Friday, August 19, 2011

"You have much to learn, Grasshopper." by Tom Bruno

It had been a few weeks since I'd been hired to my new job- my first position as a professional librarian- and my boss had asked me to produce some numbers for a project she was working on.  I remember first cutting and pasting an Excel chart into the body of an email, resulting in a poorly-formatted mess of misaligned figures and dislocated column headings.  My boss quickly suggested that I send her the actual spreadsheet instead, but no sooner did I do just that than I found myself summoned to her office for my first lesson in how to make my data dance. (I wish I could claim the credit for this beautiful turn of phrase, but I first heard it from Colette Mak, Head of Resource Access and Delivery at the University of Notre Dame the High Priestess of ILL data.)

"Rename your columns so they make sense to someone who is not you," she said.  "Use colors to delineate one set of data from another. Remember that your data is telling a story- if you want that story to be taken seriously, you must pay as much attention to presentation as you do to the data itself.  And please use formulas next time!"
Humbled, I returned to my cubicle and edited my spreadsheet following my boss' guidelines.  The result was a professional-looking report that even a librarian unfamiliar with the inner workings of resource sharing would be able to interpret, an order of magnitude more readable and user-friendly than the hairball of numbers and figures I had coughed up on the first try.  This was of course only just scratching the surface -- for example, I had yet to learn about the wonders of pivot tables -- but the lesson had been absorbed.  I may have been clever enough at that stage of my career to generate the data I needed, but I knew next to nothing about how to share that data effectively with my colleagues.

Indeed, I did have much to learn!

While my whirlwind introduction to the professional ranks of librarianship had been one epiphany after another, for some reason this particular revelation has stuck with me.  Why?  Because of all the things I learned during my first year on the job as a librarian, this is the one for which library school had prepared the least.  I can't tell you how many papers I had to write and presentations I had to give on the way to getting my MLS, but I don't remember being asked to create one spreadsheet or draft one statistical report.

This isn't just a case of my library school being behind the times, either-- when I entered the LIS program in 2004, Simmons had already crafted a technology orientation requirement, and during my time there I had coded finding aids in XML and used other cutting-edge professional tools such as the Catalogers' Desktop.  While there was (and still is) a course in database management, the emphasis there was on designing your own library databases, not how to produce, manipulate, and present the data within.  I’m not an isolated case, either. Nobody I asked had any spreadsheet assignments in graduate school.

Where is this disconnect coming from?  I suspect it has something to do with the nature of library statistics, which often come directly from the vendor or are produced by non-librarians somewhere else within your organization (such as your IT department).  Resource sharing librarians are awash in data, so we’ve got more experience with it. However, as libraries increasingly make policy and procedural changes-- not to mention budget decisions!-- based on statistical analysis, outsourcing your stats to a third party is a potentially dangerous prospect.  Why?  Because your data is telling a story.  If you do not understand where your numbers come from, not only do you run the risk of misunderstanding that story yourself, but you render yourself powerless when someone else decides to read your data in a different way.

I have a sign in my office that asks three questions of people who want statistics from me:
  1. How long ago did you need these numbers?
  2. How do you intend to misrepresent this data?
  3. If I just make something up, will you know/care?
This is a joke, like the ubiquitous "You Want It When?" signs in auto repair shops, but there's also a kernel of truth buried in their somewhere.  In my mind there is nothing more scary than the "drive-by"
data request, because there's little time to make sure that the numbers you produce actually mean what they say, nor do you have control over their distribution once they leave your hands.  Alas, these are exactly the kinds of reports that you will be asked for as library decision-making becomes increasingly data-driven.  How, then, do you ensure that the data you do offer up is sound enough to base a decision on?  Simply put, you need to know your numbers as well as you know every other aspect of your job as a librarian.


So what are the takeaways here?
  1. Own your data.  If it currently lives in some vendor's proprietary black box, see if you can find a way to get your own access to it. Vendors often mean well, but I've never met a canned report that I haven't wanted to refine in some way, shape, or form.
  2. If nothing else, learn how to use Excel.  Buy a book, take a class, whatever you need to do, but if you can't bust out a pivot table then you're missing the point of a spreadsheet.
  3. Learn how to use Microsoft Access, or buddy up with a guru.  I did both.  As a result I can produce a lot of the on-the-fly queries myself when my bosses coming looking for data, plus I know enough to explain my new ideas to my guru quickly and intelligently. 
  4. Presentation matters as much as the quality of your data, so be sure to make it dance.  (But Melvil Dewey help me if I get one more request to reformat a spreadsheet so that "it prints out nicely"!)
  5. Your data is telling a story.  Your decisions and those of your supervisors will be increasingly dependent on the statistics you generate, so it is imperative that you understand where these numbers come from and how they behave.  My stats guru friend jokingly refers to me as the "Request Whisperer" because I can look at an outlier in our resource sharing data and immediately tell her what went wrong with the ILL request in question.  Listen to your data and learn from it, so that you can be an effective interpreter of its meaning. 
Good luck, Grasshopper!


Tom Bruno is the Head of Resource Sharing at Widener Library of the Harvard College Library.  He has worked for the Harvard University Libraries for over 12 years.  Tom received his Bachelor of Arts in Latin and Greek at Boston University and his Master of Library and Information Science from Simmons College- every
now and then he thinks about getting a Ph.D. His blog is The Jersey Exile, and he tweets at @oodja.

1 comment:

  1. I heard the comment: "You've much to learn, grasshopper" on a program on White Collar and knew I'd heard it before but not where. Kung Fu was a favorite TV show years ago and that quote was used in different ways, i.e., "You've learned much, grasshopper."

    ReplyDelete