I have an accompanying new proposal: banning blogosphere substackers who have spent 0 minutes in public education systems from proposing ideas for optimizing teaching based on literature summaries.
Listen to teachers! *gasp* - we know more about our job than Freddie de Boer or Matt Yglesias (yes, I am sick of you both putting your foot into education debate whilst having spent zero time doing the actual work).
The problems we deal with are far more intersectional, relational and complex than you think. But we are also a diverse, canny and malleable bunch. I happen to think most teachers are also pretty good at their job and the direction of travel is in the main, pretty damn good.
Personally I would be a lot more open to listening to teachers (present company excepted) if teachers' organizations were not so obviously and cynically placing teachers' interests above those of students or the public.
I find it hard to convince community members of how important points one and especially two are. It seems harder the better educated they are in some other field: they understand the complexities in their own world and assume the same complexities in education.
I would push back a little on three. It's trivially true that almost any intervention (educational, social, medical, etc.) eventually fades with a large enough time and population horizon. As you say, "fade" doesn't mean "gone". Education policy makers need to be more clear about the horizons on which they are going to measure success and invest in research with similar timelines.
I get what you mean. Some reasons I find this a useful frame:
First, it's easy to dismiss education interventions by zooming out and pointing out that they fade. I wouldn't want us to dismiss 4th grade reading improvements because they have largely faded by 8th grade: that's an opportunity to focus on later grades, not a failure of the intervention.
Second, I think "all things fade" is an expression of selection effects: selection effects are persistent, and there's a "reversion to the mean" at play even after successes.
This also matters because schools should recognize that a student who is struggling and improves shouldn't just be ignored because they improved, they are at high risk of struggling in the future.
I would actually disagree with the Matthew Effect in this context. I think the Matthew Effect is mostly just selection. Improvement isn't necessarily self-compounding and needs to be actively maintained over time.
The problem with education research is that there is so much of it of such poor quality that you can find an "evidence-based" justification for just about anything. If you decided to use pure inquiry-based learning and threatened your students with an electric cattle prod when they don't make the appropriate "discovery", well, you could probably find a justification for it somewhere in the research literature.
A related problem in schools that your comment made me think of is that class culture and classroom management are some of the most important variables for learning, yet the hardest to replicate from one classroom to the next. I read once something about how teachers should more often look at data, and ask the teacher who has the best data what they are doing differently. I'm not opposed to that idea in principle, but in reality the most common answer would be "be a bit better at classroom management" and that is notoriously hard for another teacher to imitate from a distance.
I don't know what to make of that article, it sounds like in a lot of cases there's a lack of clarity around what exactly is being recommended and prioritized. It's a tough problem, they keep referring to low-performing districts without recognizing that's probably the result of selection effects. I work in a district like that, and the pressure to improve can lead to lots of short-sighted solutions. I think this all just reflects the fact that school improvement is hard.
My children go to a school with something of a bimodal distribution in achievement. To me, that seems like one of the hardest educational problems to solve, especially when applying solutions that are supposed to work for all students.
I see a lot of school leaders, especially new school leaders, start piling on initiatives in an effort to make a serious set of gains. Unfortunately, in many or maybe most of these situations, too many initiatives leads to none of them being done well, as well as teacher burnout and resentment.
Secondly, it's really difficult to get all of the different levels of education on the same page and rowing in the same direction.
I've seen situations where top down leadership has worked really well, and I've seen situations where it backfires badly. I've seen situations where grassroots movements of teachers working together have expanded and the bright spots have become bright collectives, but I've also seen situations where groups of teachers are succeeding with doing something so contrary to the rest of the people around them that they are unsupported and unable to maintain their island of success.
With this in mind, I think the report on California education is not surprising at all -- we can't often scale success across situations that are so wildly diverse. It's not just the selection effect, it's not just intervention effects, it's also the interactions of teachers and administrators and school systems and counselors and families and school boards and established policies and communities... And so on.
Good article.
I have an accompanying new proposal: banning blogosphere substackers who have spent 0 minutes in public education systems from proposing ideas for optimizing teaching based on literature summaries.
Listen to teachers! *gasp* - we know more about our job than Freddie de Boer or Matt Yglesias (yes, I am sick of you both putting your foot into education debate whilst having spent zero time doing the actual work).
