Rob Henderson on kids these days; Carolyn D. Gorman on mental health awareness; Adam Mastroianni on cybernetic psychology; Scott Alexander on smart phones and kids
"When I talked more with Bryan, he recommended hiring more nannies."
I thought Bryan had something interesting to say, but reading the SSC piece and Scott's comments I guess it all boils down to "be rich enough to afford nannies."
That's not terrible advice as public policy (money would work!) but doesn't help anyone that can't afford a bunch of nannies (you know most people in their prime fertility window).
As to free range parenting the problem is that nobody else is doing it. And if you do find one that does it they often just let them stare at screens all day because its less "free range" and more "I don't give a shit."
These days, if you give a kid actually free range, they will explore the range between YouTube, Disney+, and Minecraft. If you take away some of the "free" and try to snatch the screens away and force them to go to outside somewhere, they will go to the nearest house where they can watch the screens with some kids whose parents are OK with screens. If you restrict that too, then there is actually nowhere left for the kids to go where there are other kids in the same circumstances, best they can do is go to the playground with every other kids watched by parents or nannies. It's not that every other space is dangerous (though they will encounter dozens of adults who will imagine so, and who will then unhesitatingly and immediately call 911 and CPS if just seeing that a kid is unsupervised) but that every other space is either commercial, unaccompanied-kid-unfriendly, and too expensive for kid amounts of cash (if they even accept cash, which they increasingly don't) and, more importantly, totally boring.
If you actually want your kids to get some exercise and activity outside of the school context then you have to take over that part of their life (the very opposite of "free range") and give up the time investment of that amount of your own life in order to tackle the immense logistical involved in coordination.
This isn't just a kid problem, even adults who live in the same metro area often complain of isolation and atomization and that actually getting the boys or girls together for literally anything is a herding-cats and schedule-deconfliction nightmare unless you are treating some socializing """leisure""" activity so seriously and with necessary commitment to such a routine schedule it's practically a part-time job in itself.
The trade-off of more independent Individualism is the eradication of the very possibility of easy and spontaneous socialization opportunities. "People don't want to think it be like this, but it do."
One of the big things I did was I found a GPS tracker they could wear even in the water and didn't have a screen (their name and my phone was also written on it). This goes directly to my phone so I can see where they are wherever they go in the world (even if I'm far away). We use it if we take them to the water park or beach, that way they can roam without my having to worry about letting them run off in a big place with a large crowd.
The Scott Alexander article was actually pretty good in that he challenged the accuracy of the data multiple times. Nullius in verba.
The declining literacy issue is a long term trend, but the effect of the phones is perhaps overstated. It is less of a client device issue and more of a server issue. Millennials were made more stupid by cable TV and the early internet, which connected them with unending pap. But even the TV for children was usually more limited; you could only watch so many hours of trash before you ran out of it.
Now anyone can access more personalized trash than they can possibly consume, and it's like smoking crack all day. Debating about whether taking crack through a spoon, through a pipe, or ingested in some kind of slurry is mostly marginal. The bulk of the medium "lives" on servers and the fact that it is delivered to phones is only marginally important.
I think there’s something about the accessibility of phones that have a multiplying effect. Yes, you can be addicted to the TikTok in your house. But with phones, you can access the addicting content at all times - and it’s socially acceptable to do so outside the home. Society doesn’t think it’s weird for you to take a hit off your crack pipe when mid conversation with someone in public.
Cable TV might have been dumb, but it wasn’t addicting and you couldn’t turn it on when having a conversation with someone on the street.
Certainly it'll make up a larger part of the margin for children because their activities are so monitored and controlled compared to adults.
I think most people understood during the '90s and '00s that e.g. daytime TV made a lot of housewives and old people duller and more vapid than they would have otherwise been. But now the same person who would have been limited to 1 hour of Maury per day and 1 hour of Springer can access Maury equivalents 12-16 hours a day through multiple devices.
As a thought experiment, if you took the hammer to the server side, it'd have much stronger effects on the desirability of the client side devices. If you limited access to the client devices, but did nothing on the server side, it might not have that big of an impact because it'd be easier for the people to just use a different client to interact with the medium.
In the realm of school policies on devices, restrictions on phones might not have the impact you want if the children have to use laptops or Chromebooks throughout the day.
All reasonable concerns, which I'd be happy to respond to in a post after this one.
