Thursday, June 26, 2008

Will duct tape cure PubMed? Simplicity vs. the Curse of Knowledge

A designer knows he has achieved perfection not when there is nothing left to add, but when there is nothing left to take away.

~Antoine de Sainte-Exupery, quoted in
Made to Stick: Why Some Ideas Survive and Others Die by Chip Heath & Dan Heath (pg. 28)

Made To Stick has a large piece of fake duct tape as part of the dust jacket. Its case for simplicity as part of the measure of SUCCESs (Simple, Unexpected, Concrete, Credible, Emotional & Stories) for both creating and communicating good, lasting ideas has really struck a harmonic chord with me. Unfortunately, another good C (concise!) tends to elude my writing but I keep striving towards it.

I believe the recent changes in the PubMed's Automatic Term Mapping (ATM) came along as Unexpected for most of us in the medical library field. The source institution for PubMed (National Library of Medicine, or NLM) is certainly Credible, and the changes have inspired plenty of Emotional Stories that I will not rehash here.

Concreteness has been left in the dust. More than a month after the changes were unveiled, we have little from the source of Credibility except the midsection of a 25 minute conference video to answer the traditional journalism Who/What/Where/When/Why/How? (5 Ws/H) and a blurb in the Help page about these changes.

The intent may be to make PubMed queries easier for users to retrieve relevant results, but in reality they are anything but Simple. I want to clarify: Simple here does not merely mean the lowest common denominator, but one of "elegance and prioritization, not dumbing down." (Heath and Heath, p. 30) From the breakdown of the ATM to include so many different attributes to increasing consumer health education needs, there have been many protests of 'dumbing down' things when in reality those protesting often have the Curse of Knowledge.

The Curse of Knowledge... [is] the difficulty of remembering what it was like not to know something. Accuracy to the point of uselessness is a symptom of the Curse of Knowledge... People are tempted to tell you everything, with perfect accuracy, right up front, when they should be giving you just enough info to be useful, then a little more, then a little more. (Heath and Heath, p.57)
Sound familiar? How many of us have had users' eyes glaze over while providing them with information that we believe to be both concise and accurate? If there are any librarians out there who have never had this happen, please share with me how because I'm still learning the fine art of balance in this department.

Please, keep sending in feedback directly to NLM (at the bottom of each PubMed page) that is Simple and Concrete along the 5Ws/H! While there is always room for improvement, I have to believe that the decisions to implement PubMed changes were based on valid website usage/statistical analysis and (hopefully) usability studies before these changes left development and not because they wanted to torture medical librarians. NLM is apparently not receiving all that much feedback from anyone about the changes, and then only from librarians. Generally that means other users aren't finding much to gripe about since it's a rare day when someone submits feedback of 'Woohoo, awesome!' about anything.

Medical librarians are not the sole users of PubMed, and we have a whole arsenal of Cursed Knowledge to refine queries for us as we both assist and teach users ([MHDA], anyone? (seriously)) and that's ok. We don't want eyes to glaze over explaining everything about PubMed to our users at once. Conducting a good health reference interview, determining what filters and tabs are appropriate for the information needed, assessing our users' comfort level with query construction (I think far too many assumptions are made in this area but that's a post for another time), then suggesting and teaching appropriate modifications is the way to go. We want to provide just enough info, then a little more, and keep them coming back to us to keep learning and refine things as needed. 'Just in time, not just in case' is far more than a collection development policy.

As an FYI my duct tape and I are heading out until mid-July so things will be temporarily quiet around here, but don't be surprised if you see us pop up somewhere else in the biblioblogosphere as a guest over the next few weeks!


Rebecca said...

At MCMLA '06 in St. Louis, a veteran librarian commented at a panel discussion that PubMed will eventually be run by Google - or at least be like Google. I'm thinking this change in ATM is proving him right.