Statistical Literacy for Librarians

Last month I attended the University of Alberta Data Library’s Winter Institute on Statistical Literacy for Librarians (WISLL).  Despite my initial terror of statistics, I found WISLL to be a fun and valuable experience that I'd like to share with Chinook Arch!

http://dilbert.com/strips/comic/2004-10-14/

A Bit of Background - Data vs. Statistics

The terms are sometimes used interchangeably, but Data and Statistics are different things. Data is unpublished, not display-ready and requires processing. Statistics are derived from data, they are presentation ready, published and present facts & figures. A statistic can’t be a statistic without data.

I like to think of data being the ingredients in the recipe – statistics the finished recipe. Just like poor ingredients will make a poor recipe, it’s hard to go wrong with good quality ingredients, the same goes for statisics. 


Types of Statistics

There are two types of statistics - official and unofficial. Official statistics come from government bodies like Statistics Canada or the UN. Unofficial statistics come from places like trade associations, marketing companies, academic institutions. Both are useful.

Why is statistical literacy so important?

Look around and you will notice stats are increasingly ubiquitous. Stats are everywhere, about everything. Open the newspaper, watch the news, follow a link on twitter and you will find some bit of information with a stat attached.  Not only are we seeing written statistics but with the growing popularity of data visualizations we are seeing more and more infographics. On the reference desk we can help our users to understand where these claims come from, and help guide them to the primary sources and data sources.  

When we are writing grants and annual reports understanding how to pull together quality stats from quality data is key. 

We can use stats to tell a story. 

Here is a fun example:

Just before Valentine’s Day I kept hearing talk of Lethbridge having been declared as one of Canada’s most romantic cities.  Let’s say you have a user come to you and ask:  Hey, I heard Lethbridge is one of the most romantic cities in Canada, is that true?

An initial goolgle search brings up these secondary sources – they will help you find the primary source.

Here is a link to a quick article from a local media source 94.1 CJOC.

This article ran in several Canadian publications:

The second article does well in explaining the who what and where of the statistical claims. Now that you know that the statistic came from - Amazon a bit of internet digging will bring you to the Amazon press release page (the primary source) 

So how did Amazon create the statistic of Lethbridge being the 10 most romantic city in Canada? 

“This list was compiled by comparing Amazon.ca sales data of romance novels and relationship books (Kindle books and print books), romantic comedy DVDs and a collection of romantic music CDs from Jan. 1, 2012 - Jan. 23, 2013 on a per capita basis in cities with more than 80,000 residents*. 

New for 2012, in addition to including CDs by B.C. crooner Michael Bublé, Amazon.ca expanded its musical search to include some additional voices of love.  They include the legendary Dean Martin, the baritone of Barry White, the sexy Luther Vandross, the neo-soul sounds of Maxwell and the newest R&B sensation Miguel. “ 

*Data for cities was based on Statistics Canada’s 2011 “Population and dwelling counts, for Canada and census subdivisions (municipalities).” (Link)”

Amazon is transparent in their explanation of where they got the data and how they used it to create the list. They used unofficial sources (their own sales data) and official sources (Stats Can) But does this really mean that Lethbridge is as romantic as Amazon claims? They certainly are using data and statistics to tell a story – if that story is quality is probably debatable. However, you now have the path back to information about the data that was used to create the statistic that you patron was asking about.

What do you think? Is Statistical Literacy and important component of Librarianship?