Can data fix our high school dropout problem?
Michigan Radio social media guru Kimberly Springer sent me a write-up about a new data system being sold to school systems as a way to identify potential drop-outs. I was interested. She was suspicious.
Part of this is probably her affinity for (obsession with?) all things Matrix. But Springer has also participated in a learning analytics fellowship at the University of Chicago. One thing she says she learned there was that data is just like any other information. "It can be taken out of context," she said.
The data in question here are information about individual student performance as measured by grades, standardized tests, and attendance. The data get crunched by this company, LearnSprout, and what's given to school administrators is a birds-eye view of which kids are on track, at-risk or somewhere in the middle.
For her part, Springer would like to see a more student-centered approach, one that uses data but isn't totally driven by it. "Data is not as pure as people think it is. It's not CSI," she says. "It's real life, it's much more fallible."
Schools are under tremendous pressure to meet student achievement goals, and they have fewer financial resources to experiment. It's no wonder many of them turn to data as a way to help them better direct whatever resources they have more effectively. In Michigan, the Response to Intervention strategy is popular. It involves tracking student achievement on a large number of indicators very closely. The idea is that teachers know best who might need extra help on any given task.
The inclusion of attendance data as part of the LearnSprout warning system is what I find interesting.
There is no question that when kids miss a lot of school they are at risk for dropping out later. In Michigan, about 10 percent of kids drop out. But knowing which kids are absent a lot and being able to match that up with their performance doesn't necessarily give a lot of answers.
Absenteeism and especially truancy are often indicators of much bigger problems at home. Whether or not schools can then address those problems is, of course, another question. The upshot of schools having a data-driven approach is that if a school does try to address whatever is going on in a kid's home, they'll know if that had an effect on how that kid performs in school.