Data isn't so bad after all!

I'll be honest, I wasn't looking forward to the data unit.  Students generally don't take computer science to learn or think about data.  Honestly, a lot of the material felt pretty common sense-y.  I wanted it to be more math-y.  I wanted students to have more skills to create visualizations that were more impressive.  I wanted to make it more focused on social justice.  But, I went with it... surprisingly, it turned out well!

First of all common sense, is never very common.  By the end of this unit, I came to understand that the real learning was about being a critical viewer of data.  My goal for this unit is to build student capacity to generate knowledge from data but also question how visualizations or data collection might have a bias.  This unit certainly provided ample opportunities to build those student skills.

I was able to bring in a lot of real life articles and have a short discussion about them in class.  One of my favorites was about non-voters which as super relevant to the content for the following reasons:

• It told an excellent story and used visualizations to support the story.  
• It used a variety of visualizations (Bar charts, pie graphs, stacked bar graphs)...
• It showed how different visualizations could be used to lead to different recommendations
• The data collection methods and "cleaning" of data was similar to how we cleaned our own in-class data (free-response and then categorized). 

In reality, I think I could have used this as an example of what students could do for our final project in the unit - use several different visualizations to tell your story.

As I graded student work it was clear that next year I need to do more with "story" development.  I should use different articles that tell a story with data and have students read/analyze those.

Additionally, there is still a lot of opportunity for sense-making.  Several projects have severe gaps in sense making.  For example, they made a chart, but it isn't clear what the data actually is about.  I could even do a whole lesson on "naming charts" since a lot of students left of titles or had titles that didn't make sense.  I could give them a bunch of title-less charts and have them come up with the names.

Finally, I tweaked the original project to allow students to explore other data sets that they felt were interesting and relevant to their interests.  I actually had most students take this option which was exciting.  It helped bring out student personalities and also gave them a glimpse into big data - some data sets were QUITE large!  Overall, it turned out well, and I am thinking about how to improve upon it for next year.

For next year, I am still fooling around with these ideas:

• Making it more social justice focused - this is always a goal of mine in math, it seems like here is a great place to do it too.
• Talk more about the use of story in data
• Make more interactive visualizations either with programming or with Google fusion tables
• This would take more programming experience or we would want to walk through one together as a class in google fusion tables before setting students free.
• I could move this to the end of the year and do more in-depth applications if there is time
• Tie this into big data with Kaggle.com's data sets - it actually has a lot of great resources to use!