Data, Data, Data... and the associated pressures

Just a quick thought or two on some recent conversations I had related to data quality in our student information systems. Working around K-12 education, I reflect at times on how different things are - or how different they seem - since I was in school. There's data on everything and there's a well-intentioned desire to use that data to improve learning, assess better teaching, have a healthier school environment and things of that sort.

As that data gets analyzed, we see examples all around where the knowledge that this data exists causes changes in behavior. This school security site discusses school crime reporting... and under-reporting. As the link points out, what is my incentive to report every violent incident? Unfortunately, there's quite a bit of incentive to under-report and show that violent incidents are decreasing. No one wants to look like their school is a war zone; consequently, there are whispers that violent incidents should not be reported. That can lead to teachers having their hands full in the worst of cases with students who need to be reported and know that they will not be reported.

Another example I've heard about is the confusion around "in-school suspension". I've been told that districts are to track these suspensions in their student information tool. However, I've also talked to a couple of contacts who gave me an example where a student might have been pulled out of class... not for a 'suspension', but because they're behind in another subject and need additional time for homework or tutoring. That sounds noble, but I was hearing from a teacher who pointed out that she had good students asking her why they should do their homework. Apparently, these students saw that others were - in their eyes - simply being given more time or another chance to complete an assignment. In this case, the school didn't want to flag the students as a 'suspension', nor did they want to give the child a zero on a homework assignment.

I suppose my question is 'how do we get truly accurate and useful data when the overriding concern by some is that the data simply look good'? Perhaps if we weren't so punitive in our discourse and were simply concerned with accuracy, we could get some data that could be relied upon more often.
[Image: http://teflbootcamp.com/tefl-skills/student-discipline-efl-classroom/]