Sunday, December 30, 2012

The State's Junk Science Exposed!

In the ever continuing teacher evaluation system conflict between the DOE and the UFT.  One aspect that both sides agree is the use of the New York State Value Added Measurement (VAM) algorithm to account for either 20% or 25% of the overall teacher evaluation.  In a previous pilot study using the VAM in New York City's teacher data reports had resulted in errors as large as 87% for English and 75% for Math!  Yet the State still insist to use this "junk science" to evaluate teachers. Unbelievable but true.

While a few people really understand the VAM algorithm, there is a consensus that teachers who teach in high poverty schools with English Language Learners and Special Education students will be at a disadvantage when the VAM is used to evaluate the teacher..

The teachers most negatively affected by the VAM will be those teachers who have "high needs students" which includes the following cohorts:

  • High poverty students.
  • Academically & behaviorally challenged students .
  • Special education & English Language Learners.
  • Dysfunctional families and homelessness.
Another issue is the student attendance problem.  There are rumors that the DOE and the UFT have come to an agreement that only those students with attendance rates of 90% or higher will be used in the teacher evaluation system.  Assuming this is true, that seems to be reasonable. However, what if the DOE reneges on this agreement as they have repeatedly done in the past?  I certainly would not trust the DOE on the attendance issue since the State has refused to support the attendance rate issue in the Buffalo dispute and the DOE's real objective is to terminate as many teachers as possible..

Finally, the VAM will pit teacher against teacher as no teacher will accept "high need students" transferred from another class during the school year without an iron-clad guarantee that the student's academic progress is not included in the teacher's evaluation.

I cannot see how the VAM accounts for real student growth.  Certainly it does not account for real student learning or achievement and does not account for some of the intangibles like this

In summary, the VAM algorithm that will be used for the teacher evaluation is inaccurate, unreliable, and meaningless and with too many fudge factors. For further details on the NYS VAM algorithm please see Gary Rubinstein's blog. Why would any sane person agree to link their job performance to this "junk science"?


I noticed that... said...

Using VAM to evaluate teachers based on students with 90% attendance can be harmful, too.

The attendance rate, in a high school, is based on a student attending his/her official attendance period. I have students who are present for their official attendance period, 100%, and I have yet to see that student in my class or how about students who attend their other classes 30-50% of the time? How would VAM be applied if this issue exists in many high schools?

I strongly believe that no VAM should be used at all.

Anonymous said...

This is very scary. To lose ones job based upon an algorithm that does not reflect reality.

Anonymous said...

Which UFT caucus has been touting VAM-based evaluations as acceptable for teacher evaluations?
Which UFT caucus has spent the last year bringing Rubinstein's research to light in an on-going campaign against this ill-conceived evaluation method?
Join us; vote the MORE slate.