I was bragging about my Associate Director, Rochelle Logan. Rochelle used to crunch numbers for the Colorado State Library, and speaks nationally on the topic of library statistics. I was enthused about some of the new reports she's cooked up. Having an acknowledged statistical expert on staff is handy.
But the colleague I was bragging to was full of dark forebodings. "Beware the tyranny of numbers," he said.
"Whatever are you talking about?" I said. "I use statistics for all kinds of things. They help me track trends in use. They help me find trouble spots and -"
"Exactly," said my colleague. "Trouble spots. When you get statistics, you get comparisons. When you get comparisons, inevitably, you get competition. And then you say to one manager, 'Gee, your program stats are much lower than everybody else's.' And then that manager starts manipulating the statistics to look good."
"Pish," I said, "and tosh. It may be that some libraries act that way. But then the problem isn't with the numbers, it's with how they are used."
"Are you saying you DON'T compare the statistics between your branches?" "Sure I do!" I said.
"And you don't follow-up with the managers to find out why there are differences in certain kinds of costs or library use?"
"Well, yes," I said, "but that's useful information for everybody. I encourage our managers to experiment, to try different things at different branches. It doesn't do any good to experiment, if you don't honestly report the results. Sometimes it turns out that a program, or service, catches on at one location, and bombs at another."
"My point exactly," said my colleague, "so that means the place where it worked, you had a better manager, right?"
"I don't see it that way," I said. "Different strategies yield different results. My managers get together pretty regularly to talk about this stuff. Sometimes one key factor turned the trick: advance publicity, for instance. If one manager learns something that worked well, the other managers give it a try. But more often than not, I learn something I've suspected from the beginning."
"What's that?"
"That each of our branches operates in a unique community. Library statistics do say something about the library, but they say even more about the community."
"What do you mean?"
"Well, the fact that one branch has a high use of educational software clearly has something to do with the fact it launched the service, and our other libraries don't have any yet. But let's say that both libraries offer an identical program for seniors, or for bilingual families. In that case, the number of people who show up for it reflect not the library offering, but the demographics of the community."
"How do you sort out which is the library and which is the community?" he asked.
"More numbers, for one thing. The 2000 census stuff will be very helpful. But also, more experiments," I said. "Say a young adult program doesn't catch on one year. But it might the next. The only way to find out is to model the experiments on successes elsewhere in the district — but also on things that haven't been tried yet."
"What about costs? Once you start tracking all this stuff, don't you feel pressured to find cheaper and cheaper ways to do them?"
"Well, I'm a fiscal conservative, if that's what you mean. It's certainly not my goal to find the most expensive way to run a library."
"Aha! So you do manage by spreadsheet. The tyranny of numbers!"
"But again, sometimes you have to poke the numbers to find out what's behind them. Suppose, for instance, that staff costs are way lower in one branch than another. The reason? Way more part-timers. But it might be that having too many part-timers drives up other costs -- supervisory time for evaluations, training time for new employees, more benefits administration. That's important information."
"But if all that numbers do is drive you to look for more numbers, then what's the point? I say again, you're slaving under the despotism of data."
"Catchy," I admit. "But ultimately, that's just paranoid. What's the alternative? Ignorance? The purpose of gathering and analyzing data is to understand your environment. It may not answer all your questions, but it can at least get you to ask better questions."
My colleague sighed. "So you're going to keep running your ship according to the numbers?"
"'Fraid so. But not just on the numbers."
"Oh? What else?"
"The good judgment of my staff."
"Even when it contradicts the spreadsheets? How come?"
"Simple," I said. Then I smiled. "I've learned that I can count on it."
No comments:
Post a Comment