The Ultimate Measure of a Teacher?

" ... As I think about what made my teachers great, and listen to others' reflections, I notice that for almost all of us, what resonates with us are how those teachers made us feel -- more confident, more capable, inspired or loved -- and the life lessons and values they taught us.

The one thing I don't hear people saying is, "I loved Miss Such-and-Such because she helped me get better test scores!" It seems that what matters most to us can't be easily measured.

And yet, one of the big trends in education right now is value-added measurement (VAM), where a teacher's effectiveness is calculated by measuring the growth in their test scores. On its face, the idea makes sense. If standardized tests measure what students know and can do, and teachers are solely responsible for increasing what students know and can do, then we should keep track of how much students' scores rise while in their teachers' classrooms. That way, we can reward the best teachers and disseminate their practices, and identify the worst teachers for extra help or dismissal. Simple as Simon, easy as pie, right?

Not really. First, standardized tests offer only partial information about what students know and can do, and only in a limited range of subjects. Second, teachers aren't the only ones who teach students. Family members, friends, other teachers, librarians, tutors, counselors, friends, and the student him -- or herself (imagine that!) all influence what is learned over time. Trying to devise a way to calculate the influence of one particular teacher amidst all those others? Remember, each teacher's relative influence will be different for each learner. Good luck!

Other problems known to plague value-added measures (VAM) for individual teachers include:

High error rates. While potentially useful when assessing schools or districts, there is a one in five chance that an individual teacher's effectiveness will be mischaracterized when only three years worth of student data are used. That rate rises to 35 percent (over one in three) when just one year of data are used.

Instability. Teachers' scores can change considerably when different tests are used.


Difficulty separating student characteristics from teacher influence. For example, factors like race and giftedness have been shown to have an effect on teachers' value-added scores.

With a large enough sample size and random assignment of students, some of these issues could be minimized. But that would require school leaders to randomly assign students to teachers, and wait several years for teachers to accumulate enough data points (formerly known as children) to be sure that any inferences drawn from the calculations are reasonably accurate.

And of course, those problems are stacked on top of the other problems already associated with high-stakes standardized testing. These include, but aren't limited to: bias toward certain kinds of learners, pressure on schools to narrow the curriculum, the temptation to game the system and/or cheat, and so on. That is why groups like the National Academy of Sciences and the Economic Policy Institute agree that VAM should not be used to make high-stakes decisions about teachers.

If we truly respect teachers, and are serious about ensuring that all children are taught by high-quality teachers, why redesign teacher evaluation systems around measures that most researchers agree are unstable and misleading? (So much for being data-driven...) Why would any school district use a tool they don't actually understand to inform high-stakes decisions? Under these circumstances, it is irresponsible to use VAM for any significant proportion of an individual teacher's evaluation -- especially if the other components include a "drive-by" principal observation and little else. ...

Here's hoping the current generation of students can look back on their schooling and remember caring, creative teachers, instead of stressed-out test-prep technicians!"

-----

Read the entire piece on huffingtonpost.com.
http://www.huffingtonpost.com/sabrina-stev...f=fb&src=sp