Data Driven Instruction - Looking Forward and Predicting
Aug 01, 2011
Data Driven Instruction is yesterday’s news or is it. In K-12 we do have a tendency to embrace the latest and coolest. We can change from year to year depending on what new idea we (our administrators) saw at a conference, the Intermediate Unit now recommends, or perhaps it’s the newest ideas in the current hot selling book.
The district has workshop days to fill with training and often times there is no consistent yearly or multiyear theme. One year I went to inservices on how to incorporate music into learning, multiple intelligences, peer mediation, self defense in the class room (don’t ask), and a health fair. All were interesting and had good information but there was no underlying connection.
Staff training should be based on the needs that were determined in the strategic plan. Instead we often adopt the latest hot ideas for K-12 pushed by the most popular consultants and authors. The result is that data driven instruction is in danger of being supplanted by the next best thing. Using data in the classroom is not sexy but it is a proven research based concept. From finance to farming using data to make decisions is a normal and required process.
In K-12 we also use data but in a different way than other industries. According to Bill Eriendson, assistant superintendent for the San Jose Unified School District in California, “School districts are great at looking annually at things, doing summative assessments and looking back, but very few are looking forward. Considering that our economy survives on predictive analytics, it’s amazing to me that predictive analytics don’t drive public education. Maybe in education it’s considered a luxury, but it shouldn’t be; it should be the foundation for making decisions.”
Why do we not use predictive metrics in K-12? It seems that there are plenty of analysts that can crunch the numbers and produce the key analytic metrics for most domains. The barrier in education is not the number crunching but understanding the pedagogy and social issues which are vital to producing predictive indicators. A good K-12 analyst must know both.
When those predictive factors are discovered there will be a wealth of research based indictors. Imagine a true student early warning system that produces a hot list of students early on that gives the district time to intervene on specific determined areas in order to resolve the key issues.
I know that’s still far away but we are making progress toward that goal. Research from the Chicago and Philadelphia schools have determined a few interesting predictors. Below are two examples of recent predictive research.
Among Philadelphia sixth graders who failed math and/or English, over 80 percent did not graduate within a year of their expected graduation. Although course grades were found to be highly predictive of falling off the graduation path, by comparison, fifth and sixth grade test scores were not (Balfanz, Herzog, & MacIver, 2007).
- A few absences matter greatly: nearly 90% of freshmen in Chicago Public Schools who miss less than a week of school per semester graduate in four years,compared to just over 60% of students who missed about one week (equal to five to nine days) (Allensworth & Easton, 2007).
- Lots of work needs to be done. It will take some time for the indictors to be determined. Analyzing the data is tricky. For example an interesting part of the research showed that in 9th grade attendance was a predictor of student graduation but in the student’s sophomore year grades are a more accurate indicator.
Teachers and administrators armed with this type of data can effectively target areas that have a huge impact on student success. Predictive data opens up a new data analysis territory and data informed instruction takes on a new meaning.