On Wed, 22 Jan 2003 10:22:24 -0500, Stan Brown <[EMAIL PROTECTED]> wrote:
> Eric Lund <[EMAIL PROTECTED]> wrote in sci.stat.edu: > > > We really do not want to score all 1600 if we don't have to. > [snip] > >So, what I would like to > >know is, can we get an approximation for the population mean without > >having any baseline data, and if so, how many students would need to > >be sampled? > > [Please post right-side-up and trim quotes; see > <http://web.presby.edu/~nnqadmin/nnq/nquote.html>.] > > This prompts a further question. Putting the statistics aside for a > moment, I'll bet that if Indiana mandates you develop the test > locally the state also mandates what sort of data analysis you can > do and whether you're allowed to score only a sample of tests. So be > sure to eliminate that possibility. [ snip, rest] BE SURE to check on that possibility. It is my guess that the law and regulation either spelled it out, or else it is a *really* crappy law and regulation. A decade or so ago, my employer (hospital) called on me for advice about meeting new regulations for meeting Medicare payment standards, or some-such. These top administrators tried to describe in *general* terms, what standards needed to be met. I think it took a few hours before I figured out that they were - looking at/ sitting on/ staring in the face - some requirements and instructions that spelled out the precise sampling in a whole lot of detail. I forget now... I think the administrators might have focused on the Introduction, which stated the intentions of the regulations, whereas the Body of it went ahead and spelled out the complete measures. - the Bush administration pushed through some laws on measuring schools and students, which seem to combine the worst of the possible worlds -- according to the political commentary that I've believed. That is, they mandate a whole slew of measurement, at great cost and waste of time. It will be the waste of time because (a) they can't measure the progress that they need to measure, trying it that way, and (b) there are nasty side-effects when teachers and schools feel they *have to* 'teach the test.' Which they will. I think that the state-of-the-art says that it is very difficult to measure the quality of a school, and that it takes a few years at the very least. It might also become nearly *impossible* after you factor in 'privacy' concerns, since you have to follow the continuing progress of individuals -- if you are to achieve any good degree of confidence. So, good luck, with whatever you have to measure. Don't be surprised if the effort, all-in-all, seems stupid; it might arise from a political agenda rather than need. -- Rich Ulrich, [EMAIL PROTECTED] http://www.pitt.edu/~wpilib/index.html . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
