Quote:
Originally Posted by KHall
So here is the skinny: If you believe something to be true, prove it. Create a hypothesis, design an experiment to test that hypothesis, collect data and analyze the results to see if it supports or denies your hypothesis. That is the essence of science. Everything else is arm waving.
The next step would be to have the results peer reviewed and published in a reputable journal. This is where I believe we fall short. We can understand the basics of designing a non-biased test that will yield statistically significant results. We can separate and measure quantitative and qualitative variables. Psychologists measures attitudes and feeling such as passion and initiative all the time, its been done, published and accepted. Teachers measure learning with much the same results.
|
I think part of the issue you'll have is the level of proof. Engineers and most scientists must be very sure of their data before we publish/build. When I took statistics, the professor mentioned that Psychologists get excited about r^2 values of .15 (for those you that haven't taken stats, that means 15% of the variation in the output can be explained by the input). If I was trying to characterize a dataset as an engineer, an r^2 value of .15 would get me laughed out of the room! On the other hand, if the Education Ivory Tower folks accept an r^2 of .15 (and it sounds like they do) then this works in your favor.
Also, I'm pretty sure the NWEA tests are not particularly difficult to administer. My mom is a Teacher's Aide/School Librarian for an elementary school and has been tasked with them many times because she has some computer savvy (i.e. knows how to use one). It's just a webpage that the kids click through, and the difficulty of the questions scale with how good they are at answering them.