Category Archives: Statistics

Virtual Filing Cabinet 1.0

I trawled, I tweeted, I tried to find the best of what was out there. I have posted the first draft of my virtual filing cabinet here and in the process, learned the following:

  • There are so so many great lessons and activities out there!
  • There are far too many to include them all
  • It is hard to choose which ones to include
  • There are some incredible teachers who I would like to be more like
  • I’m really excited about trying just a handful of things I found
  • I want to do a lot more project based lessons
  • This is a working document and will evolve with my teaching

If you haven’t already I highly recommend you trawl some websites and create one of these yourself. Chances are, it will be unique to you and your teaching style and philosophy. Here is a list of websites to get you started. 

Let me know if you think there are any glaring emissions.

Gapminder is Awesome

photo (9)

I wanted something that would open up the world of scatterplots to my statistics students; something where they could really get a sense of correlation and causality. I decided to do a project based around the fantastic GapMinder World and it payed off.

First I showed this video of the master Hans Rosling at work with the graphs his foundation came up with.

I then gave these instructions to my students:

You (and max one other person) are to prepare a 3 minute presentation on a GapMinder graph of your choice.

Instructions:

  • On a computer go to the gapminder website by clicking here
  • Play around with the explanatory and response variables until you find two that you think show some sort of relationship
    • If you are struggling to find variables with a link, click ‘Open Graph Menu’ and play around with graphs that have already been created.
  • Your presentation must include answers to the following questions
    • What are your explanatory and response variables?
    • What is the link between variables at the start (before you click play)?
    • What do you notice happens over time?
    • Are all the countries close together or more spread out? What does this mean?
    • Are there changes to any particular country that are of interest to you?
    • What if you isolate by continent? Are there any changes that are of interest to you?
    • Is there anything else that stands out with your graph?
    • Are there any outliers to the trend?
    • Does this graph bring up any other questions that you would want to investigate further? What information would you need to answer these questions? Is this information available?

Your presentation must include the time series animation (when you press play) as well as PowerPoint slides using screen shots of points of interest.

You will be graded on:

  • Content (out of 6)
  • Presentation (out of 4)
What went well
  • This was a great way to get across a sense of scatter graphs and will be awesome to segway into taking about correlation and causality.
  • This was enjoyed by the students and really got them thinking about statistics and global affairs
  • It was good to give specific questions for the students to answer. In my experience just saying ‘present for 3 minutes on a graph of your choice does not give great results’
Even Better If

Next time I do this, I think it would be good to model what an excellent presentation looks like. I missed a good opportunity to teach this skill.

Popcorn and the Normal Distribution

If there is one thing Indiana is famous for, it’s corn. I saw this lesson from STatistics Education Web (STEW) on popping corn and introducing the normal distribution and loved the idea.

The basic idea is:

  • Make some popcorn in a microwave
  • Have students tally the popping count in 5 second intervals and record the frequencies
  • Have some students mark where they think certain points on the normal distribution point were reached.
  • Discuss the results

What went well:

  • All students were engaged the whole time
  • It certainly gave a sense of why the curve had a bell shape, thinking about and discussing not only direction of slope, but whether the slope was increasing or decreasing as time went by. I think this will really help with students’ understanding.
  • We got to eat some yummy popcorn

Even better if:

  • This would be great as the first lesson to introduce the normal distribution. We had already discussed the shape and real world example so some of the impact was lost
  • I would like to make more of predictions and pre experiment ideas next time.
  • There needs to be more extension questions that link to other distributions and different shapes to contrast and compare
  • Check with the schools facilities before using the microwave. Ours blew a fuse on the second run through.

How useful is the ‘percentage likes’ statistic?

This is a copy of the question I sent to the BBC’s More or Less program.

When choosing a film to watch at the cinema, I often check review sites such as Rotten Tomatoes in order to see overall percentage ‘likes’. The latest Star Trek movie achieved 87%, for example. This got me wondering, how useful is the ‘percentage likes’ statistic? How would it be different if it was weighted against certain preferences, to account for, for example, lifelong Trekkies that would have a tendency to ‘like’ a Star Trek film, however good it may be?

We’ll see if there is any response.