Tying Teacher Evaluations to Test Scores is a Good Idea

If you ever hang out on your schools playground and talk to other parents, chances are you have a good idea of who the good and bad teachers at your school are.  But the teachers everyone loves are often not rated highly on student test scores.

The New York State tests are poor gauges of teacher quality because they aren’t designed to gauge teacher quality. What the NYS exams do measure is a set of three factors:

a) What a child has learned in school
b) What a child does outside of school (studying, home life, etc.)
c) A child’s intelligence or test-taking ability

Only (a) relates to what a teacher can control, and evidence suggests that this factor is less important than the other two on tests oriented around a bell curve—such as the New York State exams. (See  “Why Standardized Tests Don’t Measure Educational Quality,” Educational Leadership, March 1999).

Small wonder, then, that statisticians who have studied test-based teacher evaluations find they don’t work. American Education Research Association, the National Academy of Education; and they American Statistical Association all say the same thing: test-based teacher evaluations are not reliable measures. (Washington Post, 4/13/2014)

Perhaps the most obvious evidence that these metrics don’t work is that the results aren’t stable.  A teacher rated “effective” based on test scores one year is quite likely to be rated “ineffective” the following year.  One third of all teachers move at least one category in a given year. Only 16% are rated highly effective two years in a row. (See the NYS Education Departments’s 2013–14 Growth Model for Educator Evaluation)

When you think about an amazing teacher, do you think that teacher is substantially less amazing the following year? Is she randomly terrible every few years?  Yet this happens all the time under test-based rating systems. (See, for example, Newsday, 8/12/2015)
This isn’t just happening in New York. Here is a photo from a press event in Albuquerque, where several “effective” and “highly effective” teachers burned their evaluations. One teacher said she was burning hers because the teacher who mentored her—someone she greatly respected—got an ineffective rating.

Why is this happening?

Governor Cuomo and others have complained that too many teachers are rated as effective. They argue that if 75% of students aren’t meeting academic expectations, a lot of teachers must be at fault.

Many of us at NYC Opt Out agree that the way teachers are evaluated and trained needs to be improved. There are lots of ways to do that but they involve developing training and mentoring programs, which cost money. Test-based evaluations are, despite their costs, cheaper in the short-term.

Meanwhile, the effects of this system are taking its on teachers. A national study by the NEA found that nearly half of all teachers have considered leaving the profession due to testing. Those that actually have quit have taken to YouTube, where their letters of resignation have practically become a genre.  Applications for teachers colleges are down across the US. In many cities, there are massive teacher shortages.

Teachers are professionals and need to be treated as such. As professionals, they must be held to high standards and rigorously evaluated. But those standards and evaluations need to come from trained, seasoned educators–not politicians and bankers.

See also:

Who Rates Teachers This Way?
(The Synapse, 2/12/2015)

How Do High-Performing Nations Evaluate Teachers?
(NEA Today, 3/25/2013)