• Gary Rubinstein | TeachForUS.org

You know that scatter plot I posted yesterday that appeared to show almost no correlation between the "value-added" performance ratings of teachers teaching the same subject to different grade levels in the same year? Well, rather than being completely random, one would have expected it to look somewhat like the graph above, with the data points clustered near a diagonal line rising from bottom-left to top-right.

Yesterday's graph illustrated the utter failure of the value-added model to measure teacher performance in NYC public schools. Today's graph illustrates the middling performance of NYC charter schools while refuting the claim that charters disproportionately serve high-needs low-performing students.

The blue markers represent NYC's traditional public schools, while the red and yellow markers represent charter schools, with the chart plotting the average change in English Language Arts (ELA) scores (0 being the 50th percentile) from the end-of-year 4th grade tests (x-axis) to the end of 5th grade, the first year of middle school. Read Gary Rubinstein's TeachForUs.org blog post for a more thorough explanation.

A few details stand out. First of all, notice that most of the charter middle schools are below the improvement trend line, suggesting that charters are not adding as much value as the average traditional public school (a second chart shows charters doing about average with math). Second, notice that the 4th grade scores of incoming charter school fifth grade students are all clustered near or above zero, indicating that charters are attracting average or better students, not the high-needs students charter backers claim they want to serve.

It is true that some NYC charter schools boast very good test results, but they also tend to start with better than average students (while kicking out the troublemakers)—and as the high correlation plotted on this chart shows, the best predictor of a student's fifth grade state test performance is his or her's fourth grade state test performance. Rubinstein elaborates:

If there was a wide variation between teachers’ ability to ‘add value’ this plot would look much more random. This graph proves that when it comes to adding ‘value,’ teachers are generally the same. This does not mean that I think there are not great teachers and that there are not lousy teachers. This just means that the value-added calculations are not able to discern the difference.

That's what we could be getting with the teacher evaluation legislation Governor Chris Gregoire just signed into law today: A standardized teacher evaluation system that fails to measure the difference between good teachers and bad. On the bright side, at least we once again managed to stave off charter school legislation, an education "reform" that consistently fails to deliver on its overhyped promises.