4 ms·

The easy answer for the "Two Sigma" problem is that you... don't try to reproduce it on a large scale by finding some other method. You provide individual, one

The easy answer for the "Two Sigma" problem is that you... don't try to reproduce it on a large scale by finding some other method. You provide individual, one-on-one full time tutoring to each student, if that is the analytically best method. It's worth it, since educational spending is close to the most effective public spending imaginable - something in the $3-$5 return per dollar spent, which you'd be hard pressed to find any other available investment that can do that consistently. This also does away with some substantial percent of the overhead of maintaining separate public schools, assuming that this tutoring occurs in either the tutor's home or the child's home.

pliny 8y ago
>You provide individual, one-on-one full time tutoring to each student So, at a modest 2 hours of instruction per day, and assuming people spend on average 12 years in education and 48 years in the workforce, you need at minimum for 1/16th of your adults to be teachers. Not 1/16th to be employed in education but just teachers and just for this education scheme. Today it's about 1/40th (in the us) and that includes post-secondary and administration, logistics etc. so you're scaling up the education sector by 3x or 4x AND you need to find something to do with the kids for the other 5-6 hours when their parents are working. >It's worth it, since educational spending is close to the most effective public spending imaginable - something in the $3-$5 return per dollar spent Marginal ROI does not work that way. Even assuming you get that return on what you're spending now, that doesn't mean the marginal dollar invested in education is getting that return nor that you can scale up the amount you spend and continue getting that return.
- jaggederest 8y ago
  Closer to 1/3rd of the working population as teachers, actually, by my numbers. Current spending is $13k/yr per student, I'd bump that comfortably up 4x to $52k per year. That's about one full time teacher per student. As for marginal ROI, if the effects are as dramatic as the stated research implies, it'd be more than worth it. It would be a larger increase than the entire implementation of formal secondary and tertiary education system combined, which cost a lot more than K-8. I don't think you can overstate how large the demonstrated effect size was. Think "industrial revolution" or "invention of the printing press". Essentially you'd be spending an extra 3x the amount we currently spend to get... 5x? 10x? better outcomes. Results from those studies were equivalent to making the current best students in the country suddenly the very lowest tier of educational attainment - top 1% suddenly becomes the minimum standard. Think Rhodes scholars being "the new special ed kids", or the new "functionally illiterate must-pass graduates". (This is all assuming that the research was accurate, and the magnitude of difference is really that large. I'd want to see a lot of follow up.)
- ProblemFactory 8y ago
  > modest 2 hours of instruction per day Isn't this close to what students could/should be spending doing homework, reading or doing some projects together with their parents - effectively one-on-one tutoring for every single one of them? I suspect that US education spending is so high compared to other developed countries because it has to compensate for lack of social support elsewhere in society. The #1 factor in student success is parental involvement. Instead of buying iPads for low-income schools, what's needed is stable 35-40/hr week jobs that you can live on for low-income parents so that they can spend evenings with their family.
jimhefferon 8y ago
> individual, one-on-one full time tutoring to each student Isn't this the promise of the various platforms? I'm talking about Canvas, Moodle, WebWork-type programs. In principle, they can have everyone in the room working at their own level, and only moving up when the system says they are ready. Not that, as far as I know, those platforms are now being leveraged that way.
Joeri 8y ago
Isn’t one-on-one tutoring by an expert by over a quarter of the population just a return to the medieval craftsman & apprentice system?
- zbentley 8y ago
  Not at all. There are huge differences: trades/subjects of tutelage aren't foreordained or narrowly restricted; much more choice exists on the part of the student and teacher ('compulsory' education now is not the same as compulsory education then--the ply-your-trade-or-starve situation is not present int he same way); teachers can leverage much more effective tools (books! Simulations!) to convey concepts quicker than a 'watch me, then do as I do' apprenticeship; teachers do not typically occupy a simultaneous position of educator authority and parent/elder authority ... the list goes on.

pliny 8y ago

>You provide individual, one-on-one full time tutoring to each student So, at a modest 2 hours of instruction per day, and assuming people spend on average 12 years in education and 48 years in the workforce, you need at minimum for 1/16th of your adults to be teachers. Not 1/16th to be employed in education but just teachers and just for this education scheme. Today it's about 1/40th (in the us) and that includes post-secondary and administration, logistics etc. so you're scaling up the education sector by 3x or 4x AND you need to find something to do with the kids for the other 5-6 hours when their parents are working. >It's worth it, since educational spending is close to the most effective public spending imaginable - something in the $3-$5 return per dollar spent Marginal ROI does not work that way. Even assuming you get that return on what you're spending now, that doesn't mean the marginal dollar invested in education is getting that return nor that you can scale up the amount you spend and continue getting that return.

jaggederest 8y ago
Closer to 1/3rd of the working population as teachers, actually, by my numbers. Current spending is $13k/yr per student, I'd bump that comfortably up 4x to $52k per year. That's about one full time teacher per student. As for marginal ROI, if the effects are as dramatic as the stated research implies, it'd be more than worth it. It would be a larger increase than the entire implementation of formal secondary and tertiary education system combined, which cost a lot more than K-8. I don't think you can overstate how large the demonstrated effect size was. Think "industrial revolution" or "invention of the printing press". Essentially you'd be spending an extra 3x the amount we currently spend to get... 5x? 10x? better outcomes. Results from those studies were equivalent to making the current best students in the country suddenly the very lowest tier of educational attainment - top 1% suddenly becomes the minimum standard. Think Rhodes scholars being "the new special ed kids", or the new "functionally illiterate must-pass graduates". (This is all assuming that the research was accurate, and the magnitude of difference is really that large. I'd want to see a lot of follow up.)
ProblemFactory 8y ago
> modest 2 hours of instruction per day Isn't this close to what students could/should be spending doing homework, reading or doing some projects together with their parents - effectively one-on-one tutoring for every single one of them? I suspect that US education spending is so high compared to other developed countries because it has to compensate for lack of social support elsewhere in society. The #1 factor in student success is parental involvement. Instead of buying iPads for low-income schools, what's needed is stable 35-40/hr week jobs that you can live on for low-income parents so that they can spend evenings with their family.

jimhefferon 8y ago

> individual, one-on-one full time tutoring to each student Isn't this the promise of the various platforms? I'm talking about Canvas, Moodle, WebWork-type programs. In principle, they can have everyone in the room working at their own level, and only moving up when the system says they are ready. Not that, as far as I know, those platforms are now being leveraged that way.

Joeri 8y ago

Isn’t one-on-one tutoring by an expert by over a quarter of the population just a return to the medieval craftsman & apprentice system?

zbentley 8y ago
Not at all. There are huge differences: trades/subjects of tutelage aren't foreordained or narrowly restricted; much more choice exists on the part of the student and teacher ('compulsory' education now is not the same as compulsory education then--the ply-your-trade-or-starve situation is not present int he same way); teachers can leverage much more effective tools (books! Simulations!) to convey concepts quicker than a 'watch me, then do as I do' apprenticeship; teachers do not typically occupy a simultaneous position of educator authority and parent/elder authority ... the list goes on.