13 ms·

AltSchool, an Experiment in Education, Raises $100M

87 points by ghosh 12y ago

AndrewKemendo 12y ago
Almost every education article posted to HN that boasts a "new approach" is invariably a school that re-purposing Montessori. In this case it's using technology and data to a higher degree to streamline the BS teachers have to deal with so they can get on with teaching. That's cool and definitely a step above 99% of schools out there, but not a revolution in the Montessori sense. I wish more people would grok the basic principles of Montessori and just focus on doing those right - because when it's done well its perfect and really doesn't take much structure. My wife runs a Montessori school from our home and has been a teacher for almost 12 years now. She has found that the more hands on she is (and other teachers are) with the students, the worse the outcomes are. As practitioners we see a lot of junk that claims to be "Montessori" but doesn't really stick to the principles. These are primarily: self directed learning and building a community for exploration and creativity. The best Montessori schools have agreements with local warehouses/mills/shops etc... where children can do internships and apprenticeships. Having "curricula" like the article states is generally antithetical to that and kills a large portion of what makes Montessori so successful - which is pupils taking prepared "vocational" paths which are liberal in educational scope. It's simpler in the long run but takes much more up-front work and I think the technology is there to make that up-front work much easier. The real challenge is fitting these types of schools into the larger "schooling" world of standardized tests etc... One thing we found was that many students leave Montessori because the clearest path into college is through the testing system and Monetssori doesn't play well with that. That is what needs real reform.
- hullo 12y ago
  If this were still May Day I would think that sending our seven year olds off to the mill would be a more obvious troll, but as things stand I'm not so sure.
  - brlewis 12y ago
    The Montessori school my oldest went to did week-long apprenticeshps in 7th and 8th grade. I'd be surprised if any school had 2nd grade apprenticeships.
  - saraid216 12y ago
    They're not expected to go to the mill to help support their family.
  - notahacker 12y ago
    I recall that after two weeks of "work experience" on a print-shop production line as a teenager I couldn't wait to get back to the relative blissful liberation of an organised lesson programme and essay tasks based on the exam curriculum...
  - AndrewKemendo 12y ago
    Apprenticeship =/ child labor.
- ironchef 12y ago
  Montessori isn't the end-all be-all of "student centered learning". There are a LOT more types of schooling out there which have similarity as they seem to have roots in Piaget, Steiner, etc. This seems to be another take on montessori + some reggio emilia and other bits. Also, as an fyi, one of the better montessoris in our area (montessori academy of chicago) has taken a big bite of the standardization area by aligning common core with the classic montessori didactic system.
  - AndrewKemendo 12y ago
    >Montessori isn't the end-all be-all of "student centered learning" I agree in theory, but have yet to see another system as flexible or robust with any actual, measurable history of good outcomes at scale. All other systems seem to be on either extreme - either too much "freedom" and the kids never learn, like Summerhill, or are too focused around some oddly specific dogma, like Waldorf.
    - ironchef 11y ago
      I didn't mean to suggest that Montessori wasn't good or that the other systems were better. All I was suggesting was that there are a multitude of student centered learning frameworks (some of which are montessori, waldorf, reggio, and summerhill). I think based on the concepts of differentiation, some of these may be more optimal for certain learners than others.
- germinalphrase 12y ago
  Hopefully we will see the influence of standardized testing companies diminish as more robust in-class assessment tools are developed. Our trust in standardized testing is dependent on the expectation that teachers cannot accurately assess the skills and competencies of their students during regular lessons. To free schools to approach education however they want we just need to find a way to reliably assess students without the need for a centralized gatekeeper. If the process of developing/proctoring/critiquing assessments was partially standardized (to allow for distribution and certification of assessment elements) and linked to student performance data then we could both reliably integrate this kind of assessment into regular lessons while also mitigating the need for expensive standardized tests. There are some hints of this in the article; as a HS teacher I'd love to see assessment overhauled in this direction.
  - aidenn0 12y ago
    Standardized curricula (which is admittedly distinct from standardized testing) is useful as prior to statewide standards, it was not unusual for a move of just 10 miles to put your children in a school district with completely different curriculum for at least one subject. The most positive thing I've heard about Common Core is that a federal standard will mean out-of-state transfer students will be more likely to be on the same page as current students.
    - brlewis 12y ago
      Alternatively, take an educational approach where it's not essential that all students be on the same page.
hmate9 12y ago
Glad to see technology used to enhance education. I like their approach and hope they succeed (with Peter Thiel on their side, it should almost be a guarantee).
diminoten 12y ago
