Thursday, October 25, 2012

VAM Fix Must Comply with Law

Note: Please read this post as a supplement to the article below it.
A teacher from Caddo sent me an email in response to my post this week on fixing the VAM system for Louisiana. She said the attorney for the Caddo School Board gave his opinion that John White's proposed fix for South Highlands Elementary and similar schools did not comply with Act 54. (The 2010 law setting up the new evaluation system which includes the 50% value added factor). The attorney believes that the law does not allow for such an exemption that removes the value added factor from the evaluation of some teachers. Unfortunately for this situation, I agree with the attorney. I don't think the State Superintendent can go around picking and choosing which teachers can be exempted from VAM, especially not after the fact of getting a bad evaluation. There is a provision for teachers of non-tested subjects and grades to be rated using student learning targets (SLTs) jointly developed by the teacher and his/her principal, but the teachers involved did not fall into this category. This just one more example of what you can expect when you put the careers of professional educators in the hands of an inexperienced unqualified individual. The email I got also pointed out other invalid results that can occur in the evaluation of a teacher using VAM, but there is no legitimate way of tinkering with the VAM formulas to correct these problems.

The same problems are being observed all over the country where a similar value added evaluation is being tried. But the most disgusting example I can think of is the following: A teacher who had previously been recognized as one of the most highly effective in her system, got an ineffective rating because of  VAM and was dismissed. Later, strong evidence was found that the teachers the previous year had cheated on the student testing so that their student scores would help them get an effective rating. Those inflated scores caused the teacher the following year to be rated as ineffective because the students showed little or no growth on state tests. So one highly effective teacher was fired while other unethical or fraudulent teachers were rated as effective!

This is what happens to our formally proud profession when we let amateurs sell the public and the legislature on a "miracle" solution that is not based on sound educational principles. Act 54 is a very bad law that has terrible unintended consequences.  If the legislators from Caddo really want to do the right thing, they need to repeal Act 54 during the next legislative session!

Monday, October 22, 2012

Can VAM be Fixed?

Superintendent John White believes he has a solution to the problem of teachers in top rated schools receiving an “ineffective” because of the VAM portion of the new teacher evaluation system. It was discovered recently that some high performing students in exclusive magnet schools sometimes experience a lull or slowdown in their academic gains caused by factors outside the control of the teacher. Even though such students are normally high performers, in some cases a classroom of high performers may perform below its VAM predicted scores on state tests, thereby placing their teacher in the bottom 10% on the VAM ranking statewide. When simulations showed that several teachers at the top rated elementary school in the state may be classified as ineffective by VAM, the whole community and their state representative were up in arms. When this story broke in the newspapers, Superintendent White immediately adjusted his schedule to travel to Shreveport and met with teachers and administrators at South Highlands Elementary Magnet School.

Within just a few days of the discovery of the flaw in VAM for high performing schools, Superintendent White had a solution. He will be proposing to BESE that in cases where some teachers who teach high performing students get an “ineffective” from the VAM, the state will simply wipe out the VAM portion of the teacher's score and use the qualitative portion of Compass as the only factor in determining the rating of the teacher. This exemption to VAM will be applied as long as a teacher's students score at the top two levels on state assessments, even if her/his students do not show the growth the VAM formulas demand. That's a great victory for the teachers at South Highlands Elementary school and the teachers there certainly deserve to be exempted from this flaw in the evaluation system.

I believe White's recommended change for South Highlands and other similar schools will be approved by BESE, not because of the extenuating circumstances that may cause an unfair evaluation of those teachers, but because the legislator representing that district is a solid supporter of Governor Jindal and his “reform” programs. White, who has no training in teacher evaluation, no training in statistical analysis, and who has never evaluated teachers, came up with an instant solution that will fix the problem for that select group of teachers. But White's solution will still allow thousands of other teachers in the state to be vulnerable to ineffective ratings based only on the VAM even if they have extenuating circumstances in their classrooms.

For example let's look at another teaching situation. (The following is a hypothetical example because I don't have access to actual the value added growth that VAM predicts for various socioeconomic groups.) Lets consider a high poverty middle school in the inner city that has had most of its high performing students transfer to magnet schools. A 20 year math teacher (teacher B) is assigned a group of 25 students where the VAM formulas project that those students should show composite growth of .8 years in math for that year. But during that year, two of the girls get pregnant, 3 of the boys are picked up on drug charges and detained in a juvenile institution for three weeks to a month, and three other students' families were evicted from their home because their unemployed parent could not pay the rent. Those students had to live with relatives and friends where they were lucky to find a bed in which to sleep, much less find a quiet place to study. One other student's mother was murdered by her estranged husband causing traumatic shock waves throughout the community. LEAP testing produced a composite growth in math of this class of only .4 years. I guess you could say this teacher's class also experienced a lull or slowdown in academic gains probably caused by factors outside the control of the teacher. Teacher B received a good rating from her principal on the qualitative portion of Compass, but her low rating on VAM placed her in the bottom 10% statewide, so she received an overall ineffective rating. There is no appeal for an ineffective rating on VAM. That teacher immediately loses her tenure, has her salary frozen, and is placed first on the list to be laid off next year, in the event the school system is forced to reduce teaching staff. That teacher is all alone. She did not have a state legislator who was aligned with the Governor to campaign for her and get a special rule change for her extenuating circumstances. There will be no adjustment in her evaluation.

Here's another example deserving consideration. Some teachers have pointed out that there seems to be a difference in average state test scores for students from one grade level to another. That difference could be caused by several factors. It could be that the state tests do not increase smoothly in difficulty from grade to grade. LEAP and ILEAP tests are constructed by contracted testing companies that make a huge profit regardless of variations of their tests. Or it could be that students in one year where state policy requires retention of students who do not make the state cut off score work a little harder to pass the test that year than they do the previous or succeeding years. That can have both adverse and beneficial effects on ratings for teachers in different grades even if teacher performance is generally the same from grade to grade. Would White recommend a change in the teacher evaluation system in such a case? Specifically using real data if we compare the 4th grade LEAP scores of all students statewide in 2011 to the ILEAP 5th grade scores we find that 24% of students scored below basic in ELA in 2011 while 30% of those students scored below basic in 5th grade in 2012. Statistically this means that more 5th grade ELA teachers are expected to fall in the bottom 10% of VAM than 4th grade teachers because their students don't perform as well on state tests. If a disproportionate percentage of 5th grade teachers are rated as ineffective because of the factors I mentioned above, will some legislator go to bat for those teachers and get them an exemption from VAM?

One more point. It turns out that the Governor's supporters in the legislature generally represent the more affluent areas of the state. Opponents of the Governor generally represent impoverished communities. Which teachers do you think have the best chance of getting exemptions from or adjustments to VAM? The ones teaching in high poverty schools or the ones teaching in more affluent schools?

Can VAM be fixed so that it will be fair to all teachers and still produce the mandatory 10% of teachers rated as ineffective? Obviously not, because for every group of teachers granted special exemptions, the system will have to pull more or the rest of the teachers down to the the ineffective level.

I hope all teachers and administrators will stick together in opposing the continued implementation of this fundamentally flawed evaluation system. Right now I believe most teachers are appalled at the lack of accuracy and now also the lack of integrity in the decision making process relative to this evaluation system. It cannot be fixed. It must be junked. Let's not allow some good teachers to be thrown under the bus while others get an exemption.