Tuesday, October 20, 2015

Fantasy Standards

Reminder: The elections for BESE are this Saturday, October 24. 

An enormous amount of out-of-state money (From the Wal Mart heirs and the Broad family) is being spent on electing John White controlled members of BESE! Make no mistake about it. These candidates supported by the LABI/Lane Grigsby PAC intend to increase privatization of our schools and set up a system where the "market" will determine winners and losers in public education. If it continues to work according to the trend we already see in Louisiana, our schools will become more segregated according to wealth, and our tax support for the schools serving the most at-risk students will decline.  During the most recent period of education "reform" in Louisiana under Jindal and Superintendents Pastorek and White, Louisiana's ranking on the National Assessment of Education Progress (NAEP) has declined to just third from the bottom, just above Mississippi and Washington D.C.  And yet the John White controlled candidates are claiming that Louisiana test scores are going up in their campaign literature.

The nation of Chile has already tried this experiment and it was a miserable failure. Now the whole educational system in Chile is in turmoil in an effort to repair the damage.

It is easy to tell the LABI/Grigsby/John White candidates. They are running numerous television ads promising to grade our schools A-F and to reward the good teachers with pay raises. They are also mailing slick oversize cards promoting their controlled candidates and nasty hit pieces on the pro-public education candidates. Voters may be interested to know that BESE cannot raise any new money to give pay raises, that's a function of the legislature, so they would have to do it by robbing Peter to pay Paul.

Remember the last time when Jindal implemented merit pay for teachers without providing a penny of new state funding? The local school systems had to mostly abolish their teacher step increases so that they could put money into the merit pay system. Then it turned out that the teachers teaching the VAM subjects were at a huge disadvantage compared to teachers rated with SLT's.

Please vote for the good pro-public school candidates Saturday Oct. 24, that are listed on the Flip BESE Facebook page. Click here.

John White's Fantasy Standards

White sent an email to local superintendents addressed to "Dear Colleagues": In this latest propaganda piece he claims that with the adoption of the PARCC-like cut scores, Louisiana is on the path to higher standards. He has set a policy that Louisiana will gradually move from level 3 performance to a level that schools must average level 4 performance on the new standardized tests in order to get rated as A schools by the year 2020. He directs his readers to an article by Mike Petrilli of The Fordham Institute, that supposedly demonstrates that PARCC states including Louisiana are steadily moving to higher standards. But the article points to Ohio as one of the real laggards in pushing for higher standards. But we are supposed to assume that Louisiana is forging ahead to true higher standards.

The PARCC Consortium is Rotten to the Core. . . .  Pun Intended.

I have pointed out in this blog that apparently when the testing bosses at Pearson (the testing company running the PARCC Consortium) realized what terrible scores were resulting from the PARCC test nationwide, they took immediate steps to soften the blow or to sugar coat the results. Much of this is accomplished by using the phony scale score system of 650 to 850 for all tests in the PARCC group.  So even the students scoring extremely low, or even zero, on any of the tests start off with 650 points right up front. Then the Consortium set extremely low raw cut scores averaging about 29.8% for level 3 and about 45.5% for level 4 performance. (level 4 performance is the goal which the PARCC Consortium would designate as the minimum performance for "proficiency"). The percentage scores listed above are the Louisiana version of the raw scores needed to get the ratings of 3 and 4, but we have been assured by White that the conversions are similar for all PARCC states. We don't really know for sure because the raw scores for most other PARCC states are still secret and the PARCC Consortium would have preferred that all the raw score to scale score conversions had been kept secret. What exposed the ugly underbelly of PARCC was that the local superintendents in Louisiana thought they they needed to know the raw scores of their students and then the citizen's group filing a public records request thought they had a right to see the conversion tables for raw to scale scores and vice-versa. So those are the only reasons we can calculate the old fashioned raw scores equating to the totally artificial scale scores.

So not only can the testing bosses change apparent student performance from year to year by lowering the underlying raw scores that produce the scale scores, but they can also change the ratio of easy questions to difficult questions. Also, they can change the order of questions to place most of the easy questions on the front end of the test before the students get weary and start just skipping over questions. That's why I call these fantasy standards. The testing company and the LDOE can make them up to come out any way they want. Who's checking on them to make sure they are honest in making any changes to the tests? Well just trust them, they are constantly running field tests and high tech statistical analysis much too complicated for us to understand. That's why they wanted to keep the raw scores secret from us. Raw scores are just too difficult for us to really understand.

How does Louisiana Compare to Ohio?

One of the biggest criticisms of certain states by the Petrilli article is that Ohio lowered the high standards of PARCC by choosing level 3 as their definition of proficient instead of the level 4 that the PARCC consortium recommended. Well it just so happens that Louisiana is also starting with level 3 for satisfactory performance this first year, but White has set a goal of level 4 average performance by the year 2020. Well that may look good to our gullible business community who wholeheartedly support anything White comes up with, but I wonder if they realize that Louisiana would be expecting our students to average getting only 45.5% of the questions right on our PARCC-like tests by 2020? That's not all students getting that high a score, that's just the average White wants to reach. How is that going to make our students all college ready? 

So the reason I call these "fantasy standards" is that they are just made up standards to begin with. No one really knows whether or not they will prepare our students for college. But the biggest part of the fantasy is that teachers and principals are supposed to double their efforts at test teaching, when the real underlying scores behind the standards can be changed at will by the LDOE and the testing company. Oh, and did I mention that last week White got BESE to remove all the previous safeguards in BESE Bulletin 118 that may have prevented test tampering and score manipulation?

1 comment:

Alter ego said...

Slides 34 and 36 are revealing in this BESE presentation.
White is selling the idea that level four will be the new requirement. PARCC level four is much wider than the old LEAP level four. So while PARCC level four is more selective than LEAP level three, it is neither as selective as LEAP level four or NAEP level four. At least according to the results from the cut scores and scaling that BESE approved.
To translate this into a White-speak, PARCC level four is a lower bar than either NAEP level four or LEAP level four.
So even here, his "raising the bar" is not to the level he has implied.
And all of this is predicated on the new Louisiana "PARCC" score scaling having any accuracy at all. Compared to Ohio and Illinois, Louisiana's scores are much more unstable grade-level to grade-level. If PARCC accurately assesses students as being on track for success in college, how is it that only 22 percent of our seventh grade students reach level four in math, but 32 percent do so in the eighth grade? Are our eighth grade teachers that amazing? (No offense intended to either seventh or eighth grade teachers here.) Meanwhile in Ohio there is a significant drop-off from seventh to eighth grade Math. Bucking Ohio's trend so sharply does not seem reasonable.
I imagine that the spin on these results, when released in final version, will be dizzying.
Herb Bassett