Wikipedia is the single-most common reference site we share with patrons. I don't really have a problem with Wikipedia as a source, I just want to see librarians using the full range of their abilities.

I wondered, are patrons more satisfied or less satisfied when we share Wikipedia?

In 2006, 148 of 861 survey-takers said they were not satisfied with our service (17%). We can be sure of this figure within 3%.

In those sessions where we shared Wikipedia, 6 of 29 were not satisfied (21%). We can be sure of this figure within 18%, or rather, we can't be sure at all.

I think this will help me sleep, though - even though Wikipedia is the most common site we share, it comes up once in only 20 transcripts and so many reference sessions use sites that are not Wikipedia my fears of librarians not using their skills are unjustified.

Related, we can be sure that people are only half as likely to fill out a survey if we share Wikipedia (3% compared to 6% overall). My guess is that survey behavior and Wikipedia are actually unrelated - it's probably that some group of users is filling out surveys less often, and that group is also the ones we're sharing Wikipedia with, i.e. K-12 students.

Wikipedia entries, other than the main page, that show up in at least 5 transcripts (since July, 2004):

13 Ferdinand_Magellan
12 Cotton_Mather
11 Christopher_columbus
9 Vasco_da_Gama
9 Plessy_v._Ferguson
9 Twins
8 John_Cabot
8 Bombing_of_Dresden_in_World_War_II
7 Inuit
6 Samuel_adams
5 Marco_Polo
5 British_colonization_of_the_Americas