Ed Sector proposes that we simply ditch those teacher aides whose jobs are created via the contract. The research support (which Ed Sector has been patting itself on the back about generally) comes from the Tennessee STAR research. I love Tennessee STAR. I take its findings regarding teacher aides very seriously. I want to suggest that those findings are being applied improperly here.  

STAR had three comparisons. The control group consisted of classes that were at the status quo. The experimental groups were classes with fewer kids and classes with dedicated full-time aides. The research did show that full-time aides generally didn’t improve achievement more than the status quo. But, I want to unpack that finding.

An AERA paper called “Do Teacher-Aides Improve Student Performance? Lessons from Project STAR” by John Folger and Carolyn Breda found that classrooms in the control group actually had an aide present at least part ime for 17.9 days a month out of 23 instructional days--there is simply not an “aide/no aide” comparison in the data. STAR shows that the presence of a full-time aide, as mandated by the state, is not better than the presence of a part time aide put in place as a result of both local and state programs.  Understanding the real comparison discredits the idea of cutting aide positions created as a result of the local process of collective bargaining. 

Folger and Breda also found that the new aides “did not have to be certified, or have any specific educational background," had “no special training programs . . . most aides received no special training in their duties, and teachers did not have any training in how to utilize an aide effectively.”  It seems to me that this is where the most important and generalizable conclusion of the research is found. If you don’t train people appropriately and use them in a way the research suggests will help, then things won’t go well. 

Another paper by Chuck Achilles found that the classes with full-time aides had the best outcomes overall for students who had been retained in grade. My suspicion is that the additional aides tended to work one-on-one with students, although there is no documentation about classroom practice to back that up. Other research shows that aides are quite effective when used as tutors. For example, tutors are an integral part of the Success For All (SFA) program. While SFA creator Bob Slavin prefers certified teachers to do this work, many official SFA schools use aides.  Slavin and Toks Fashola's book Show Me The Evidence cites a number of studies (some going back to the 1960’s) that back up this practice. Other programs you might have heard of that use aides are Direct Instruction, where aides work in small groups (you’re messing with DeRosa now), and METRA, where aides, again, do tutoring.