The problems we deal with are far more intersectional, relational and complex than you think. But we are also a diverse, canny and malleable bunch. I happen to think most teachers are also pretty good at their job and the direction of travel is in the main, pretty damn good.
Personally I would be a lot more open to listening to teachers (present company excepted) if teachers' organizations were not so obviously and cynically placing teachers' interests above those of students or the public.
Love the optimism. Happy 5/12!
Came for the critical thinking. Stayed for the tomato monch reference.
https://www.youtube.com/watch?v=NIt1qgUflxo
I find it hard to convince community members of how important points one and especially two are. It seems harder the better educated they are in some other field: they understand the complexities in their own world and assume the same complexities in education.
I would push back a little on three. It's trivially true that almost any intervention (educational, social, medical, etc.) eventually fades with a large enough time and population horizon. As you say, "fade" doesn't mean "gone". Education policy makers need to be more clear about the horizons on which they are going to measure success and invest in research with similar timelines.
I get what you mean. Some reasons I find this a useful frame:
First, it's easy to dismiss education interventions by zooming out and pointing out that they fade. I wouldn't want us to dismiss 4th grade reading improvements because they have largely faded by 8th grade: that's an opportunity to focus on later grades, not a failure of the intervention.
Second, I think "all things fade" is an expression of selection effects: selection effects are persistent, and there's a "reversion to the mean" at play even after successes.
This also matters because schools should recognize that a student who is struggling and improves shouldn't just be ignored because they improved, they are at high risk of struggling in the future.
I would actually disagree with the Matthew Effect in this context. I think the Matthew Effect is mostly just selection. Improvement isn't necessarily self-compounding and needs to be actively maintained over time.
The problem with education research is that there is so much of it of such poor quality that you can find an "evidence-based" justification for just about anything. If you decided to use pure inquiry-based learning and threatened your students with an electric cattle prod when they don't make the appropriate "discovery", well, you could probably find a justification for it somewhere in the research literature.
A related problem in schools that your comment made me think of is that class culture and classroom management are some of the most important variables for learning, yet the hardest to replicate from one classroom to the next. I read once something about how teachers should more often look at data, and ask the teacher who has the best data what they are doing differently. I'm not opposed to that idea in principle, but in reality the most common answer would be "be a bit better at classroom management" and that is notoriously hard for another teacher to imitate from a distance.
You make a lot of sense. What do you think of the recent Stanford study on California’s education system? It seems their conclusion is too many interventions, not enough bandwidth/coherence. https://edsource.org/2026/stanford-research-california-education/757513
I don't know what to make of that article, it sounds like in a lot of cases there's a lack of clarity around what exactly is being recommended and prioritized. It's a tough problem, they keep referring to low-performing districts without recognizing that's probably the result of selection effects. I work in a district like that, and the pressure to improve can lead to lots of short-sighted solutions. I think this all just reflects the fact that school improvement is hard.
My children go to a school with something of a bimodal distribution in achievement. To me, that seems like one of the hardest educational problems to solve, especially when applying solutions that are supposed to work for all students.
I would think a bimodal distribution would be well suited to tracking?
I would think so too...unfortunately the school does not, and continues to try to 'solve' a much harder problem.
I see a lot of school leaders, especially new school leaders, start piling on initiatives in an effort to make a serious set of gains. Unfortunately, in many or maybe most of these situations, too many initiatives leads to none of them being done well, as well as teacher burnout and resentment.
Secondly, it's really difficult to get all of the different levels of education on the same page and rowing in the same direction.
I've seen situations where top down leadership has worked really well, and I've seen situations where it backfires badly. I've seen situations where grassroots movements of teachers working together have expanded and the bright spots have become bright collectives, but I've also seen situations where groups of teachers are succeeding with doing something so contrary to the rest of the people around them that they are unsupported and unable to maintain their island of success.
With this in mind, I think the report on California education is not surprising at all -- we can't often scale success across situations that are so wildly diverse. It's not just the selection effect, it's not just intervention effects, it's also the interactions of teachers and administrators and school systems and counselors and families and school boards and established policies and communities... And so on.
I tend to agree. Then you have intra-school diversity, as I mentioned above...non-trivial systems indeed.