that you can use photos and graphs to document what worked in class. We should be able to do this sort of scrapbooking all the time as teachers.
Next time: take some photos of the kids in class doing the activities, too (assuming that’s ok with them).
I’ve used a similar activity from Activity-Based Statistics (Schaeffer et al, Springer 1996), in which students draw (without looking, and with replacement) 10 strings out of an opaque bag of 25 using the “Touch method” (i.e., shake the bag, then just grab the first one you touch) and the “Random method” (I give the students a list of the lengths of the strings, with the strings labeled 1 to 25). They then compute the mean for each sample and put each mean on the appropriate stemplot on the board or to be shown on the doc cam. There are also two follow-up exercises for discussion, identifying what selection method would be like the touch method in different scenarios.
I made up several bags so the activity would go quickly (they only need a bag for the touch method; I give the the lengths on the reverse of the handout). I saved the bags and used them in subsequent semesters, so there was only a one-time preparation effort. (I think someone else inherited them when I retired.)
It seemed to be very effective, making the point that what might appear to be random sampling is in fact biased. I often gave an exam question asking if some specific sampling method analogous to the touch method gave a random sample, and most students would get it.
Martha:
One reason I like the candy weighing version is that it is so easy to compute the total by simply weighing the bag.
True, but it has the drawback that each pair has to take their attention away from what’s going on in class when they get the scale. (I wouldn’t like that if I were a student.) With the strings, the activitiy can proceed almost simultaneously with all the pairs, and those that finish early can discuss the follow-up questions.
Yah, good point. Distraction is an issue when I do this activity in class. It can take a lot of effort to keep the class focused while the candy weighing is going on.