The American Educational Research Association brought journalists to Chicago for Data at Your Desk training, giving us an overview of nationally collected education statistics and a serious dressing down for media misdeeds.
The most common failing by the collective us involves saying because two things occur together, that one causes the other.
For example, if 61 percent of third-graders in this county are not reading at grade level, which is true, and high numbers of third-graders in this county eat grilled cheese sandwiches, which is also true, then eating grilled cheese sandwiches must cause reading problems.
We can all laugh at that one, but then look at the number of times political talking points will be laid out in the same way, and we – the media – keep quoting them.
Premium content for only $0.99
For the most comprehensive local coverage, subscribe today.
Another pet peeve was sloppy studies that, in essence, ask anybody anything and then lay it out as representative statistics.
“You’re going to see people taking incredibly poor-quality data and making decisions on it,” said Jack Buckley, senior vice president of research for the College Board.
Interconnected loops of learning stood out at this conference devoted to research about education. These folks study educators, learning what works teaching kids, to better teach educators to better help children.
The topics studied touched on use of technology, evaluating teachers, principal effectiveness and every hot-button issue out there, each examining with microscopic precision a thin slice of much larger pies.
Here are a few tidbits of knowledge I gleaned from a morning spent dropping in on different researchers’ discussions on data use by schools:
Data-based decision-making in classrooms appears to work best when teachers and principals gather and analyze their own data, rather than having an outside vendor crunch the numbers. Much seemed to be gained by doing the analysis rather than just having the numbers.
Other positive correlations were found when there was a culture of collaboration and lots of diversity in the mix, allowing teachers with different skills and points of view to help each other.
Strong social networks and low centralization – in other words, well-connected teachers who made decisions about instruction – were also strong positives.
Where teachers have a high level of comfort in working with data, it built capacity and worked well. Where data skills were modest, dumping a load of numbers on their desks appeared to have a negative effect on classrooms, and where teachers had little or no training, it was just a struggle to use, researchers found.
Note: Teacher skills and training correlated with positive or negative effects from having lots of data. I am not saying they caused it.