Beautiful Statistics and Storytelling

One of my favourite university courses was my first full course in statistics.  I still remember the first day when our professor indicated that there were two ways that she could teach the course.  One was through number crunching and the other was through story telling.  She indicated that her approach would be story telling and that there was another section of the same course offering the other approach if anyone wanted to change.

I didn’t change – I’d been through the whole registrar process a previous time and didn’t want to go through THAT again.  I decided to stick it out and I’m glad that I did.

Her message and it’s stuck in my mind all these years is that you’ll be more effective in communicating the statistical results of anything if you can present it in a story.  It made sense at the time and it makes even more sense as we get bombarded with statistics and numbers on a daily basis.

In education, I’ve sat through so many presentations about research results.  So many of the presenters probably took the “other course”.  Persuasive discussions are shown via Excel spreadsheet which you’re supposed to get onside with because Excel is “the industry standard”.  To help you understand the results, watch the presenter show how she/he highlights rows and columns and cells to make the point!  You might even be lucky enough to get your own printed copy with the highlighting already done!

On the other side, I once worked for a superintendent who turned out the most engaging presentations.  Every one was a story!  While I’m unable to remember a single presentation from the first method, I can still remember some from the second.  One, in particular, was on the district’s love with paper, printers, and photocopiers.  Grounded in statistics, you would sincerely have to dig to find them, (they were there…); the message was embedded in the story which was filled with imagery – trees, recycling bins, home fridges, …  It was so powerful and memorable.

One of the reasons why I’m a fan of the infographic is that they tend to take that approach.  Of course, at the heart, you’ll find statistics but the presentation of a good infographic tells a story and takes you along for the ride.

You may have noticed some Twitter messages recently from Vizify.  Hopefully, you’ve been notified or tagged in one because statistically you showed up significantly in someone’s timeline.  My most recent tagging was from Teresa Marrello.

After receiving a few of these, I decided to check it out myself.  If you want to jump to the end of the story, my video is stored here.

It was with a little hesitancy that I proceeded because Vizify wanted me to grant access to my Twitter account.  But, I realized that I could revoke the access afterwards so I went forward.  It was only a few seconds later and my movie of 40 seconds was created.  I watched it to see where my major interactions were from the past year.

A lot of it made sense.  I’m an early riser and am most active personally first thing in the morning over breakfast and the morning news reads.  The rest of the day is random, scripts, and could be at any time.  It would be interesting to see how much was actually done on weekends when my days aren’t exactly scripted!

After the movie, I wondered “Is that it”?  My whole year summarized in 40 seconds.

Fortunately, Vizify lets you do what a statistician would call “drill down”.

So, is there more than just my top three topics?

Definitely, but I am happy that I am seen to be promoting the wonderful efforts of Ontario Educators.  How about my “Golden Followers”?  Top three and that’s it?  Poor Brandon – he tries so hard.

The site does allow you to dig a little deeper.  I didn’t realize how much “coffee talk with Linda” had transpired!

There’s Brandon!  Just missed the cut.

In fact, Vizify does allow for a little editing of their results or you could even add another scene to your story.

Now, that’s what I call a great storytelling approach.  Now, certainly, I could access the entirety of my Twitter history if I wanted and then build my own story using my software tools.  Vizify does a nice job of taking on some of what most people would call their highlights.  If you’re interested in this statistical approach to analysing yourself, do it at their website.  If you have a classroom Twitter account, I’ll bet your students would get a kick out of the results.

When you’re done, you should probably consider revoking access to your Twitter account until the next time you want to run the routine.  It’s just security common sense.  While you’re there, you should probably take a look at all of the applications that you’ve allowed access to your account.  If there’s no clear reason why they should, here’s the chance to turn it off.

Big Data Sets

One of the things about teaching Computer Studies is having data sets for students to run with their program to get the types of results that you should.  During program development, I always found that it was desirable to talk to the students about how to create their own data sets to meet the specifications of the program.  It shows that they can read and understand what is needed and generating data that allow them to work through a problem manually and compare their results to what is generated by their program is a great technique to master.

Of course, when it comes to testing the accurateness of their program, you want to have your own data sets.  I would make it available as a file that they would read and present the results.  Since I know what the results should look like, it’s a quick and easy way to test their programming skills.  It’s also handy to have a couple of data sets to completely test the program and also to check if the student requires multiple runs in order to get it right.

Sometimes, generating these test data sets can be a real chore in themselves.  Not so, if you use  Using this  utility, data sets can be generated so quickly.

From a menu, it’s a matter of selecting just what types of fields that you need for your file.  With a wide variety of choices, I think you’ll find this very functional.

Keep reading, and I’ll show you how easy it is to create a mailing database.  In the Column Title, generate a meaningful title for the field and then pick the data type from the pull down menu.  In this case, I really appreciate the Canadianization of the data – I can specify Postal Codes instead of Zip Codes and the abbreviation for the provinces.

Note the options for output of your file.  I would suspect that CSV and SQL would be most popular but you’ve also got options for HTML, Excel, and XML.  If you require additional rows for your data set, just add them.  When you’re good to go, click “Generate” to get your results!

Your data is good to go.  Save it and you’re ready to mark.  I would suggest running a few versions so that you have a choice of datasets with those parameters.  It only takes seconds.

The website will generate datasets of 200 records.  If you like what you see, the script is downloadable for installation on your own server.  For a small donation, you could get into the realm of big data and create a dataset of 5000 records.

What could be easier?  If you’re a Computer Studies teacher, do yourself a favour and bookmark this one today.

Powered by Qumana