Marianne Bitler, Hilary Hoynes, and Thurston Domina have a very interesting preliminary paper that looks at the differences between the effect of Head Start for kids with good language skills as compared to the benefit for kids with not such good language skills. The authors examine an experiment where some kids were offered Head Start and others were not. (The random assignment experiment was implemented at Head Start centers that were too crowded to take all eligible students.)
The authors first remind us that there is now lots of evidence that Head Start gives kids a boost, although there is an open question about whether the boost fades out in elementary school. Next, the research team describes what they had to do to make up for the not-very-clean data collected in the experiment. The not-clean data probably made it harder than necessary to get clear answers from the experiment and the results reported in the paper sometimes differ depending on exactly which students are looked at and how long the students were in Head Start. But one particular finding is quite clear. One year of Head Start for three year-olds had a large effect in the short-run at least. And most of the gains were won by the kids with the worst language skills. Here’s the team’s picture.
The red line shows students at all ability levels gained. Look at the left side of the graph though. The kids who were furthest behind had the biggest gains. I suspect that’s consistent with the goals of Head Start’s founders.