My recent post on comparability and Title I is here. Researchers have cautioned that correlations at high levels of aggregation do not need to be true at lower levels of aggregation, and can even reverse, Simpson’s paradox. And so we found with spending on teachers. When data are analyzed in the aggregate, it appears like spending on teachers is lower in high-poverty schools. When data are analyzed at the district level, spending on teachers is about the same regardless of whether schools are high or low poverty. Our finding that teacher experience is about the same regardless of school poverty is interesting too.
If you enjoyed this post, please consider sharing it.
