The Seabaugh Solution Apology and Explanation

VAM (Valued Added Modeling) is garbage.  It does not work. Louisiana’s system is especially flawed. The underlying premise behind VAM is also flawed, and no VAM assessment (good or bad) should be trusted.  Before I wrote about the Seabaugh Solution I wrote numerous articles about this.

I probably have have dozens of articles where I discuss the fallacies of VAM.  Others around our state have written dozens more.  As a data analyst by trade myself, this misuse and misapplication of data is especially infuriating.

Recently, a series of articles I wrote about an Louisiana Department of Education conspiracy to adjust the entire VAM system to benefit 3 teachers in Caddo was rediscovered, and made popular, but without all the backstory and context.  (I hope those that are promoting that story will also promote this one.)

This conspiracy was actually named the Seabaugh Solution by John White’s staff.  John White, and several of his executive TFA staffers recruited from out of state, carried out this deception after discovering it was flagging our best teachers as our worst teachers.  Please let this sink in.  They understood that VAM was identifying our best teachers as our worst teachers, and they have continued to promote this charade to this day.  Their behavior is well into the loathsome territory here, folks.

A number of native Louisiana citizens working at the department at the time were outraged by this perversion of the VAM system. They could not have disclosed it without jeopardizing their jobs and careers.  Nevertheless, at great personal risk to themselves, they notified me and fellow blogger Tom Aswell, at Louisiana Voice, so we could alert the public to this travesty being perpetrated against our teachers.  Please read Tom’s story for more specific details and background.

My intent was never to involve the specific teachers.  (VAM only has a 25% accuracy rate at best.)  Internally these teachers were sometimes referred to as ineffective by VAM calculations, and by other less flattering terms.  However the truth is in fact the exact opposite.  I was trying to make an ironic point by referring to them as “crappy” when both John White and Alan Seabaugh knew, or claimed to know, the exact opposite was true. The students of these teachers scored consistently at the top – for the entire state.  These teachers initially labels as “ineffective” were in actuality some our best teachers. VAM had classified them as our worst.

Legislators need to understand this and ban VAM from being used in the future for any punitive purpose.  They are knowingly persecuting and hanging innocent teachers in a politically motivated witch hunt that is none of their business in the first place.  LDOE is a state agency and not them employer of these teachers. LDOE should not be making judgments about them from afar; especially based solely on a  few pieces of data that were never meant for the purpose they are being used.

Sadly, these truly outstanding teachers were not alone.  Many teachers across the state are classified as ineffective because their students scored so well it was impossible for them to improve.  Others were classified as horrible because they were teaching some of our most disabled, neglected, homeless, limited English, and poverty stricken kids and learning and improvement is not always linear. VAM assumes all kids will improve at a completely linear rate regardless of teacher or circumstances and teachers are responsible for any and all deviation from that rate.  That’s just ridiculous assumption.  Furthermore, for a child to contribute the maximum points to each teacher ever your would require exponential improvement; which is impossible.  These tests have a finite range.  You can’t improve beyond 100%.  As illustrated by the need for the Seabaugh Solution, for VAM to work for teachers with students performing in the upper ranges, the tests would need to have no upper boundary, they would have to be worth an infinite amount of points.

Teachers have resigned in shame and have even committed suicide  across the nation after being unjustly defined by VAM systems as inferior.

LOS ANGELES — Colleagues of Rigoberto Ruelas were alarmed when he failed to show up for work one day in September. They described him as a devoted teacher who tutored students before school, stayed with them after and, on weekends, took students from his South Los Angeles elementary school to the beach.

When his body was found in a ravine in the Angeles National Forest, and the coroner ruled it a suicide, Mr. Ruelas’s death became a flash point, drawing the city’s largest newspaper into the middle of the debate over reforming the nation’s second-largest school district.

When The Los Angeles Times released a database of “value-added analysis” of every teacher in the Los Angeles Unified School District in August, Mr. Ruelas was rated “less effective than average.” Colleagues said he became noticeably depressed, and family members have guessed that the rating contributed to his death.

I was actually trying to highlight a real problem for out teachers across the state.  I am trying to prevent more, Rigoberto Ruelas tragedies, in our state.  Great teachers labeled have been and continue to be labeled as terrible by a terribly inaccurate and unjust data system.

Part of the problem is how the state has expanded its reach into places it has no business being in.  The state needs to get out of our local classrooms.  It is not helping.  John White’s department of Education is tearing our teachers and students down while claiming it is building them up.  As your teachers what they think about the LDOE’s involvement in their classrooms.  As the Seabaugh Solution shows, even LDOE can’t trust LDOE’s own data.  John White even said this in his conversation with Alan Seabaugh.

I beg you, please do not refer to these teachers as crappy, inferior, ineffective, or anything other than mistreated by a system I was trying to expose as outrageously unfair, dehumanizing and debasing.  The legislature may not have understood this at the time, but they should understand this now.  Anyone who supports VAM is attacking our teachers and children for political points.  It’s not a coincidence that a psychologist, and not a professor of mathematics, designed and endorses this system.  It’s not just that VAM is “a little off”, it’s actually completely backwards and entirely unreliable.

The outrage here is that John White and his executive staff fully understood the implications of what they were told.  Top teachers were being lambasted and shamed by an unjust data system.   John White knew his staff tried dozens of ways to calculate the VAM numbers and could not find a credible way to prevent some of our greatest teachers from being classified as “crappy” so White gave these teachers “bonus points” as he often does for charter schools he’s trying to save from his accountability system, which is also flawed.

Initial supporters of VAM may have had the best of intentions, but VAM is not the answer and never will be because the underlying premise is flawed.  VAM is victimizing our teachers.

Instead of conceding this, LDOE and John White simply added bonus points to certain teachers and shifted the curve downward to classify a new set of teachers as ineffective.  This new set of teachers might have been outstanding too, but they did not have an Alan Seabaugh willing or able to speak for them.

The State needs to discard VAM once and for all.  Not only is evaluating teachers none of their business, LDOE and John White are knowingly doing a horrible job, and playing favorites in the process.

My deepest apologies to the teachers involved.  My intention was to highlight the flaws in VAM to prevent what happened to you from happening to others.

Advertisements

My release of the John White, Alan Seabaugh taped conversation

My release of the John White, Alan Seabaugh taped conversation

I have been getting requests to release the taped conversation between Alan Seabaugh and John White discussing the manipulation of VAM, or the Value Added Modeling system devised by the state to evaluate teacher performance to serve a few of Seabaugh’s constituents. I did not receive the full 14 minutes that Tom Aswell received, just 3 or so minutes of some of the more interesting points, but without the full context. I think I have mined my piece pretty well for interesting nuggets but feel free to review it yourselves to see if you spot anything else interesting.

Audio removed by request as of 2/12/15  (I will need to create a revised one.)

I think it’s kinda cute how John White is using his big boy voice to talk to one of our legislators. You tell he’s really straining his diaphragm to summon up a deeper voice, and at points I worry he will let out a little squeak or perhaps exhaust himself before the conversation is over. Boy, that would be embarrassing. As it is, he just calmly explains how you can’t always rely on the data, especially when the data doesn’t show what you want it to show. No problem, and no need to bring this point up to BESE. Its much easier to randomly assign “bonus points” to favored teachers or teacher groups to get them to have the scores you want them to have. What’s nice about this approach is you can simultaneously sully the reputations of teachers teaching more challenging, poor, or black students and drive them out of the profession. This makes it easier to bring in charter school operators, who don’t require certified teachers at all! It also makes voucher schools look like better options by comparison. If you drive down the quality of public schools enough, well eventually those voucher schools which score 30 points lower on standardized (LEAP) tests public schools are given, will start to look like a good deal, or at least muddy the narrative a little by making them look not so obviously, dreadfully, worse.

I don’t have the full 14 minutes of the conversation.  For that you will have to contact Tom Aswell at Louisiana Voice, but this might still provide some interesting insight into how things really work between, BESE, Jindal, John White and the legislature.  And for future reference, if you are a constituent of Alan Seabaugh up in Shreveport, you might try contacting him for personal favors since that appears to work.  Perhaps he can get DOTD to repave your driveway, or DEQ to take out your trash?  Hey, it’s worth a shot, right?

thrown me something

Bobby Jindal, the Seabaugh Solution and MFP

Bobby Jindal, the Seabaugh Solution and MFP

When I first heard about the “Seabaugh Solution,” a term John White invented to describe a request by legislator Alan Seabaugh to rig the entire teacher evaluation system to favor three teachers that taught at his daughter’s school in Shreveport, something didn’t quite make sense to me. Why would John White risk everything to do something he knew was wrong and probably illegal and unnecessarily complicated?

During the course of his interview with Sentell, White confided “in an off-the-record remark” that the three teachers were ineffective and that Seabaugh was “pushing hard” to fix it

I think John White went to a lot of trouble to set up Alan Seabaugh for the fall on this one. White took the unusual step of openly referring to this “fix” as the “Seabaugh Solution” and even planted evidence for reporters to find.  When he met with reporters about this situation he queried them about it innocently, and set up Seabaugh again. It could be that John White has is completely clueless, but there may be some political intrigue going on here as well.