Big picture, you've put your finger on the conundrum of K-12 Education: How have National Assessment of Educational Progress (NAEP) Long Term Trend (LTT) scores remained flat or slightly down (before the lock-down dip) for decades while inflation-adjusted spending per pupil has risen 2-6x over the same decades?
Importantly, based on Programme for International Student Assessment (PISA) scores, other countries deliver higher achievement for less money, with Finland and Singapore famously setting the pace and Canada beating the US easily over the years, so better is possible.
It seems the choices are: 1) the NAEP LTT test(s) are deeply broken (i.e., the 100+ year old profession of psychometrics has no idea what it is doing) or the connection between input and output in K-12 Ed is 2) non-existent/broken or 3) blocked (e.g., by the teachers unions). I'd welcome any other hypotheses, as I'm working on addressing number two.
To bring it back to Hattie (2009), one factor is the public's mistaken belief that smaller class sizes lead to better outcomes. Yes, there is an effect size (d = 0.21) but it is small and below the average intervention, because within the three ranges (small group, classroom, and large lecture) student and teacher work practices don't change much I believe.
To the extent television is a proxy for smart phones, in his syntheses of meta-analyses Hattie (2009) reports the effect size of TV as a Cohen's d = -0.18 and as d = -0.15 in Hattie (2023). Thus, lots of TV reduces achievement. This contrasts with the average intervention effect size of d = 0.40, which is roughly equal to a grade level.
I highly recommend either of these books as containing everything you ever wanted to know about Education.
Thanks for the citation. Perhaps I am just old and cynical but reading through the amazon reviews left me "underwhelmed." Feedback is important? Who knew? The big problem is to get the student to care about the feedback.
One of the reviewers said something that really stayed with me. Most of the interventions Hattie reports have a positive effect. You say in your first comment, a Cohen's d of 0.40. That means the intervention increased student achievement by 0.40 standard deviations. Put a few of those together and you've moved the mean achievement to where the top few percent used to be. But we don't observe that in real life. What we have observed is almost no movement at all.
The books are meta-analyses of meta-analyses. The lack of real world effect makes me wonder if these books, though a tremendous intellectual achievement, are "garbage in/garbage out". The original studies are, perhaps, not very useful. (I'm trying to avoid vulgar language.)
All reasonable concerns, which I'd be happy to respond to in a post after this one.
Big picture, you've put your finger on the conundrum of K-12 Education: How have National Assessment of Educational Progress (NAEP) Long Term Trend (LTT) scores remained flat or slightly down (before the lock-down dip) for decades while inflation-adjusted spending per pupil has risen 2-6x over the same decades?
Importantly, based on Programme for International Student Assessment (PISA) scores, other countries deliver higher achievement for less money, with Finland and Singapore famously setting the pace and Canada beating the US easily over the years, so better is possible.
It seems the choices are: 1) the NAEP LTT test(s) are deeply broken (i.e., the 100+ year old profession of psychometrics has no idea what it is doing) or the connection between input and output in K-12 Ed is 2) non-existent/broken or 3) blocked (e.g., by the teachers unions). I'd welcome any other hypotheses, as I'm working on addressing number two.
To bring it back to Hattie (2009), one factor is the public's mistaken belief that smaller class sizes lead to better outcomes. Yes, there is an effect size (d = 0.21) but it is small and below the average intervention, because within the three ranges (small group, classroom, and large lecture) student and teacher work practices don't change much I believe.
Slightly more technical points as promised, which Hattie (2009) addresses explicitly for the most part:
1) To your point, Hattie (2009) claims natural human development without formal intervention yields a d = 0.15 as best we can estimate, so in one sense everything works, which makes the negative effect of TV even more troubling.
2) Since nearly everything works, also to your point, it's a matter of trade-offs since some interventions cannot be executed simultaneously (e.g., inquiry-based learning and Direct Instruction are incompatible)
3) Thus, we should optimize the combinations of interventions with high effect sizes, perhaps using something like effect size per minute of intervention time.
A) Feedback is indeed important. Unfortunately, the feedback students actually get is rare and not especially helpful. Grade inflation has a lot to answer for here. In a higher Ed context, one professor of mine didn't believe in grades, so we got no feedback all semester and just had to assume she was going to hand out As automatically. This is an abdication of responsibility on her part, since constructive feedback has little to do with grades.
B) My understanding of meta-analyses is they effectively add to the sample size to enhance the ability to detect the signal. Yes, most studies aren't great but the presence of consistent differences suggests more than just noise. And decisions are getting made daily based on what one of my professors call "the folklore of teaching."
