We explored the statistical data sets in a program called gapminder.com. At first, we looked at sugar consumption by country, by year, and compared to the country's gross domestic product. The gapminder website allows you manipulate the graph to show information you are interested in viewing. Whenever you see graph information, one has to be careful not to confuse correlation with causation. I think it is really easy to make this error, and so the gapminder graphs provide a great way to teach students to really question the data.
To prove this point, my group specifically chose two data sets that one would generally assume influence each other. The y-axis was percentage of the population living on less than $2/day. The x-axis was women's literacy levels. We searched for data where women's literacy increased, so did the percentage of the population that lives on less than two dollars a day. One could look at this data and argue that women's increasing literacy has a negative effect on those living in the country.
I learned how powerful data can be, but also how we need to be cautious in using data and not make it say more than it does. I think the correlation/causation lesson is important for students to learn, so I see the relevance for our classroom instruction. I would like to know where the data sets come from and if they are all a high caliber.
Agreed. Data has a story to tell, but we have to be careful how we interpret it. The way we frame things, consciously or unconsciously, has great effect on the implied meaning of the data. Very subtle stuff.
ReplyDeleteDoubly agreed that students need to think about this. It would be fun to let them create a narrative based on a graph without labels, and then let students share out to the whole class to highlight the variability of interpretation.
Is that too nerdy? Yes?
Excellent.