|
|
Interversity Quotes"I think it's time to ask the Standardistos: Where's the test for compassion? For honesty? For curiosity? For moral commitment? Until they can answer, let's tell them to shut up." —Susan Ohanian (One Size Fits Few: The Folly of Educational Standards) Interversity ProjectsEchoes of Orwell:The purpose of this webforum is to promote a lively discussion among students and faculty alike about George Orwell's 1984 and the parallels or echoes found in today's society. |
[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index] Re: CTA President on state assessment system
At 06:48 PM 12/1/2002 -0500, Karen Cole wrote: As I understand it, a score on a norm-referenced test compares the test It's conceivable - we can imagine it, but it can't happen, at least until an infinite number of monkeys have typed the complete works of Shakespeare. Considering the laws of chance, there is a significant probability it will happen in some classes. It can happen in a whole school if the population now taking the test is sufficiently different from the norming population. If the number of test takers is large enough, you should expect some to fall below the 50th percentile no matter what. There actually appears to be something like the opposite of the Lake Woebegone effect with the SAT-9 at the high school level. In all the states I've heard about, the average student score is below the 50th percentile. Perhaps the norming group was not really representative. At all grade levels, California's student population is significantly different from the norming group for SAT-9 - with a much higher percentage of students from low-income families and an immigrant population that is orders of magnitude greater than the percentage of immigrants in the norming group. Thus, for the average score in California to equal the "national average" - the average score of the norming group - would be a tremendous accomplishment. When we talk about all students scoring above average we are in the realm of what George Schmidt calls "unicorn studies." George Sheridan -------------------------------------------------------------------------- To unsubscribe from the ARN-L list, send command SIGNOFF ARN-L to LISTSERV@LISTS.CUA.EDU. Data collection is not a viable alternative to human engagement.
Post a Message to arn-l: |
eric |