Grade Retention Can’t Be One-Size-Fits-All

Low standards do cheat students, but a motion set to be heard today by the Los Angeles Unifed School District board misses the mark. Board member David Tokofsky’s motion seeks to retain any second- or eighth-grade student whose Stanford 9 reading score is ranked at the lowest 5% nationally (that is, below the fifth percentile).

Closer inspection reveals that a fifth-percentile threshold for student retention is both too low and too high. The fifth percentile is too low in that we should ask more from every student. It is too high when it causes any student to be retained who should not be retained.

The general premise of the motion seems to be that students who read below grade level should not be promoted to the next grade. The motion relies on several critical assumptions that do not hold water, three of which are examined here.

First, the motion assumes that students who score below the fifth percentile in reading on the Stanford 9 cannot read. Yet, recent work by David Rogosa, a statistician at Stanford University, shows that the test is not a very reliable measure of the true ability of individual students. My review of sample data on LAUSD students supports this finding. As many as 2% of the students who scored at or below the fifth percentile in 1997-98 scored above the 40th percentile in reading on the test the next year. This raises the specter that many at or below the fifth percentile can actually read at grade level, notwithstanding their Stanford 9 score. Perhaps low-scoring good readers were ill, distracted, worried or hungry when they took the first test. Retaining such students in the same grade has no clear benefit.

A second assumption of the motion is that if a student cannot read, he should be retained even if he performs well in other subjects. Hundreds of LAUSD students who score at or below the fifth percentile in reading score above the 40 percentile in math, for example. Retaining second-grade students in the same grade for another year to improve their reading means that they would also be subjected to a low-level math curriculum. This is akin to hammering a fly on a plate-glass window. Fortunately, many schools employ effective swatters that harm only the fly. However, a centrally mandated retention threshold would force these schools to pick up their heavy hammers again.

A third critical assumption is that students who do not learn to read at grade level would be better off remaining in that grade. To buy into this we would have to ascertain that high-quality teachers and programs are in place to re-teach the student. What sort of education would the student get the second time around with the same teacher? Consider further that retaining a poor reader in second grade could prove detrimental if the student’s prospective third-grade teacher is a lot better at teaching reading than the second-grade teacher.

The upshot is that a centrally mandated student retention threshold has the potential of doing more harm than good.

Rather than use Stanford 9 reading scores as thresholds for retention, the district should continue using them as a criterion for targeting students for closer attention. (It is worth noting that the district apparently already keeps a close watch on students scoring below the 30th percentile. Indeed, many students scoring way above the fifth percentile have been retained.) For students who score below the threshold, teachers should have to justify their recommendation to either promote or retain. Moreover, the district might consider the feasibility of requiring that schools develop individualized education plans for all students who perform below, say, the fifth percentile.

More than anything, what’s needed is a set of policies that seed the district’s lowest-performing schools with the highest-quality teachers and principals and then frees these professionals to exercise their intellect and creativity to educate children at the highest levels. The board and central office must continue to wean itself from the urge to mandate change from on high.