I'm an Assistant Principal in a high school in Bergen County, NJ. We use D3M in just about every aspect of our functions.

For example, we analyze the Special Education students' early warning test data to place students in support classes for either Math or English. We also use this data to provide their teachers with specific weaknesses that each student has in these areas; this allows for clinical differentiated instruction.

We have also used D3M in creating a Suspension Alternative Program in our high school. I blog about this here: http://mikeparent.blogspot.com/2008/01/alternative-to-suspensions_02.html. I also have the exact data posted on Slideshare: http://www.slideshare.net/mikeparent/suspension-intervention-program/.

I have found that using relevant and current data strengthens the probability that action will be taken and ensures that a comfortable decision has been reached by program proposers and Boards of Education.

What's most important to me is the vehicles through which the data is derived. For much of my data, I rely on our IMS (PowerSchool) and student surveys. Lately, we've begun to use Scantrons for student opinion and survey data collection - this avoids the "eyeballing" method which can be inaccurate and tainted.