a bit a really look at the data with the knowledge of a statistician and you see something very different. You might as well be extrapolating GDP trends from the results of a Cosmo or Glamour survey about what men want in bed...
The biggest problem with metrics are the strategy behind measurement (why we're measuring and what it will be used for); the definition of the variables being measured) and how the data is collected and analyzed. The amateur statisticians in the world - and there are many - recoil when told that their strategy, variables, data collection process, data analyses, and presentation vehicles are flawed, and lash out when told that the validity and reliability of their "project" are dubious at best.
If you've read this far without becoming catatonic then remember what Claudia wrote... IF YOU CAN'T MEASURE IT, YOU CAN'T IMPROVE IT.
Diana, you'd be surprised how much statistics there is in intuition. But the more quant data we have - and smart people really have a hard time with this - the more a statistical model and solution separates itself from
humans in identifying trends.
Did you notice that I said "quant"?
Because recruiting is generally and historically so poor at collection and measurement when old farts (FYI, I'm getting my AARP membership card next month) talk about metrics, it's usually accompanied with the same face as when one has a case of hemmaroids or acid reflux.
I say fight it the urge to medicate!
Let the data set you free!
Lookat Crispin's and Mehler's Sources of Hire surveys and tell us if you have this data for your operation. Wouldn't this make a difference?
[FYI, in my doctoral program, one of my foci was in quantitative psychology - three solid years of univariate and multivariate statistics and two years of research design. I'm by no means a card carrying member of the ASA but I can say for certain I'm more advanced than the average recruiter]
Sandra McCartt said:Trends would seem to make a lot more sense than the tactics and drill. Bottom line is if it works do it.