Thursday, October 9, 2008

"We're number 1...or number 50...or in the top 10% or" ?!?

The annual HAPLR rankings have been released, prompting a spate of local press coverage in communities where the libraries found enough HAPLR justification for bragging rights. I don't begrudge any library being able to show reasons why they're good. It's a challenge to find ways that quantify the positive impacts public libraries have in their communities. But we have to keep ranking systems in perspective. I sympathize with any librarian who may be criticized for having low rankings, since ranking systems have built in biases.

HAPLR, which is Hennen's American Public Library Rankings, has been published for years by Tom Hennen, the Director of the Waukesha County Federated Library System here in Wisconsin. I've told Tom since the start that I'm not a big fan of his rankings, which are based on national statistics gathered by the Public Library Statistics Cooperative. The reason I don't much like them is that I believe they emphasize circulation of popular materials and totally disregard library programming and electronic service delivery. I've written about this in the past ("Innovators suffer under HAPLR", Jan. 2007)

In the October issue of American Libraries magazine, Tom posts a spirited defense of his rankings. He makes a good case for why it's important to use input measures as well as output measures, and how weighting numerous factors in a complex formula helps paint a complex picture. The new Library Journal ranking system looks interesting, though it has not yet published rankings. The LJ system intends to include programming (a strength here at APL) and public workstation use -- just about the only objective technology measurement we've been able to agree on as a profession.

It's hard to predict whether APL will fare better or worse in the LJ rankings. I know that we have very few computer workstations relative to both population and demand. That doesn't matter to HAPLR, but it will in the LJ rankings. Most importantly, it matters t our patrons! Any ranking system will reflect some arbitrary choices by those designed it.

For the record, our library's scores in the latest HAPLR:
  • raw score 762
  • percentile 90%
Rank relative to other Wisconsin libraries:
  • in our population class #1
  • of all libraries #50 (of 381)
  • of public library system resource libraries #7 (of 16)
% scores of libraries in the area
  • Neenah 94%
  • Menasha 96%
  • Oshkosh 90%
  • Appleton 90%
  • Kaukauna 80%
  • Kimberly/Little Chute 67%
  • Brown County 81%
These are all pretty good scores, which is fine because these are all good libraries. Is Menasha the best library around here? Maybe -- it depends. Like a lot of statistics, I find the trend lines over time more interesting than a single year snapshot. The most important point is that measurements and rankings are arbitrary.

A few years ago, I visited a top-rated HAPLR library and I was expecting to be blown away by a paragon of library service. I found a relatively run-down and messy place and felt confused. But it had a lot of tourist use with a small permanent service population, driving up all the per capita outputs heavily weighted in HAPLR, and it had a lot of donated materials, substantially reducing the cost per circulation. Ya gotta love statistics.

Each community is the measure of its own library.


Unknown said...

It's not clear to me from your article, but I hope you heard that the recent American Libraries article was published using the same data as last year. Tom admits to making a mistake, but it wasn't caught before the magazine went to press. See

Terry Dawson said...

Yep, I used the corrected information Tom has linked to.