I finally have a clean set and am doing correlations. I’ve got a clean ‘change in visits per population’ from 2000 to 2008, with 8624 libraries in the sample. Of the ~9500 total population, libraries were removed because they had zero population, zero budget, zero visits, or they didn’t exist at either end. (opened after 2000, closed before 2008).
So far I’ve had one assumption shattered. Actually, it’s the implication of a claim made by the ALA. See, I got the impression that almost all libraries were seeing increased service. No, 27% of libraries have had declining visits per capita. Over 1/4. Sure the majority have seen an increase, but not “almost all”.
I’ve also further confirmed some of the tentative suggestions from earlier. Neither changes in computers nor actual numbers of computers have any impact on increased service. The same goes for population size or total budget. There’s an increase in funding per capita from sources other than state and federal but it’s less than 15% correlated, and I’m not sure if it’s cause, effect, or parallel consequence.
I’ve still got a dozen or so stats, direct and determined, to evaluate. Right now, though, there is nothing to say “these are what growing libraries are doing.”
That quantitative survey may be more critical than I thought.