"Um...no...data isn't all that important." - Bill Gates

Don't worry.  Every decision will be data driven.

We've heard that a million times lately.  It is meant to reassure us.  But we are finding more and more of this data, that somebody just HAD to collect,  that is not seeing the light of day and that should concern us.

Forbes just reported that Department of Health and Human Services (HHS) has been delaying the release of the final results of the Congressionally-mandated evaluation of the federal Head Start program, which impacts millions of American children. The study of 5,000 preschoolers was completed in 2008, but the final results have been delayed for four years. Why so long HHS?  Could it be the "data" showed this?

"The study found that, compared to their control group peers, Head Start failed to boost students’ cognitive abilities across 41 measures. Moreover, first grade teachers reported that former Head Start students were actually less prepared in math than the non-Head Start students."

 $180 billion over the life of the program and the data are beginning to show that it doesn't do very much good.  The part of the report we are still waiting on is whether the program had effects on third graders which is due out this year (maybe).  Strangely, ed reformers like Michelle Rhee and Howard Dean are out pushing their 0-6 agenda, trying to get children into government run formal education as early as possible. The data should have every parent asking, "Why?"

More data that is not seeing the light of day comes from the popular Value Added program for teacher performance evaluation. Using the actual value-added scores of thousands of New York City elementary school teachers from the school years that ended in 2006 and in 2007, one researcher found absolutely no correlation between student performance and value added scores.

From the author himself,

In any introductory statistics course, you learn that a graph like the one above is a textbook case of “no correlation”. I had Excel draw a line of best fit anyway, and calculate an r-squared correlation coefficient. Its value? 0.057 — once again, just about as close to zero correlation as you are ever going to find in the real world.

In plain English, what that means is that there is essentially no such thing as a teacher who is consistently wonderful (or awful) on this extremely complicated measurement scheme. How teacher X does one year in “value-added” in no way allows anybody to predict how teacher X will do the next year. They could do much worse, they could do much better, they could do about the same.   -  Gary Rubenstein

If the people at the administrative top are going to keep pushing to collect more data, like the longitudinal data system, then we need to start demanding that they follow the data. When something they're trying out through the free laboratory of our school system doesn't work, they must follow the data and eliminate it.