C) There are good studies at scale, especially in Tennessee, and Hattie lists his sources, so one can select from the large scale, serious studies if one prefers.
Finally, I agree that many of the interventions can be combined to generate a very big lift in achievement. Most of the people on this blog could have gotten through K-12 many years sooner if they'd been encouraged and allowed to. This would free up resources to be used for those students most in need. See accelerationinstitute.org for counters to the usual objections to accelerating.
Sorry to take so long to reply. I was offline all holiday weekend.
One of the basic beliefs of people who run modern education is that age-peers should not be broken up. None of us would be encouraged or allowed to be out sooner (unless a parent put up a fuss).
The folklore of teaching may often be more realistic than what gets published in education journals. In fact, I'd suggest that much that is actually correct in education research is present in classrooms. A lot gets taught in ed school and professional development and then teachers keep what they find is practical and doable. And it doesn't make much difference because--heresy! heresy!--American schools are doing about as well as they reasonably can.
Speaking of heresy, American PISA scores look disappointing when you compare on a country-wide basis, but Americans of European descent generally do better than Europeans in European countries. American of Asian descent generally do better than Asians in Asian countries. Americans who the government define as Hispanic generally do better than people in Latin American countries. And American blacks do better than the few Africans who participate in PISA. A colorful representation of how this looks for the 2018 PISA is:
I disagree strongly with Mr. Mastroianni. I think the mind, which is to say human consciousness is an emergent system that cannot be fully understood by just mechanically assembling its component parts. It may not be fully comprehensible to humans at all. It is one thing to say that we behave "as if" we have an array of control systems located in various places in our brain. It's entirely different to conclude that the brain actually functions this way
I have not yet read his article. Thank you for providing the link I am very interested in this stuff.
Mastroianni is reviewing another book, which has the paradigm of the mind being a series of cybernetic control systems, a theory he strongly suspects is wrong.
But he thinks it’s closer to science, and more falsifiable, than most personality stuff that is either naive or impressionistic.
If you really disagree, you should read it and suggest how to show what consciousness is.
I think the brain acts like the controls about 70% of the time. Proof or false?
"When I talked more with Bryan, he recommended hiring more nannies."
I thought Bryan had something interesting to say, but reading the SSC piece and Scott's comments I guess it all boils down to "be rich enough to afford nannies."
That's not terrible advice as public policy (money would work!) but doesn't help anyone that can't afford a bunch of nannies (you know most people in their prime fertility window).
As to free range parenting the problem is that nobody else is doing it. And if you do find one that does it they often just let them stare at screens all day because its less "free range" and more "I don't give a shit."
These days, if you give a kid actually free range, they will explore the range between YouTube, Disney+, and Minecraft. If you take away some of the "free" and try to snatch the screens away and force them to go to outside somewhere, they will go to the nearest house where they can watch the screens with some kids whose parents are OK with screens. If you restrict that too, then there is actually nowhere left for the kids to go where there are other kids in the same circumstances, best they can do is go to the playground with every other kids watched by parents or nannies. It's not that every other space is dangerous (though they will encounter dozens of adults who will imagine so, and who will then unhesitatingly and immediately call 911 and CPS if just seeing that a kid is unsupervised) but that every other space is either commercial, unaccompanied-kid-unfriendly, and too expensive for kid amounts of cash (if they even accept cash, which they increasingly don't) and, more importantly, totally boring.
If you actually want your kids to get some exercise and activity outside of the school context then you have to take over that part of their life (the very opposite of "free range") and give up the time investment of that amount of your own life in order to tackle the immense logistical involved in coordination.
This isn't just a kid problem, even adults who live in the same metro area often complain of isolation and atomization and that actually getting the boys or girls together for literally anything is a herding-cats and schedule-deconfliction nightmare unless you are treating some socializing """leisure""" activity so seriously and with necessary commitment to such a routine schedule it's practically a part-time job in itself.
The trade-off of more independent Individualism is the eradication of the very possibility of easy and spontaneous socialization opportunities. "People don't want to think it be like this, but it do."
One of the big things I did was I found a GPS tracker they could wear even in the water and didn't have a screen (their name and my phone was also written on it). This goes directly to my phone so I can see where they are wherever they go in the world (even if I'm far away). We use it if we take them to the water park or beach, that way they can roam without my having to worry about letting them run off in a big place with a large crowd.
The Scott Alexander article was actually pretty good in that he challenged the accuracy of the data multiple times. Nullius in verba.
