Last summer, I stumbled across a clever 1993 paper by education statisticians Mark Dynarski and Philip Gleason that proved it was possible to adjust average state SAT scores for variations in the test participation rate and demographic factors, making them comparable to one another. Barely able to contain my excitement (hey, don’t judge), I set about extending their method so that it could discern trends in state SAT scores over time, and improving the validity of its estimates by using more data, fewer assumptions, and more exhaustive methods. Last week, I released the technical paper presenting those extensions. Yesterday, I released a paper that uses them to chart academic achievement and spending trends, for every state, back to 1972. How did your state do? Find out here.
To pique your interest, here’s the chart for Masschusetts: