On the dangers of using data

Here's a really nice explanation by Jonathan Gray's Data Blog on the Guardian, spelling out five cautions for use of data in decision-making.

http://www.guardian.co.uk/news/datablog/2012/may/31/data-journalism-focused-critical

And here's an example of why data is bullshit. You can't measure this program with an end of year test. In fact, there are lots of things you can't measure that way.
  http://www.slate.com/articles/technology/future_tense/2012/06/maker_faire_and_science_education_american_kids_should_be_building_rockets_and_robots_not_taking_standardized_tests_.html

And when you ignore data showing that what you do doesn't work, here's what can happen. http://www.startribune.com/local/156812385.html