[Image courtesy of bradmarley]
I've already highlighted some of the terrific presentations that occurred at last week's GeekNet in Minneapolis ... very quickly, here are a few quick hits on short but important discussions that also occurred:
A/B Testing - You'll remember I profiled the concept of A/B testing a while back ... so I asked the group at one point what they thought about the idea. The consensus was that, unlike the giant online sites that have essentially infinite visitors to their sites, election websites likely aren't going to get the traffic necessary for A/B testing to make sense or be useful.
Wait times - During an open discussion about data, the subject of whether (and how) to measure voter wait times got a lot of airtime. While there was a rough consensus that asking people about wait times in hindsight is generally not ideal, there was not a lot of consensus on what (if anything) should replace it. In short, the challenge is to find a way to measure how long people wait to vote without somehow compromising their privacy or adding still more complexity to the voting process. I heard some discussion of "geo-fencing" - a concept that is increasingly popular given the rise of smartphone-based geolocation applications - as well as slightly lower-tech ideas of barcoded documents or other techniques to capture start and end times of different aspects of the process. Finally, almost everyone agreed that defining "wait time" would be important; is it the time it takes to get from the back of the line and cast a ballot, or is it a door-to-door "time to vote"?
Access (and fees) for data - one other fascinating topic of discussion centered on the procedures for obtaining election information in different states. Not surprisingly, experiences run the gamut - but at least a few states indicated that their work to make large data sets available in a defined format had enabled them to get out of the business of generating custom reports for different players in the election process. There was also a spirited discussion of the fees that are required in different places; some states said their policymakers and public would never accept fees for obtaining such data, while others saw such fees as a vital source of secondary revenue and in some cases a deterrent for time-consuming data requests.
All in all, a great week - thanks again to everyone for taking part, and especially to The Pew Center on the States and my colleagues at the University of Minnesota for making GeekNet happen!