Friday, February 06, 2009

The Learning Continues

It's a Sisyphean task, trying to find and read SPS documents that help expand understanding and explaining this district.

Yesterday, I spend the day as part of the accreditation team for Roosevelt (every high school has to do this every 6 years to be listed on the NW Association of Accredited Schools. (I learned that the Association started in 1917 and Roosevelt, started in 1922, has been on the list ever since.) The school had to put together an accreditation report as well as gather community members to conduct walk-thrus of the building. I did learn a lot because the report gathered some stats about SPS high schools that I hadn't seen gathered in one place (I'm sure Charlie has all this at his fingertips).

There were charts with the following:
  • enrollment
  • ethnic background of students
  • Free/reduced lunch
  • Not living with both parents
  • limited English
  • Special Ed
  • percentage of returning students
  • assignment information
  • certificated versus classified staff
  • WASL breakdowns
  • SAT breakdowns
  • transfers in and out
  • suspensions
  • expulsions
  • cumulative GPA
  • graduates/dropouts
You wouldn't have to guess on most of these items for most high schools. But some items are striking. (Figures from 2007-2008)

  • Sealth has the most students with limited English proficiency at 19%. Sealth also has the most even distribution of ethnic backgrounds with each (Native American, African-American, Latino, Asian and White, all between 22-29% except Native American). The whitest comprehesive? Ballard at 64.7% (but that fits because the 2000 Census shows Ballard to be the whitest neighborhood in Seattle).
  • The school with the lowest number of students not living with both parents? Roosevelt at an astounding 26.9% (given the divorce rate, that's pretty good). The highest was RBHS at 65%.
  • The highest free/reduced lunch? Cleveland at 62.9%. Nova and Center were at a virtual tie at about 11%. Among comprehensives, Hale is the lowest at 14.9%.
  • Special Ed? Cleveland is the highest at 16.6% with the lowest being Garfield at 5.8%.
  • SAT scores: Nova was the highest on the Verbal SAT at 611 (with Center one point behind at 610), Roosevelt was highest on the Math SAT at 572 and Garfield was highest on the Writing SAT at 567 (with Center right behind at 565 and Nova at 564).
  • WASL scores: Center School had the highest math pass rate (76.6% with Roosevelt a shade behind at 76.3%), Roosevelt had the highest reading pass rate at 92.7% and the highest writing pass rate at 94.4%. Center School had an astonishing 73.3% pass rate for science with the closest school to that being Garfield at 50.4%. What are they teaching at Center? (In all fairness, Garfield has a much more diverse population than Roosevelt or Center and their scores are really high.) But math continues to plague all the schools with gaps of between 30-40% points between math and reading/writing.
  • What was interesting to me was the chart with the transfers in/out. At the comprehensives, 9 out of the 10 had 100+ students transfer out (with the average coming in at about 30). That seems like a lot of movement to me of students either going to another SPS high school or leaving the district. I'd have to ask Tracy Libros if that is about average.
  • The school with the lowest suspension rate? Ingraham (and not that it's a bad school but I wouldn't have guess it) at 2.7%. For this rate the principal may be the difference. The highest was Cleveland at 21%. Nova had zero but I'm not surprised because of the closeness between faculty and students.
  • Lowest dropout rate? A tie between Hale and Garfield at about 2.2%. Highest? RBHS and Cleveland at about 10%. Interestingly, this rate goes up for the Class of 2007 with Hale at 15.4% and Garfield at 10%. Ballard had the lowest Class of 2007 dropout rate at 9.6%. The highest dropout rate for the Class of 2007 for a comprehensive high school was Sealth at 30.9% (but Center was the highest overall with 33.8%). How confusing is that? Center has the best WASL science pass rate by far and yet they lose a lot of seniors who should graduate?
I wonder if someone at the district sits down with all this information and creates a picture of what is going on at each high school and for all the high schools as a whole. Roosevelt, Garfield, Ballard and Hale seem to have the most even stats with Nova and Center performing very well, too. It would seem that there is a lot to be learned from these stats like what Nova and Center seem to be doing that their students do so well on the SAT. For smaller schools, these are great scores but for larger schools with more challenges, Garfield and Roosevelt really do well.


