By Dylan Wiliam
I know you’re from the UK and have spent a lot of time in various countries overseas. Can you share a little bit about your experiences and key accomplishments in the United States?
In 2003, I started working as Senior Research Director at ETS in Princeton, NJ. My vision was to try to make classroom formative assessment a reality. As a result, we started doing lots of exploratory work with local school districts to find out how they could best be supported to develop classroom formative assessment.
We did some rapid prototyping of ideas, worked in local urban districts like Trenton to see what worked, made changes based on what we saw, and tried again. I also worked with lots of schools around Philadelphia on a math/science partnership that was funded by the National Science Foundation, and work in Vermont and on the west coast in Seattle in a couple of districts out there. Working in these districts as a whole helped us develop and refine the two-year PD Pack that is now distributed exclusively in the US by Learning Sciences International.
In Florida, third-grade students take a reading and math test at the end of the year and in fourth grade, they take a reading, math, and writing test. I know a third-grade teacher who taught a lot of writing and as a result, all of her students scored at the highest level the following year on the fourth-grade writing test. Although it may look as if the fourth-grade teacher made a huge impact, the outstanding results could actually be traced back to the third-grade teacher. How do teachers usually react to stories such as this one when you do presentations for them?
When presenting to a group of teachers, I try to move away from focusing on individual teachers and try to get them to focus on the moral imperative. If you just focus on raising test scores, some teachers say, “Well, I get the best test scores in the district or the state, so why should I improve?”
So when I present to teachers, I try to help them understand that even if they’re the best teachers in the state, they still need to get better, because when they get better, their students are healthier, live longer, and contribute more to society.
What is really interesting is that when administrators hear that, they sometimes ask me to skip over this part because “they already know that” and try to get me to focus on the tips that teachers can use in their classrooms. And of course, I refuse to do this, and teachers always tell me afterwards that it was very valuable.
The specific example from Florida is extremely important. Everybody knows that evaluating teachers on the scores that their kids get is a really bad idea because too many variables are out of a teacher’s control, but this example shows why even “value-added models” don’t work. If a third-grade math teacher spends lots of time on problem solving, autonomy, and independent learning, that won’t show up in the third-grade scores, but it will show up in the fourth-grade scores, so the person who gets the credit is not the person who is actually adding the value.
These are great examples of why you can’t judge a teacher’s quality by just comparing students’ test scores when they started with a teacher to their scores when they finished with that teacher.
A closing thought on the moral imperative: as a society, if we only help the weakest teachers improve, we’re not going to improve the average levels of achievement by enough to ensure our students can find work and thrive in the 21st Century. And every country faces this problem—not just the US.
Can you explain what you mean by the “love the one you’re with” strategy?
We know you can’t identify good teaching by looking at a student’s test scores. It could be that those kids were just smart when they started, as we often see in magnet schools. We’ve also seen why you can’t identify good teachers by using “value-added.” In fact, the latest research by Dan Goldhaber and his colleagues shows that value-added models can just about differentiate the very best teachers from the very worst, and that’s it. It can’t help you distinguish good teachers from average teachers.
Another recent paper, written by Marcus Winters, showed that value-added models are very volatile measures. If your policy is to fire the bottom 5% of teachers identified two years in a row with a value-added model, Winters discovered you’d get rid of only one teacher in every other county (school district) in Florida, which is not enough teachers to make a difference. So firing teachers is not the answer.
Another option is to raise the bar for entry into the profession. This would improve the quality of teachers, but it could take 30 to 40 years to have an impact on our students. So this is why we came up with the idea of “love the one you’re with.” This is why every single teacher needs to improve; not because they’re not good enough, but because they can be even better.
Next month, I’ll answer a few more questions, addressing TLC leadership, prioritization, the possibility of Dylan Wiliam Center demonstration schools in the near future, and a question that people often ask me about classroom formative assessment. Check your inbox at this time next month to continue this discussion with me.
Have a Question for Me?
I’m ready to answer it! Send me any questions you have about embedding formative assessment, college and career readiness, Common Core State Standards, or any other hot topic in education. In next month’s issue, I’ll offer my thoughts on the subjects that matter most to you.