The declining literacy issue is a long term trend, but the effect of the phones is perhaps overstated. It is less of a client device issue and more of a server issue. Millennials were made more stupid by cable TV and the early internet, which connected them with unending pap. But even the TV for children was usually more limited; you could only watch so many hours of trash before you ran out of it.
Now anyone can access more personalized trash than they can possibly consume, and it's like smoking crack all day. Debating about whether taking crack through a spoon, through a pipe, or ingested in some kind of slurry is mostly marginal. The bulk of the medium "lives" on servers and the fact that it is delivered to phones is only marginally important.
I think there’s something about the accessibility of phones that have a multiplying effect. Yes, you can be addicted to the TikTok in your house. But with phones, you can access the addicting content at all times - and it’s socially acceptable to do so outside the home. Society doesn’t think it’s weird for you to take a hit off your crack pipe when mid conversation with someone in public.
Cable TV might have been dumb, but it wasn’t addicting and you couldn’t turn it on when having a conversation with someone on the street.
Certainly it'll make up a larger part of the margin for children because their activities are so monitored and controlled compared to adults.
I think most people understood during the '90s and '00s that e.g. daytime TV made a lot of housewives and old people duller and more vapid than they would have otherwise been. But now the same person who would have been limited to 1 hour of Maury per day and 1 hour of Springer can access Maury equivalents 12-16 hours a day through multiple devices.
As a thought experiment, if you took the hammer to the server side, it'd have much stronger effects on the desirability of the client side devices. If you limited access to the client devices, but did nothing on the server side, it might not have that big of an impact because it'd be easier for the people to just use a different client to interact with the medium.
In the realm of school policies on devices, restrictions on phones might not have the impact you want if the children have to use laptops or Chromebooks throughout the day.
All reasonable concerns, which I'd be happy to respond to in a post after this one.
Big picture, you've put your finger on the conundrum of K-12 Education: How have National Assessment of Educational Progress (NAEP) Long Term Trend (LTT) scores remained flat or slightly down (before the lock-down dip) for decades while inflation-adjusted spending per pupil has risen 2-6x over the same decades?
Importantly, based on Programme for International Student Assessment (PISA) scores, other countries deliver higher achievement for less money, with Finland and Singapore famously setting the pace and Canada beating the US easily over the years, so better is possible.
It seems the choices are: 1) the NAEP LTT test(s) are deeply broken (i.e., the 100+ year old profession of psychometrics has no idea what it is doing) or the connection between input and output in K-12 Ed is 2) non-existent/broken or 3) blocked (e.g., by the teachers unions). I'd welcome any other hypotheses, as I'm working on addressing number two.
To bring it back to Hattie (2009), one factor is the public's mistaken belief that smaller class sizes lead to better outcomes. Yes, there is an effect size (d = 0.21) but it is small and below the average intervention, because within the three ranges (small group, classroom, and large lecture) student and teacher work practices don't change much I believe.
To the extent television is a proxy for smart phones, in his syntheses of meta-analyses Hattie (2009) reports the effect size of TV as a Cohen's d = -0.18 and as d = -0.15 in Hattie (2023). Thus, lots of TV reduces achievement. This contrasts with the average intervention effect size of d = 0.40, which is roughly equal to a grade level.
I highly recommend either of these books as containing everything you ever wanted to know about Education.
I don't see any books in your comment.
Sorry, I was trying to save space.
Hattie (2009) Visible Learning: A synthesis of over 800 meta-analyses relating to achievement.
Hattie (2023) Visible Learning: The sequel: A synthesis of over 2,100 meta-analyses relating to achievement.
Note that Hattie uses "Visible Learning" as a sort of brand in many of the titles of his books, so double-check the titles before ordering.
Thanks for the citation. Perhaps I am just old and cynical but reading through the amazon reviews left me "underwhelmed." Feedback is important? Who knew? The big problem is to get the student to care about the feedback.
One of the reviewers said something that really stayed with me. Most of the interventions Hattie reports have a positive effect. You say in your first comment, a Cohen's d of 0.40. That means the intervention increased student achievement by 0.40 standard deviations. Put a few of those together and you've moved the mean achievement to where the top few percent used to be. But we don't observe that in real life. What we have observed is almost no movement at all.
The books are meta-analyses of meta-analyses. The lack of real world effect makes me wonder if these books, though a tremendous intellectual achievement, are "garbage in/garbage out". The original studies are, perhaps, not very useful. (I'm trying to avoid vulgar language.)
All reasonable concerns, which I'd be happy to respond to in a post after this one.