Dorothy Neville said...

Gotta take all those stats with a grain of salt or something.

SAT. Not everyone takes SAT. Nova and Center, small alternative schools, kids choose to attend and have a high percent of college bound students. Therefore their SAT scores probably are indicative of the school as a whole, but they would not be comparable to the SAT scores at a comprehensive high school. Not one bit.

Comparing GHS, RHS and Ballard on SAT and WASL is more reasonable, but still imperfect.

I am so far not impressed with RHS's writing instruction in 9th and 10th grade. I have heard that GHS has an emphasis on writing, and there are some good writing teachers at Washington. That may explain the higher Writing scores from Garfield.

I am not surprised that RHS has highest math scores as they have so far thwarted attempts to reform their math curriculum (and kids rising from Eckstein would have also had mimimal exposure to math reforms).

If they ever do any real analysis on the PSAT scores, that will be a better indicator. Especially if the grant from Boeing lasts another couple years, we could truly have value added data. Garfield would have that now for its sophomores and juniors.

Any urban comprehensive high school anywhere that claims a 2% dropout rate is suspicious. As we know from Texas and other scandals, it is all too easy to hide drop-outs. What about the 100+ kids "transferring out?"

FRL: Elementary APP has had extraordinarily low FRL scores and high scores for living with both parents. This will affect the high school APP cohort as well. It would be very useful if one could separate out FRL by program (for those with enough students that privacy is maintained). Since at high school they don't specify former APP or Spectrum kids, you lose that piece of information on FRL and living with both parents. It would be useful to have to compare schools.

MathTeacher42 said...

Is there a URL to this data?


Melissa Westbrook said...

As I said it is compiled in RHS'accreditation report. I can ask if it will be on-line but I doubt it. I'll try to see if the high school charts are at the website somewhere.

hschinske said...

As I've said before, I don't think *average* SAT scores (especially on self-selected populations such as those at Nova or Center) tell you a lot. It would be a lot more useful to get a sense of the distribution. For instance, if you're wondering where a student who's likely to score extremely high on the SAT is most likely to find intellectual peers (insofar as you *can* get data like that out of the numbers), you won't find that out from the averages, which are all far too low from that point of view. National Merit data, on the other hand, is too narrow, except for schools like Garfield. If you had some sense of the average for the top ten or twenty percent at each school, that would be more useful.

Helen Schinske

Helen Schinske

CJF said...

These stats all come from the "Individual School Summaries" section of the 2007 Data Profile, which is compiled annually by the Research, Evaluation, and Assessment department at SPS. You can find links to the last 16 years of profiles here, including the 2008 profile. You can get a bound hardcopy by contacting REA (252-0140).

SP said...

Yes, this huge data report comes out each year aound the end of December. It certainly is not easy to locate in the district's website.

Does any one know of any charts showing HS academic grades (ie % of A's, B's, etc) for each of the HS's? In the District Summary report there is a page "Academic Grades Awarded by Ethnic Group", but not broken down by each HS. It lists the total for the district in '07-'08 as:
A's 43.9%
B's 26.5%
C's 18.9%
D's 10.7%

It seems very strange that almost 44% of the grades given were A's. I wonder how that compares at each high school? How does that compare to other school districts?

Note that the % of "N" grades are not listed (at our HS it has been similar % to D's) which changes the overall picture, more than doubling the actual "no pass" grades.

It will be very interesting when '08-09's report comes out, the first to include N grades in many years. It will greatly also impact the GPA's.

More alarming, is to be able to look at the % of students earning one or more "N" and/or "D" grades. The REA releases these figures to all the schools. It is serious when approx. 40% of all students earn one or more N's (and/or D's). How do our schools compare?

