(Guest Post by Matthew Ladner)
A careful analysis of test score gains by David Figlio of Northwestern University has found a modest but statistically significant gains for Florida tax credit students. The data in this study are messy, and Dr. Figlio admirably goes about sorting through the various issues in an even-handed fashion.
Figlio employs a regression discontinuity design to analyze the data, and his finding of a small but statistically significant academic gain fits quite comfortably with the larger random assignment literature, which find small year to year gains which accumulate over time.
One of the under-appreciated features of the random assignment literature: the studies usually fall apart after three or four years due to attrition in the control group. Our window into the academic benefits of choice is therefore limited. Figlio’s employment of a different analytical technique provides confirms previous findings, and may (?) open the door to longer term assessment. The challenges with the data described in this paper, however, suggest that it may not be easy.
Money quote from the study, with a definite echo of previous random assignment studies:
These differences, while not large in magnitude, are larger and more statistically significant than in the past year’s results, suggesting that successive cohorts of participating students may be gaining ground over time.
Good discussion of the results over at RedefinED, including a discussion of the baseline results (tax credit students are poorer and less Anglo). Emerson also puts this study in the context (Figlio also found positive public school effects associated with the Step Up for Students program).
So, the Step Up for Students program has now been found to help improve public school results, help improve participant academic gains, generates high levels of parental satisfaction. Sounds like a rock solid justification for expansion to me.