Apparently, according to the taped conversation between  John White and Alan Seabaugh below,  it seems likely the directions for this scheme came from Governor Jindal’s office in the form of Policy Director Stafford Palmieri, who was Jindal’s Education and Policy advisor prior to that position and is simply referred to as “Stafford” on the tape between Alan Seabaugh and John White.

I suppose it’s possible this is simply a favor called in by the governor to support a fellow conservative legislator.  However this situation could have been handled seamlessly, as DOE has often handled such things in the past.  I asked one of my sources to comment on this situation. I know how these types of “situations” usually work – through manual “adjustments.” These “situations” are often handled in accountability and SPS scores for politically connected individuals. These situations are usually handled quietly, with a manual adjustment to raw scores, and without much fanfare. When done that way they are almost undetectable to anyone on the outside without access to the true “raw” data which DOE never releases.

I’ve never heard of  something like this being discussed and e-mailed about as a”Seabaugh Solution.” That’s almost begging for someone to inquire if they came across that snippet.  However if things ended there I could chock it up to White’s youth and arrogance, but the planting of evidence that was meant to get handed over to reporters?  John White even pushed this carefully crafted narrative by telling advocate reporter Will Sentell that representative Alan Seabaugh was forcing him to change the VAM system,, for the sake of 3 ineffective teachers. This all makes me wonder. . . .was this John White trying secretly to alert us to the undue influence of his boss, Bobby Jindal, and Alan Seabaugh?

Yes, orders came from the governor making him work with Seabaugh to “fix” the problem. Their process was all screwed up. It started as a policy adjustment but Seabaugh didn’t want his name tied to it so they wanted to hide it in the model. I suggested a sound adjustment but because the teachers were so bad they were not helped by my proposed adjustment. I did suggest exactly that, simply adjusting those teachers scores, but that was not acceptable. Apparently the Supt had voiced that there would be a cap put on student predicted scores and we had to do it that way even when we pointed out this was wrong and would negatively impact effective teachers. {emphasis mine}

On May 14, 2013, at 4:31 AM, Crazycrawfish <crazycrawfish@yahoo.com> wrote:

Who the hell put White up to doing this crazy scheme? It seems like he risked an awful lot for something that really shouldn’t have been a big deal for him. Did the Jindal make him do it?

Also, why not just simply tweak those teachers scores with an adjustment and be done with it?

Consider this:

  • If this tape had not come forward, and others had not spoken out, Alan Seabaugh would have been left holding the bag for this.
  • It’s quite obvious John White could have handled this quietly, and this suggestion was even made to him by his own staff and rejected by him.
  • Instead, what happened is John White destroyed the entire VAM system, screwing up the scores for thousands of teachers, just to fix a situation for three.
  • White then went to great pains to refer to this as the “Seabaugh Solution” ensuring Alan Seabaugh and this rigging would be inextricably linked.
  • White then planted evidence with a former DOE staffer named Rayne Martin, to be “found” by Will Sentell with The Advocate.
  • White then feigned ignorance about how this situation transpired, but disclosed more evidence that makes Seabaugh look corrupt to a reporter interviewing him about this exact situation.

Is John White really that inept, was he trying to do the right thing and expose Alan Seabaugh and Bobby Jindal’s meddling, or was this part of a larger plot of the Governor’s to discredit Alan Seabaugh for some reason?

Your guess is as good as mine.

Now showing!  Courtesy of John White's Department of Education
Now showing! Courtesy of John White’s Department of Education

The only thing that is clear is that MFP should be rejected based on this continuing evidence that the formula is based on manipulation and fraud. Did you know that VAM (Value Added) scores are used throughout the MFP formula to distribute funding? This is the same VAM formula that was manipulated for personal and professional gain and has lost all credibility. It won’t be used for teacher evaluations this year, but if the MFP formula is approved as resubmitted by Chas Roemer it will lead to widespread funding inequities and open the state up to countless lawsuits.

I ask the legislature to once again reject the MFP formula resubmitted by Chas Romer and to tell him to restore the previous MFP formula . . . you know, the one without all the controversy, unconstitutional funding mechanisms, and potential fraud.

A. Student Performance

A weight is provided for student performance using the Value Added Model (VAM) and the LAA1 and LAA2 accountability data from the latest available data. The weight is provided under the following circumstances:

i. Category 1 provides a 135% weight times the number of students whose score in English Language Arts (ELA) or Math “exceeded expected achievement” in the Value Added Model or whose LAA1 or LAA2 test results improved one achievement level or more, or

ii. Category 2 provides a 175% weight times the number of students whose performance in English Language Arts or Math “significantly exceeded expected achievement” in the Value Added Model.