Jay Greene initiated a discussion of federally-funded education research. Harriet would have left a comment, but it went over-long.
The tax-funded education research apparatus provides an endlessly entertaining playpen for the statistically literate puzzle addicts who become quantitative researchers and to the proponents of non-quantitative research such as Jim Horn, who's authority depends on the authority of the statistically literate empiricists. Both groups deserve the mistrust which Jay Greene advocates in his essay.
Professor Greene's recommendation, that legislators restrict the Federal role to accumulation and distribution of raw data, likely would lead eventually to the situation which he laments. What data would the Feds amass? Why suppose that the people who collect these data would perform any more honestly than climate scientists who report favorable data from non-existent stations in China and omit stations which yield data contrary to their preconceptions? Substitute "school" for "weather station" or "proxy source" (fossil tree, stalagmite, glacial gas bubble), and "standardized test" for "thermometer" or "tree rings", or "isotope ratio".
Some Texas financier once said: "Money is like manure: when you spread it around it can do a lot of good; when you heap it in one place, it stinks."
Government policy may effectively address problems which arise from free rider issues and tragedies of the commons, such as overfishing, or from other sources where free rider problems inhibit market solutions, for example, the threat from Earth-crossing asteroids. I expect, however, that, given a sufficiently large budget for a sufficiently long time, the Asteroid Defense Agency would spend its budget on plush conference venues and overpriced, redundant studies, and, come crunch time, perform as well as the French army in 1940. We're screwed, or we're robbed, then screwed.
Education does not belong in either of the above categories. Parents have an alternative to the wasteful and abusive, tax-funded K-PhD school apparatus. Homeschool. The market (meaning, whatever happens in the absence of State coercion beyond the protection of individuals and property) will generate more reliable information on effective practice than will PREL (nee the Pacific Regional Education Lab, once an offshoot of the Northwest Regional Education Lab and now the "Pacific Resources for Education and Learning").