Monday, May 02, 2011

More thoughts on STAR testing

As we enter week two of STAR (standardized) testing in California, I have a few more thoughts. Actually, Pennamite and I came up with these ideas together.

First, listen to Diane Ravitch's interview on Fresh Air. She spoke about the intent of standardized testing--to collect data--and the results of standardized testing--to punish schools. It was never her intention that standardized tests would be used as a blunt instrument and that they would become a tool for those who aim to privatize the US public school system.

After my What the STAR tests really tell us post, Pennamite suggested that schools (and school districts) that are known for their ultra high test scores become prisoners of their past results. They must get ever higher scores--or at least stay flat--or risk the wrath or displeasure of the parents.

It takes a bit of massaging to maintain such steady upward progress in the face of statistical noise and educational policy and demographic changes. Remember, people's jobs (and pensions) depend on the pleasure of the parents. Take a look at the math classes taken by 8th graders in California, Redondo Beach and Manhattan Beach.
Former Governor Schwartzenegger wanted to have ALL California 8th graders to take Algebra or beyond. All is a tall order, but most districts are trying to comply.

Statewide, 57.3% of 8th graders are taking Algebra 1 and another 4.8% (have already taken it and) are taking Geometry. 62% are meeting the new goal.

In RBUSD, it's 57.3% and 18.9% respectively. 78% are meeting the new goal. RB parents are wealthier and better educated than the statewide average (but not by much), so these numbers are not that surprising.

In MBUSD, it's 35.0% and 9.2% respectively. 44% meet the new goal. Despite having some of the wealthiest and best-educated parents in the state, the students are taking lower level math classes than the state-wide average. They are a full grade level behind comparable schools with similar demographics.

This is even more alarming because MBUSD used to exert strong pressure on parents to hold kids born after April back a year. See Implications of academic redshirting.

It wasn't always this way. Look at the data from prior years. The rest of the state leapt to accelerate students in math while MBUSD moved more slowly. They are where the rest of the state used to be.

Pennamite pointed out that the ultimate losers in this game are the kids. They are being held back until the school is sure that they will perform very well on the standardized tests. The default is to not challenge/stress the students. Otherwise, the kids might not score so highly and the schools will face some angry parents.

There is some grumbling on the soccer fields about the lack of challenge in math curriculum in MBUSD. One MB mom who has her kids on permit in RB has cited that as the main driver for leaving their school district. She works in the education field. I wonder how long it will take for the rest of the parents to become aware of how fixation with test scores is shortchanging their kids?

A (IMHO) good use of standardized test scores:
The RBUSD superintendent and the administrators of RBUHS are concerned that the middle schools are pushing the kids too far and too fast into higher level math. When 9th graders take Algebra 2, some of them do very poorly in the class. Would they have done better with an extra year of maturity? Did the system set them up for failure?

The standardized test scores support that view. The average test scores of the kids in the accelerated math track in RBUSD are slightly above the statewide average for kids of the same age, and they are twice as likely to be on an accelerated math track. But they are far behind the kids in MBUSD, which takes a much more selective approach. Are higher test scores (which I hope means mastery of the subject) a consequence of the extra year of arithmetic and pre-pre-Algebra? That's a subject worth further study.

The district and the HS want to try a more selective approach. Giving only kids that we are confident will succeed a chance to succeed will certainly result in higher test scores. But is that a laudable goal? As I explained in Math class is tough!, my bias is to put kids in the most challenging math class they can handle. That might mean putting some kids on the edge to see if they surprise us--and themselves. I've triumphed over difficult math classes and it has changed me profoundly for the better.

RBUSD MS math teachers suggest that they intervene with struggling students earlier, before students start to fail. They already use benchmark tests several times a year to discover what the students do and do not fully comprehend. (The STAR test results come out during the summer so they are useless for individualized course correction.) Why don't the schools use the data they already have to call in students who have lapses in their math skills?

The teacher said that my conjecture in What the STAR tests really tell us was wrong. They do cover the entire Algebra 1 book before STAR testing in early May. The problem is that the pace may be slightly too fast for some of the kids. Math is a cumulative subject. Because Geometry is so orthogonal to Algebra, those gaps do not necessarily impair performance in Geometry. But missing some key concepts in Algebra 1 could really impair understanding in Algebra 2.

Teachers can use quiz and benchmark tests to identify struggling students and set up appointments at lunch and after school to help them. The teachers are already available at lunch and for one hour after school each day for tutoring. (The teachers cooperatively work out a schedule so that one teacher for each subject will be available each day.) But the students in most need of help may not be the ones savvy or confident enough to ask for help.

So they only thing they want to do differently is to reach out to those students rather than wait for them to come in on their own. It can be done at no extra cost--the teachers at this school work off the clock so much anyway. Only after exhausting all these options, would they recommend that the kids take a step back in math.

I think that's worth a try.

UPDATE:  STAR test scores and external influences

3 comments:

  1. Marie-Christine06:14

    Every good mathematician I've ever met who had to go through US schools had a major struggle... with boredom. It's perfectly possible to flunk out of a subject you'd be really good at because you're not being challenged.

    I like your proposal very much. Few kids (or parents) go in and ask for help in time. Few parents even realize the cumulative effect, and how you're just not going to catch up later if you don't get it now.

    I'd like to point out that I was part of a French generation that got the very accelerated guinea-pig version of math. They stopped that pretty quickly, but it was really good for me to do set theory and advanced algebra and such in 6th grade. I really believe that math, like languages, is easier when your brain is more malleable, and that we do kids a disservice to keep the interesting stuff hidden away till well into grad school.

    ReplyDelete
  2. THIS! Testing essentially no ability to describe a students' long-term outcomes, and thus little utility in really measuring school performance. However, that doesn't mean that testing can't teach us something about what we are doing in our schools- a very subtle dual point that few are willing to acknowledge on either side of the debate.

    This blogger had a great guest post on the subject:

    http://www.anurbanteacherseducation.com/2011/05/teach-for-america-corps-member.html

    ReplyDelete
  3. Thanks so much for your email! I really enjoyed your post. I agree that it's good for students to struggle (or at least, it's better than being bored). My own students really struggle with critical thinking, and with generalizing problems to different contexts.

    Coddling students doesn't do anybody any good. The quicker ones end up bored, and the slower ones end up with some content knowledge, but no ability to think.

    ReplyDelete