Ed-Spin: Reading the Tennessee Star Study

by Martin Harris 

Martin Harris

Probably no single researcher has used his academic chair and analytic skills to make sense of the 40-year-long edu-crat campaign to reduce class size more effectively, statistics-wise, than Professor/Doctor Eric Hanushek, formerly at the University of Rochester, then de-fenestrated for his politically-incorrect findings, and now at the Hoover Institution of Stanford University, where he continues to undertake meta-analyses (statistical compilations of large numbers of specific studies) of the “smaller-class-size-produces larger student achievement gains” argument. Close runners-up are economist Richard Vedder, formerly of Ohio University, who drafted the costs-up-results-stagnant 40-year record in terms of industry-wide productivity decline, and who even so managed to make it through to retirement without penalty; and Caroline Hoxby, formerly of Harvard, whose research reported that public schools faced with fleeing students were more statistically likely to improve their teaching (as shown by test scores) than those which weren’t so existentially threatened. Facing imminent of de-fenestration, she joined Dr. Hanushek at Stanford. In 1998 he published “The Evidence on Class Size” which remains the single most succinct de-bunking of the supposedly “common-sense” thesis that, with fewer students, a teacher can get better results. It also remains almost unknown, its politically-unwelcome message making it worthy of media suppression.

That’s the same media which, handed each year the latest achievement-test-scores from the Federal tests –in the low- to mid- to a few upper-200’s out of a possible 500– consistently and happily publish in print and electronic venues the “we-are-excellent” assertions from State and local educators, even though such scores mean that the majority of their students can’t make “Proficient”, which means, roughly, that they’re non-functional in grade-level math and reading.

Of the thousands of studies, large and small, purporting to confirm the seemingly logical “small-classes-are-better” thesis, the mid-‘80’s Student/Teacher Achievement Ratio study in Tennessee schools has been and still is quoted most frequently. In Harvard Magazine, professor/author Frederick Mosteller cites the “substantial improvements” in those classes set up with small section sizes (13 to 17) compared to the then-usual section sizes of 22 to 25, but doesn’t print the findings in numerical or graphic form. Nor does he mention the overall poor level of reading-and-math achievement, in TN schools just as in all the other States’ K-12 operations, especially Vermont, with smallest-classes-in-the-nation. Writing in “The Evidence on Class Size” Eric Hanushek does: in a telling little chart mapping the smaller and the larger groups by SAT results, he shows them as mostly matching, major gap maybe a dozen points, tops, out of 480 to 640, for grades K-through 3. That’s after a 1/3 class size cut. Hardly “substantial”. The score lines for larger classes, with and without teacher aides, show zero benefit from the presence of aides, a subject Hanushek doesn’t pursue in his report. He does discuss in greater detail a subject avoided in the Harvard Magazine piece: the impermanence of the 10-12 point K-3 advantage declared “substantial” by small-class advocates like Mosteller. Parallel achievement scores of the same students as they moved on through grades K-3 showed that, as Hanushek writes, “…if small classes were valuable, the achievement gap would widen. It does not. In fact, the gap remains essentially unchanged through sixth grade, even though the experimental students return to larger classes for the fourth through sixth grades. The inescapable conclusion is that small classes at best matter in kindergarten. ” Humble Scribe add-on: the stats show that reducing class size did little to no good, achievement wise; raising it didn’t hurt achievement levels, either.

* * * * * * *

The little chart of K-3 math results in “The Evidence on Class Size” (showing no achievement-differential beyond Kindergarten),and the statistical references to no cumulative improvement or backsliding beyond K-3 either, even as the same students re-enter larger class sections, bring to mind that other 40-plus-year-long experiment in public education, the introduction of pre-K classes in the mid-60’s. Just as with the “smaller-classes-are-better” doctrine, Head Start advocates have long claimed “substantial” (and long-standing) gains for students who had the advantages of those introductory programs, and, similarly, detailed studies (never deemed worthy of media attention) confirm that the gains are non-existent. For Head Start, the most recent of many was the US Government Accountability Office study of 2010, which found not only widespread outright fraud (non-poverty students enrolled) but widespread long-term ineffectiveness. When the Heritage Foundation published a report on the GAO study, defenders of Head Start dismissed it by calling Heritage names: right-wing ideologues, and so on, but didn’t dare challenge the GAO’s own statistical findings. Some quotes from the Heritage report:

“The national evaluation tracked the progress of 3- to 4-year-olds entering Head Start through the first grade. Overall, the program had little to no positive effects for children granted access…compared to similarly situated children not allowed access…[HS] failed to raise the cognitive abilities of participants on 41 measures; specifically, the language skills, literacy, math skills, and school performance failed to improve. Alarmingly, access to HS for the three-year-old group actually had a harmful effect on the teacher-assessed math ability of those children once they entered kindergarten. Teachers reported that non-participating children were more prepared in math skills than those children who participated in Head Start.” (Parental pre-K?)

As with the weakness-under-analysis (confirmed by more than 40-years of shrinking class sizes and stagnated Federal test scores) of the “smaller-classes-are-better” doctrine, the governmental-sponsored Head Start too fails under impartial governmental analysis, and neither set of findings gains much media attention, although both have gained a lot of educator denial and misrepresentation. Meanwhile, the educational establishment is now faulting parents for inadequately pre-K-ing their kids (my generation of parents was told to keep our clumsy hands off and deliver the kids, illiterate and undamaged, to the professionals at school) and anecdotal reports of quite successful private-sector pre-K programs go unreported by local and national media. You can guess why.