About this blog Subscribe to this blog

Thompson: The CEP CAAT Calculates Your Students' Pain

Empathbear The forty fire alarms that have disrupted instruction this year at the District of Columbia’s Ballou High School (where students have a Reading proficiency rate of 24%) illustrates the problem with Value Added Models for evaluation purposes. While not prejudging that principal’s policies, the Ballou teachers who find themselves, once a week, twiddling their thumbs outside of an evacuated building rather than teaching progressing towards their growth targets are a reminder that classroom instruction is only one determinant of students’ performance.

As the Center for American Progress acknowledges, it would be tasteless to calculate the value-added of teachers comforting their students in hospitals and funerals; assisting in suicide interventions and with psychotic episodes; and in grief and drug abuse counseling as well as guiding students through the legal system. Were we to calculate the pain endured by the students of inner city teachers, the toll would be so horrific that some would be tempted to abandon the blame game and invest directly in kids. That might sound too liberal. 

Besides, the CEP implies that teaching during a gang war and a recession, which produces an avalanche of homelessness and domestic crises, could make it easier to meet growth targets, even though urban teachers are too "risk adverse" to recognize that statistical possibility. And since "there is a point at which caution becomes negligence," if VAMs are to be ready for the spring teacher-termination season, they need to be rebranded as "Context-adjusted Test Effects" or CAAT.

In order to maintain the "full speed ahead" urgency desired by the CEP, the most likely scenario will be to incorporate principals’ evaluations into teachers' projected growth targets. If an evaluation rubric determines that a principal’s "spinelessness rating" is below X, then his teachers’ performance goals could be adjusted by the algorhithym of Projected Score = MY + b1(X1 – M1) + b2(X2 – M2) + … = MY + xiT b divided by MayorFentyFavorabilityPolling equation. If the administrator’s "Magpie Quotient" or the infatuation with the flashiest new "quick fix" reaches Y, then teachers' targets would be adjusted by the RheeCronyism Index where MY, M1, etc. are estimated mean scores for the response variable (Y) and the predictor variables (Xs).  When the central office’s AvoidingTellingTruthtoPower rating exceeds the highest sycophancy levels recorded by an "Organization Man" of the 1950s, however, the test-driven evaluations of principals should be adjusted by .... whatever is necessary to avoid a complete exodus of administrators above the age of 21 from urban schools.

The New Teacher Project would question the need for such adjustments because their "research" indicates that 98% of principals respect their teachers and 87% are trusted by their teachers. If that self-reported figure is not fully accurate, (only 68% of teachers said they trusted their principal) VAMs CAATs must be adjusted for school leaders who can’t or won’t enforce attendance or discipline policies, who go along with nonstop test prep, and narrow the curriculum in a way that drives down student performance, but who might not want to accept responsibility for those policies when interpreting the test scores that could end teachers’ careers.

We could borrow a technique from Jesse Rothstein who determined the correlation between students' test scores and their subsequent teacher. When determining teacher tenure, three years’ of test score growth could be calculated as if that teacher was serving under the subsequent principal, destined to replace an ineffective school leader. That would incentivize "high expectations" for struggling central office administrators so they would make better choices in the future.

We should also draw upon recent research demonstrating the loss of a job by a student’s parent can increase his chances of being retained by 15% due to "deprived self-actualization." So, we need new metric adjusting growth expectations in relation to out-of-school factors deprived self-actualization. The loss of the grandparent, who was the only father a student knew, should reduce his anticipated test score growth trajectory by X% during that school year. The recurrence of cancer in the single mother of another student would lead to an adjustment of Y. Nonfatal shootings or the death of a cousin would bring an adjustment of Z. When the full range of those traumas are inflicted on a class during a year, in a school where those tragedies are annual reoccurrences, the formula used for another class where such tragedies are rare could then be adjusted - a few years later after the statisticians work the kinks out of the models.

So maybe we should follow the logic of the CEP, and think of "a cheap tire-pressure gauge, one can sort inflated tires into bins labeled 28 psi, 29 psi, 30 psi, and so forth." Test scores of teachers in Bin #1 would be compared with other neighborhood school results where students have routinely attended more funerals than birthday parties. Bin #2 teachers, while facing socio-economic challenges that are nearly as sad, would do so in selective schools where principals are empowered to enforce attendance, behavioral and academic policies. Teachers in Bin #3 would be those who "prefer to work in schools with better working conditions (where) principal reasoning and parental jockeying usually informs the development of class rosters."  In return for exceeding growth targets, the top Bin #3 teachers could be forced to transfer to Bin #1 schools, replacing equally effective teachers driven out of the profession by VAMs CAATs. 


TrackBack URL for this entry:

Listed below are links to weblogs that reference Thompson: The CEP CAAT Calculates Your Students' Pain:


Permalink URL for this entry:


Feed You can follow this conversation by subscribing to the comment feed for this post.

The comments to this entry are closed.

Disclaimer: The opinions expressed in This Week In Education are strictly those of the author and do not reflect the opinions or endorsement of Scholastic, Inc.