ICME14 Day 2 Part 2

After lunch on the 2nd day, I attended TSG12, the topic study group on teaching and learning statistics. This is not because I am doing research on it (I'm not), but rather because it would be nice to learn more about this, and some of the talks seemed interesting. 

The first talk was by Lonneke Boels, presenting "Designing embodied tasks in statistics education for grade 10-12". Her research question was "How can literature—on misinterpreting histograms and on embodiment—in combination with results from an eye-tracking study inform embodied task design in statistics education?" Many students confuse bar graphs and histograms, and many think one of the axes should be time. Boels discussed an experiment with eye tracking, asking students to find the means from different diagrams. Eye tracking made it very obvious that the students were interpreting the histograms as if they were bar charts, and also that the students carefully read the text and the labels on the axes). The results were depressing. Thereafter, they did some experiments where students had to make boxplots themselves by dragging each data point to its correct position. They also had tasks where students had to find the balancing point (which is the mean) of the boxplots. (In a way, it doesn't make sense to summarize in this way, because the devil is of course in the details in such tasks. For instance, the software gave feedback when the balancing point was set correctly. Such details will influence the learning in some way. However, the point of my blog is obviously not to replace the article but to give a small taste which may lead you to look at the article...) Working with students merging and splitting classes seems to be a useful strategy.

(Being a novice in this area, I notice words such as CODAP, iNZigt, Tinkerplots and VUStat, only some of which I understand. I should investigate...)

In the discussion, it was asked what an "embodied task" is - is that dependent on the task or on your perspective when looking at the task? That is a good question, and I would perhaps answer that no mathematics task can be "non-embodied", but of course the way and degree to which a task is embodied, is an important consideration in task design. (This reminds me of a recent fad: to call everything "semiotic". I have read about "semiotic representations", but never of "non-semiotic representations". But of course, I am not saying that looking at embodied learning is an unimportant "fad".)

The next talk was Hanan Innabi, "Teaching statistics and sustainable learning". Her starting point was variation theory. The idea is that (based on Marton) students work on variations lead to sustainable learning. She shortly presented a research project with Marton, and mentioned the special thing that variation is an inherent thing in statistics, because uncertainty is an important part of statistics. (So in a sense, it is hard to teach statistics without much variation - although textbooks of course have statistics tasks where the data are given and variation is lost.) At the end she mentioned a couple of examples of how to work on this using variation.

The third talk of the TSG session, was Daniel Frischemeier's "Reading and interpreting distributions of numerical data in primary school". After giving a background of previous research, he presented his project of supporting primary school students to read the data and read between the data (not including the third level, reading beyond the data), with 19 primary school students (age 10-11), with use of real data, TinkerPlots and collaboration. The teaching was based on the PPDAC cycle (another FLA that I didn't know before). The conclusion was that ... and here the time was up, sadly. I suppose an article will be available at some point in the future. (In the discussion afterwards - in the chat - the distinction between stacked dotplots (such as in TinkerPlots) and messy dotplots (such as in Minitools) was stressed. Children find stacked dotplots more difficult to understand than messy dotplots.)

Then followed Carlos Monteiro and Karen Francois' "Statistical literacy as central competence to critically understand big data". Monteiro pointed out that students need to understand that they are also producers of big data, understand the power issues and to be able to analyse them critically. This is not an issue of how to handle the data technically, but to understand the origin of the data. 

(At this point, a small misunderstanding regarding time was sorted out, and Frischemeier got five more minutes to finish his presentation. The posttests were, not surprisingly, better than the pretests.)

The final presentation of the first session of TSG12 was Florian Berens, Kelly Findley and Sebastian Hobert's "Students beliefs about statistics and their influence on the students' attitudes towards statistics in introductory courses". In their experience, students have negative attitudes towards statistics when entering university, even though they have not had statistics before. They measured attitudes by the six-dimensional SATS-36 instrument, and created their own instrument for measuring the beliefs about statistics (descriptive perspective, investigative perspective, confirmation perspective and a rules-based perspective). They did find some correlations between attitudes and beliefs of statistics. In particular rules-based beliefs are connected to negative attitudes, while investigative beliefs are connected to positive attitudes. This raises interesting questions for teaching of statistics - probably also on lower levels than university.

Then the day ended with the Plenary panel on "Actors for Math Teacher Education: Joint Actions versus Conflicts". To be honest, I have attended very few good panel discussions in my life, mostly because they often end up not being panel discussions but instead turn into a series of barely related short lectures. Thus, I decided to take some time off from my blog-writing duties to just follow the plenary panel away from my keyboard. Instead, I watched it with pencil in hand, making some notes as it went along. Here it is:

That was the second day - and the first full day - of the ICME14. Five more days to go.