Hi Doug,
I am not at all sure that a 3 point difference is significant. It could well be random fluctuation. Without knowing something about the standard error and 95% confidence interval limits, it is very difficult to say anything at all about a 3 point difference. Furthermore, there's something else to factor in: the difference between the previous test and the present version, keeping in mind that there is a standard error associated with the testing tool itself. That is, the ability of the testing instrument itself to accurately measure the thing it is supposed to measure is always imperfect. If you go to the store and buy two measuring cups, they will not measure out 1 cup of something with equal precision, though they will measure it and argree with one another closely enough for most practical applications. But with tests, that precision and reproducibility in measurement is much less certain. There are many difficulties in test construction such that it is not at all a straightforward task to measure some trait or extent of knowledge and so there is a fairly large amount of uncertainty and error variance.
So you have several possible unknowns here that could affect the "3%" in unpredictable ways, and because of these and the probable magnitudes of the error variance associated with them (compared to the 3% measured difference), it could even be that today's HS seniors are better than the previous batch.
So, hard to know what to make of the specifics in the results.
But you and I agree on the basics: our public school system is a disaster. And whether it is 18% or 20% or even 30%, any way you slice it is completely abysmal and a tragedy.
Got a call from Doug (!!), so I headin' for SB tonight.... you in??
T |
