Apparently Medicare is just dicovering what we’ve all known in library science for decades: satisfaction and accuracy are not the same measure! A GAO study of 1-800-MEDICARE, found accuracy rates of 61%. Remarkably close to 55%, I’d say. Indeed, their methodology was practically lifted from Hernon & McClure (1986): test questions developed for the study, asked by proxies, of different answerers.

And then comes my favorite part: Medicare responds with the finding that 85-90% of callers are satisfied! Well of course they’re satisfied; they got an answer! Even if it’s dead wrong. And how is the caller going to know it’s wrong anyway? I mean, if they knew the right answer, they wouldn’t be calling the help line, now would they?

Can there be any question that we need to educate the public better about what we do in this profession? We in LIS could have told you 15 years ago that satisfaction ≠ accuracy! No one asked us, though, and so the GAO and Medicare have to re-invent methodology for evaluating reference services.

Medicare hot line fails government test, from MSNBC

The GAO study is here. I’ve been unable to find Medicare’s own survey with the 85-90% finding.