Unfortunately, "grade inflation" issues and concerns make it difficult to track & ask for accountability.

SP said...

There is also another report I'd like to find an update for:
Seattle students attending 4-year WA. state colleges, showing the % of students who need to take remedial classes in their first 3 years of college. This report is broken down by each HS.

The most recent report I've seen is from 2005, when 6.1% of all Seattle HS students needed 1 or more remedial classes.
At that report, Sealth came up highest (26.5%),Cleveland (13.3%), Nova, RBHS & WSHS (all at approx. 10%). NOVA surprised me, as their SAT's have always been very strong. Surprisingly, Franklin came in needing less remedial classes, at 8.2%.

No surprises with the HS's with lowest % needing remedial classes.

Charlie Mas said...

First, I never really feel complimented by the presumption that "I'm sure Charlie has all this at his fingertips". On one hand it makes me feel like a bit of a freak - which I may well be, but that doesn't mean I'm open to discussing it - and on the other hand it makes me feel like I would be disappointing people if I didn't know these numbers. I guess I'm a freak because I did know a lot of them.

I get them mostly from the annual Data Profile report. It is a GOLDMINE of data.

As for the high rate at which NOVA students need remedial college classes, I wonder if it isn't partly a reflection of the absence of grades at the school. Some colleges and universities are comfortable with it, some are not.

My number sense was alerted by the grade breakdown provided by seattleparent:

A's 43.9%
B's 26.5%
C's 18.9%
D's 10.7%

Right away I saw that these numbers would add up to too much - 100%. What? Nobody failed any classes? That's not very credible.

In a standards-based learning system, it would be no problem for 44% of students to achieve an A in any or all classes. It would NOT, however, be possible for all students to pass all classes.

Josh Hayes said...

I agree with Charlie's WTF on those grade distributions. I don't necessarily think letter-grading is a great idea, but I'm struck by a couple of things.

One is, as Charlie points out, nobody fails, unless the reported numbers are simply for those students who don't fail. This strikes me as a pretty useless set of numbers if so: how many kids DID get failing grades? Who knows?

Second, look at the distribution of grades. I can't stick in graphics here, but just look at the numbers: nearly HALF of students got A's. Does this mean that "grading on the curve" is out the window? Not that I'm unhappy with that, I'm just concerned that it's more likely to represent grade inflation than anything else, especially since the assessment tools (WASL, SAT) should be geared to the same standards on which the curriculum is supposedly based. If they're doing that well, they should be kicking ass on the WASL. And on the SAT (and yes, I know that's a sentence fragment). But I don't think they are.

Just seems like there's a disconnect here.

SP said...

You are right, Josh & Charlie. The reports are not showing what really counts, the number of failing "N" grades from last year. In Seattle, you can still pass with a "D".

I also was astonded by the 44% A's, knowing that the WASL scores don't reflect that, and concerned about grade inflation. According to the old Seattle Times School Guides, the UW compensates for Seattle's high grade inflation (unfortunately the UW no longer gives the district ratios out to be published in the School Guide).

The REA Department tracks the # of N grades, and it should be available if requested. At the individual school level, they break down the % of D's & N's by demographics, % of D's & N's to all marks, but also % of students with 1 or more D's and N's (which is very disturbing). Finally, the REA also can give reports on grades by Department, which is another very interesting breakdown.

To me, the question is why these 2007-2008 annual Data Profiles do not include the failing grades in their totals? This current year, when "N" grades now count in the GPA, there will be a huge difference.

hschinske said...

I don't know which courses students most often need remediation in, but I would guess math and English. I think a number of Nova students may avoid math to some extent. An intelligent, test-savvy student can easily get as far as a 500-550 score on SAT math without having done any trig or precalculus, and if they got to college without having covered that material, they would need a remedial course or two before getting to calculus.

Helen Schinske