Ban Individual Student Testing Forever!
Recently, during a Washington Post panel discussion, Montgomery County Public Schools Superintendent Joshua Starr called for a three-year moratorium on standardized testing used to evaluate teachers. Click here to read more about the ban proposal.
I’m puzzled why this is news.
Starr leads a public school district that does not use student test scores to
evaluate teachers. MCPS already has a moratorium in place, and the moratorium
or practice—or whatever word one wants to use to describe what MCPS does—was in place before Starr came to Montgomery County. And, by the way, I have no
problems with the ban.
With the ban in place, MCPS takes itself out of the running for any of the federal Race to the Top money. But then one could make the argument that MCPS is such a well-funded school district anyway—why bother getting bogged down in those politics? And frankly, no matter how many times Starr publicly says using test scores to evaluate teachers is bad science, he is not really going to change federal policies.
Does Starr really believe President Obama is listening? Click here to read more about Race to the Top.
But, if Starr wants to make real news, why not just propose a permanent and complete ban on standardized testing of individual students in MCPS? (Important: individual here means all or every student.) That’s right—a complete ban. Why stop at three years? With a complete ban in place test results can never be used to evaluate teachers.
And so, I’m all in for the complete ban in exchange for the following:
- Annually, test random samples of students in reading
and math. One could limit this testing to grades 2, 4 and 6. - Periodically, at certain grades—grades 8 and 11, using National Assessment of Educational Progress exams, test random samples of MCPS students and compare MCPS student performance to national outcomes. Click here to learn more about NAEP.
- In our high schools, using the Trends in International Mathematics and Science Study, test random samples of students so we can compare our students to global outcomes. Click here to learn more about TIMSS.
Because all of what is proposed here is limited to random samples of students—impacting very few students, results could never be tied to individual teachers or used to evaluate teachers. And the beauty of this strategy is one could use results to continue monitoring MCPS progress on closing achievement gaps. A total win-win!
JM
12:33 pm on Thursday, January 31, 2013
Random testing might work for evaluating high level education policy, but even there I would worry about educators influencing the random sample to influence the results. We already know that cheating happens (see DC, Atlanta, etc), so it is not a stretch to think that a certain group of students would end up on a field trip the day the testing students are picked.
My biggest worry is for the students themselves. Testing is very important to GT and LD students. The system likes to tell parents that their kid is a normal average kid that is doing fine. It really is nice to have a test that says, 'no my kid is above or below average and this kid needs help'. If you have an otherwise smart kid, that seems to really struggle in reading, it is great to see where his/her reading is benchmarked to his/her peers. As a parent, I have no clue without the test. I would have to listen to a teacher give a qualitative judgement.
That is why I really like testing as a parent. I am happy to not use testing for teacher evals, but I think it is vital for policy and for helping individual kids.