Big picture, you've put your finger on the conundrum of K-12 Education: How have National Assessment of Educational Progress (NAEP) Long Term Trend (LTT) scores remained flat or slightly down (before the lock-down dip) for decades while inflation-adjusted spending per pupil has risen 2-6x over the same decades?
Importantly, based on Programme for International Student Assessment (PISA) scores, other countries deliver higher achievement for less money, with Finland and Singapore famously setting the pace and Canada beating the US easily over the years, so better is possible.
It seems the choices are: 1) the NAEP LTT test(s) are deeply broken (i.e., the 100+ year old profession of psychometrics has no idea what it is doing) or the connection between input and output in K-12 Ed is 2) non-existent/broken or 3) blocked (e.g., by the teachers unions). I'd welcome any other hypotheses, as I'm working on addressing number two.
To bring it back to Hattie (2009), one factor is the public's mistaken belief that smaller class sizes lead to better outcomes. Yes, there is an effect size (d = 0.21) but it is small and below the average intervention, because within the three ranges (small group, classroom, and large lecture) student and teacher work practices don't change much I believe.
Slightly more technical points as promised, which Hattie (2009) addresses explicitly for the most part:
1) To your point, Hattie (2009) claims natural human development without formal intervention yields a d = 0.15 as best we can estimate, so in one sense everything works, which makes the negative effect of TV even more troubling.
2) Since nearly everything works, also to your point, it's a matter of trade-offs since some interventions cannot be executed simultaneously (e.g., inquiry-based learning and Direct Instruction are incompatible)
3) Thus, we should optimize the combinations of interventions with high effect sizes, perhaps using something like effect size per minute of intervention time.
A) Feedback is indeed important. Unfortunately, the feedback students actually get is rare and not especially helpful. Grade inflation has a lot to answer for here. In a higher Ed context, one professor of mine didn't believe in grades, so we got no feedback all semester and just had to assume she was going to hand out As automatically. This is an abdication of responsibility on her part, since constructive feedback has little to do with grades.
B) My understanding of meta-analyses is they effectively add to the sample size to enhance the ability to detect the signal. Yes, most studies aren't great but the presence of consistent differences suggests more than just noise. And decisions are getting made daily based on what one of my professors call "the folklore of teaching."
C) There are good studies at scale, especially in Tennessee, and Hattie lists his sources, so one can select from the large scale, serious studies if one prefers.
Finally, I agree that many of the interventions can be combined to generate a very big lift in achievement. Most of the people on this blog could have gotten through K-12 many years sooner if they'd been encouraged and allowed to. This would free up resources to be used for those students most in need. See accelerationinstitute.org for counters to the usual objections to accelerating.
Sorry to take so long to reply. I was offline all holiday weekend.
One of the basic beliefs of people who run modern education is that age-peers should not be broken up. None of us would be encouraged or allowed to be out sooner (unless a parent put up a fuss).
The folklore of teaching may often be more realistic than what gets published in education journals. In fact, I'd suggest that much that is actually correct in education research is present in classrooms. A lot gets taught in ed school and professional development and then teachers keep what they find is practical and doable. And it doesn't make much difference because--heresy! heresy!--American schools are doing about as well as they reasonably can.
Speaking of heresy, American PISA scores look disappointing when you compare on a country-wide basis, but Americans of European descent generally do better than Europeans in European countries. American of Asian descent generally do better than Asians in Asian countries. Americans who the government define as Hispanic generally do better than people in Latin American countries. And American blacks do better than the few Africans who participate in PISA. A colorful representation of how this looks for the 2018 PISA is:
https://www.unz.com/isteve/the-new-2018-pisa-school-test-scores-usa-usa/
I disagree strongly with Mr. Mastroianni. I think the mind, which is to say human consciousness is an emergent system that cannot be fully understood by just mechanically assembling its component parts. It may not be fully comprehensible to humans at all. It is one thing to say that we behave "as if" we have an array of control systems located in various places in our brain. It's entirely different to conclude that the brain actually functions this way
I have not yet read his article. Thank you for providing the link I am very interested in this stuff.
Mastroianni is reviewing another book, which has the paradigm of the mind being a series of cybernetic control systems, a theory he strongly suspects is wrong.
But he thinks it’s closer to science, and more falsifiable, than most personality stuff that is either naive or impressionistic.
If you really disagree, you should read it and suggest how to show what consciousness is.
I think the brain acts like the controls about 70% of the time. Proof or false?