“Teacher Quality Widely Diffused” trumpets the headlines in the New York Times today (February 25, 2011). The headline and the article strongly suggest that the disadvantages of poverty and literacy-poor home environments are not critical influences on student performance on standardized tests. Rather that “teachers who were most and least successful in improving their students’ test scores could be found all around–in the poorest corners of the Bronx, like Tremont and Soundview, and in middle class neighborhoods of Queens, like Bayside and Forest Hills” (1).

The same article qualifies the results of the “value-added” assessment of students and their teachers by reporting “the margin of error is so wide that the average confidence level around each rating spanned 35 percentiles in math and 53 in English. . .”  This technicality may be conveniently ignored by the Times, but it is more than an inconvenience to teachers who are now publicly evaluated by their students’ test scores.

The media, the Bloomberg administration and the Obama administration are so hungry to get the goods on bad teachers, that they are willing to sanctify any statistics that appear to support their case.  “Value-added” statistics are a clear improvement on evaluating teachers on the raw data of their students’ test scores, but with a confidence level that spans 53 percentiles in English, there is still much to question about publishing such data.

Suppose the verdict of a jury had a 53% variance with the truth?  Suppose the testing of a drug to cure HIV had a 53% confidence level of success? Suppose the computer models of an air assault on the nuclear resources of Iran had a 53% chance of disabling their nuclear program?  Would anyone take these risks? Are these test scores any less damaging of the reputation and the professional survival of a school or a teacher?

The cases that seem to fall outside the range of probability in the field of statistics are often referred to as “outliers.”  Outliers are often subjects of further experimentation, because they may speak to the validity of the data that falls within the confidence levels of the data.  Thorough scientists do not ignore outliers, because they may reveal flaws in their original hypotheses. They investigate outliers more rigorously to learn what they can from the deviations.

That is not what is happening with the “value-added” data offered up by the New York Public Schools. The data is being privileged with a public showing and sanctified by a headline like “Teacher Quality Widely Diffused.”  In criminal prosecution this would be called a “rush to judgment.”

In the media, we should call this “out lying.” The data is out, even though some of it may be lying.  It is all well and good for schools to use the data for discussion and give it further scrutiny to see what it really says. It is another thing to pretend that the data is evidence that poverty is not a mitigating influence on teaching.  This is what I get from “Teacher Quality Widely Diffused.”

Let’s not use blunt instruments to execute teachers. Let’s investigate the outliers, not lie about them.