Search This Blog

Thursday, September 06, 2007

Discount madness

Man, USAC needs to do some training on how to calculate your discount. I just saw an appeal by yet another school district extrapolating their NSLP (USDA National School Lunch Program)numbers based on the percentage of NSLP forms returned.

For those of you who don't know, USAC rules allow applicants to send out an income survey, and if you get half of them back, you can make a projection based on the percentage of low-income students in the returned surveys. So if half the kids return the survey, and 75% of them are low-income, then you can say that 75% of the students in the school are low-income. Nice!

But you can't use the NSLP application as a survey and make a projection based on it, and I think this is like the 3rd appeal I've seen where some district did just that.

Why can't you use the NSLP form? Here's an analogy: Let's say you sent a survey home to parents, asking if their child was a boy or a girl, and said that you would give $250 to anyone who returned the survey if their child was a girl, but $0 if it was a boy. Of course 90% of the responses you got would say "girl"; the parents of boys just wouldn't respond

The NSLP form is like that: Low-income families who return the form, you get about $250 worth of free lunches (and maybe free breakfast, too). Families that are not low-income get bupkus. So of course the district doesn't get a random sample. They get almost all the low-income families responding, and almost none of the non-low-income families. It's a non-random sample.

I think the USAC Web site needs to state outright: "You can't use NSLP applications as a survey instrument." And put it in a NewsBrief, too.

No comments:

Post a Comment