Data Workshop Series: Predictive Analytics and Early Warning Systems
TOPIC Predictive Analytics and Early Warning Systems for End of Year Academic Performance
PRESENTERS Steven Glazerman, Mathematica Policy Research
AUDIENCE School Data Users (Preference will be given to LEAs)
DETAILS Do you want to feed all your data streams into a predictive model for early warning, to find out in real time what each student needs to be successful on the end of year exam (PARCC)? This workshop will describe an approach, called Bayesian analysis, that produces statistically sound, actionable results -- even from small samples -- that can be used by educators. This is a model where you input prior beliefs about performance expectations, add data at various points, and after each new refresh of the data, update those beliefs based on all the data in hand. Workshop participants will have a chance to see an early version of the predictor and contribute their ideas for improving the next generation of this tool.
Light refreshments will be provided.