Instability in Value-Added Modeling: The Race to the Top Leads to Anarchy and Chaos.

Three Stanford researchers including Linda Darling-Hammond, plus a UC Berkeley colleague just published:

  Value Added Modeling of Teacher Effectiveness: An Exploration of Stability Across Models and Contexts,” in Education Analysis Archives, September 30, 2010.

  They found that teacher rankings were “significantly and negatively correlated with the proportions of students they had who were English learners, free lunch recipients, or Hispanic, and were positively correlated with the proportions of students they had who were Asian or whose parents were more highly educated.”   In other words, the more high-needs students a teacher taught, the lower her value added percentile rank was likely to be.

In addition, English teachers were more highly ranked if they had greater proportions of girls in their classes, and math teachers got higher VA scores if they taught more “fast track” math students.

 One mid-career English language arts teacher’s ranking changed from the bottom decile (1) to the top decile (10) from the first to second year. In the first year, her students were 58% ELL, 75% Latino, and 42% free and reduced price lunch. The second year she had 4% ELLs, and half the percentage of Latinos and low-incomes than her previous year. Parent education levels of her students were also significantly higher the second year.

 Teacher rankings also varied across courses.  A teacher could teach algebra better than geometry or literature better than composition. And teacher rankings varied greatly from one year to the next. Depending on which value-added the model used, 39-54% of teachers varied by three or more deciles across courses; and 45-63% of teachers varied by two or more deciles across years.

 Noting the LA Times claimed that teachers’ VA scores will are not affected by “low performing students, ELLs or other students with challenges,. they write, “This statement is not supported by our analysis.”

 They conclude, “We might need to broaden our definition of teacher effectiveness from a generic perspective to a differentiated perspective, acknowledging that teacher effectiveness is context specific rather than context free.



Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s