Earlier today a Facebook friend and teacher posted the following note on my page:

"The element of gifted is listed as a component in the development of my score as a teacher. I emailed the state department to find out how it is used. The person answering my email sent me links to material that mentioned gifted. None of it answered my question as to HOW gifted was factored in. Therefore, I emailed AGAIN with the information that the material they sent had not answered my question. I never did get an answer. Yesterday at a meeting, I was told that gifted would have to make more progress on their scores than other students. . . which makes no sense to me since many are already at the top of the range. . where are they supposed to go? It's very worrisome for me."

For those of you (probably even most teachers) with little knowledge of the provisions of Act 54
which legislates the use of the new teacher evaluation system called COMPASS, I have included some information below that "explains" how student test scores are used to PREDICT the progress they should make during a school year and upon which PREDICTED progress the teacher is evaluated.

I asked for advice and received this from Tom Aswell author of the blog

Send this to Supt. White (Rene Greer, communications director, resigned last Friday):

Pursuant to the Public Records Act of Louisiana, R.S. 44:1 et seq., I respectfully request the following information:

Then simply type in what record(s) you are requesting.

Below is the public records law:





§1. General definitions

A.(1) As used in this Chapter, the phrase "public body" means any branch, department, office, agency, board, commission, district, governing authority, political subdivision, or any committee, subcommittee, advisory board, or task force thereof, any other instrumentality of state, parish, or municipal government, including a public or quasi-public nonprofit corporation designated as an entity to perform a governmental or proprietary function, or an affiliate of a housing authority.

(2)(a) All books, records, writings, accounts, letters and letter books, maps, drawings, photographs, cards, tapes, recordings, memoranda, and papers, and all copies, duplicates, photographs, including microfilm, or other reproductions thereof, or any other documentary materials, regardless of physical form or characteristics, including information contained in electronic data processing equipment, having been used, being in use, or prepared, possessed, or retained for use in the conduct, transaction, or performance of any business, transaction, work, duty, or function which was conducted, transacted, or performed by or under the authority of the constitution or laws of this state, or by or under the authority of any ordinance, regulation, mandate, or order of any public body or concerning the receipt or payment of any money received or paid by or under the authority of the constitution or the laws of this state, are "public records", except as otherwise provided in this Chapter or the Constitution of Louisiana.

(b) Notwithstanding Subparagraph (a), any documentary material of a security feature of a public body's electronic data processing system, information technology system, telecommunications network, or electronic security system, including hardware or software security, password, or security procedure, process, configuration, software, and code is not a "public record".

(3) As used in this Chapter, the word "custodian" means the public official or head of any public body having custody or control of a public record, or a representative specifically authorized by him to respond to requests to inspect any such public records.

The custodian of the records (in this case, White) must produce the records immediately, if available, subject to fines, attorney fees and possible imprisonment. If not available, he must let you know within 72 hours when they will be available.

ACT 54 Teacher Evaluation - an excerpt from the LDOE website.

How Are Value-Added Scores Calculated?

Louisiana’s Value-Added model incorporates students’ information with their test scores to determine the average student growth for children with similar histories. The actual achievement for each student is compared to the average achievement of similar students to determine if a student has made more, less, or a ‘typical’ amount of progress, and the results for all students in a teacher’s assignment are then combined for that teacher. If a teacher’s students make more progress than would be expected, the teacher would receive a positive value-added score; if the collective progress of a teacher’s students is less than expected, he or she would earn a negative value-added score; and if student results are consistent with their prior performance and other similar students, the value-added score would be a zero.

Louisiana’s Value-Added Model uses each student’s test scores in core subjects and compares actual performance to the expected achievement of each student, based on his or her prior achievement history, as well as other individual factors. This means that unlike other methods, the state’s value-added method takes into account whether a teacher’s students have specific challenges outside the teacher’s control when figuring how much the teacher accomplished with the student.

The Value-Added Model considers the following factors when identifying expectations for student growth:
•Available prior achievement data (up to three years);
•Gifted status;
•504 status;
•Disability status;
•Free and reduced meal eligibility;
•Limited English proficiency;
•Prior discipline history;
•Classroom composition variables;

In order for a student's assessment results to contribute to the value-added assessment of a particular teacher, the student must: have been enrolled in that school from early fall until testing time; have a prior year of standardized test data; take the regular state assessments (LEAP and iLEAP). In addition, the teacher must agree that s/he taught the students, through the roster verification process.

This is an excerpt from a report to be found on the Louisiana Board of Regents website. You can find other reports there by searching for "value added." It describes the complex metric used to determine value-added and the PREDICTED PROGRESS that each student should make. I would like to have Supt. White explain this.

Level 1: Students

Yijk = π0jk + Σ(πpjk)apijk + eijk where Yijk is the achievement of student i in class j at school k in the target subject

π0jk is the mean achievement for classroom j at school k πpjk are the p coefficients that weight the contribution of the student level data in the prediction of Y for p = 1 to the total number of coefficients

apijk are the student level data (prior achievement, demographic variables, and attendance) that predict achievement for p = 1 to the total number of data points
eijk the student level random effect, the deviation of the predicted score of student i in classroom j in school k from the obtained score

Level 2: Classrooms

π0jk = β00k + Σ( βq0k)Xq0jk + r0jk where π0jk is the mean achievement for classroom j at school k

β00k is the mean achievement for school k

βq0k are the q coefficients that weight the weight the relationship between the classroom characteristics and π0jk, q = 1 to the total number of coefficients

Xq0jk are the classroom level data that are used to predict achievement; this is also the location in the model at which codes for recent TPP completers are entered (described below)

r0jk the classroom level random effect, the deviation of classroom jk’s measured classroom mean from its predicted mean

Level 3: Schools

β00k = γ000 + Σ( γs00)Ws00k + u00k where β00k is the mean achievement for school k

γ000 is the grand mean achievement in the target subject

γs00 are the s coefficients that weight the weight the relationship between the school characteristics and β00k for s = 1 to the total number of coefficients

Ws00k are the school level data that are used to predict achievement

u00k the school level random effect, the deviation of school k’s measured classroom mean from its predicted mean

Coefficients for variables retained in the model for each year are scaled to the approximate standard deviation of the educational assessments (iLEAP and LEAP) used in Louisiana: 50. There is general consistency in which variables have been retained in each content area across testing years. The previous year’s achievement for a student in a given content has been the strongest predictor of the current year’s achievement among prior achievement scores. Having a special education diagnosis is a consistent negative predictor of achievement and in some cases (e.g., Mild Mental Retardation), the effect is large. Interested readers are referred to previous reports for descriptions of the base models from 2005-2006 to 2007-2008

The perpetrator of this metric who no longer works for LDOE and therefore is no longer required to explain it is Dr. George Noell (Noell et al., 2007; Noell et al., 2008; Noell et al., 2009) I have not been able to find any of Dr. Noell's research that is associated with this model that has been peer reviewed. Peer review is the gold standard for research. If I am incorrect in this finding, I would invite Dr. Noell to make that information public.

1 comment:

  1. Thanks, Lee. When the mechanism for judging someone is this complicated, it's no wonder teachers are upset. This whole charade by Jindal is a smoke screen. Even George Noell resigned before it went into force knowing that his work would be used under inappropriate circumstances and did not want to suffer the fallout amongst his peers. His reputation would have been under scrutiny and he could not risk losing his standing in the academic world. This way he can claim that he only provided the research but did not implement it or even supervise its implementation. Debbie Meaux


Your Comments Welcome