Data Analysis made Almost Interesting
Here’s a way to generate high interest data for grade-conscious middle school kids and get them engaged in analysis: Use their own test scores and graph them against all kinds of variables that might have affected their scores.
Here’s how it went:
1. Hand their tests back.
2. Go over the test in the usual way.
3. Ask what might have affected their test score. (Groups, then whole class.)
4. Set up the X-axis with an appropriate scale. (If you have squirly classes, you can do this ahead of time for common suggestions ex. number of hours studying, number of hours of sleep the night before, hunger on a 1-5 scale, interest in the subject (careful with this. Don’t use if you are having management issues in the class), paying attention in class etc.)
Here’s the basic graph set-up, plus the notes from the end of class discussion on what they learned:
5. Demo how to mark their data point on the graph.
6. Pass around the different graphs – 2 per table of 4. Have each student mark their data points on each graph, then pass the graphs on to the next table. (Might be good to have a sponge activity to take up slack time, maybe starting on some homework or reading a science magazine.)
7. Once all the graphs have data points marked, hand out one or two per table group and have students analyze the data like this:
8. After about 10 minutes, have each group put their graph under the document camera and present their findings to the class as above.
The ‘best fit ellipse’ was a good way to get to a basic understanding of line of best fit and standard deviation, correlations and t-tests. Nice to get them set up for it, and my math teacher colleague Sandy approves 🙂 (she actually knows what she’s doing with this stuff, I’m just faking it.)
I was so excited by some of their findings and the glimpse into their real lives – how much of a distraction texting is for example. Also their grasp of whether data was likely to an over or under-report: were kids being ‘cool’ to their table mates in answering “How interesting is the topic?”, or were they being more honest than in their journal writes? We discussed the effect of anonymity and all sorts of other data bias. It was a genuinely interesting discussion to all of us, with unexpected insights and opportunities for further research:
According to our data,
– hours of study made very little difference to test outcome.
– being hungry and being really tired did make a difference, not a good difference.
– using index cards does not guarantee an A. Disappointing because something THAT boring MUST be useful!
– being interested in the subject, some correlation but ‘motion’, not that exciting apparently. Although 2nd period said it wasn’t the activities, just the actual topic that was meh.
Anyway, give it a try! Share the cool stuff you discover about your students and about effective studying. If you want more specifics, feel free to e mail me.