If a 5th grader "decides" to learn some algebra, those are the cards the teacher will start preparing for that student, right? But to "learn" algebra, one traditionally learns a set of concepts and practices those concepts in the form of large quantities of practice problems. Every math teacher I've ever met has said the key to mastery of math concepts is practice and repetition. So, if we go back to our 5th grader, and algebra comes up for him/her in the customized lesson plan, does that mean for the next year, algebra cards are going to appear, with a set of problems to solve, and links to... what, wikipedia pages that explain how to solve them? What happens when the kid decides he/she doesn't want to "explore" math anymore? Does that kid just not get a math education, at that time? Are we going to have 16 yr olds who never figured out that "exploring math" is going to be a key to being a competent adult? I'm sure what I wrote is quite commonly brought up and an answer exists, but I do wonder what that answer is -- to sum: what do you do when a kid doesn't want to "explore" an area of knowledge that we as adults know is going to be critical to functioning as one of us in the world?
- Marcus316 12y ago
  Not sure where you got the idea that it is entirely the child's choice what he learns. What I got out of the article is that the child's interests, strengths and weaknesses are taken into account to produce an individualized curriculum. If a child is weak in mathematics, the curriculum is adjusted for _that child_ to emphasize mathematics. It's not as simple as "Johnny wants to learn poetry and only poetry, so that's what we let him learn." At least, that's what I got out of the explanation. It's chaotic in that every student needs to be guided differently, and AltSchool is using technology to make that guidance easier for the teachers to apply across individuals. The idea is not to lose students in the crowd, but also not to let them lose themselves in the chaos.
anonbanker 12y ago
This school looks like a terrifying extension of Snow Crash's "living-in-a-storage-container" lifestyle, combined with training on how to be a perfect worker drone for the new tech/service-based economy (they're even teaching the kids to clock in and out!). If this is where private/public school is headed, I'll stick to homeschooling/unschooling my kids, thanks.
- ryoshu 12y ago
  Seemed a bit more Diamond Age to me.
glup 12y ago
A surveillance-state Montessori with a big data fetish and a bottom line doesn’t count as reform, sorry AltSchool. I was particularly struck by: > “It’s not hard to model language acquisition if you can listen to every word a person is saying,” This is patently false big data hubris, which I’m afraid bodes poorly for the rest of the rest of their approach. I study early language acquisition and have worked on some very large datasets (e.g. Speechome, http://www.ted.com/talks/deb_roy_the_birth_of_a_word http://www.ted.com/talks/deb_roy_the_birth_of_a_word), and there’s a lot you don’t know about the language system of a child even with a complete record of utterances and videos of everything that happens. Higher density/quality data is fantastic, but to understand language acquisition or to understand the progress of a single child you have to work out a complex system of latent variables. Education is similar— the biggest thing is that there is s different mixture of explicit pedagogy vs. observation. I’m all for school reform that leverages technology but this misunderstanding of data, in conjunction with the for-profit aspect, leaves me pretty convinced that this isn’t the right direction. The other thing that caught me off guard was: > “and wearable devices that help keep track of students off campus” I did a double-take. Why does this school track the kids off campus? The complete lack of criticism eliminated the last shred of credibility Wired had in my mind. I don’t care if you think this is a good idea or not, tell me who doesn’t think it’s a good idea and why.
- pbreit 12y ago
  > Why does this school track the kids off campus? So they don't get lost on field trips?
Aloisius 12y ago
I'm all for doing whatever we can to improve education, but it is extremely difficult to compare the quality of a education between public and private schools because of selection bias. Private schools tend to pick and choose the best kids and have significantly higher suspension and expulsion rates eliminating the "problem" children. Public schools on the other hand, have to accept everyone from the hyperactive to those with learning disabilities. When we try to control for these effects, often there is little to no difference between these custom approaches and standard ones. Montessori for instance does customization and outcomes are effectively the same as public schools. I suppose though even if the end result is a cheaper and highly replicable education system, it could become quite successful as a company.
- pbreit 12y ago
  Isn't the main point of private schools to pay for the selection bias?
- brlewis 12y ago
  Please cite a source for "Montessori...outcomes are effectively the same as public schools". Here's a study that found the opposite: http://www.scientificamerican.com/article/students-prosper-with-mon/ http://www.scientificamerican.com/article/students-prosper-w...
  - Retric 12y ago
    At the study level sure, you can support any argument, a meta analysis like: http://www.irp.wisc.edu/publications/dps/pdfs/dp114197.pdf http://www.irp.wisc.edu/publications/dps/pdfs/dp114197.pdf Generally shows little to no benefit and often points out flaws in individual studies. Indeed, in our data, previous authors’ specifications rarely explain a substantial portion of the selection into the relevant privateschool sector, raising the possibility that the weak-instrument critique may hold for this literature. This point is corroborated in work by Ludwig (1997), who also finds compelling evidence that prior authors’ instruments are weak.
    - 12y ago
coldcode 12y ago
Everything can't be improved by adding Silicon Valley to it.
- michaeledwards 12y ago
  But everything CAN be improved by only adding silicon to it!
  - crpatino 12y ago
    Or silicone...
danans 12y ago
There are several distinct problems in the public education system. There is the problem where motivated and largely well-off parents find that the public education system doesn't suit the unique interests and learning styles of their children. This sort of school seems like a wonderful experiment to address this sort of problem. A second, largely unrelated problem is the under-funding (in CA for sure, perhaps elsewhere) of public schools, given the expectation that these school systems rectify the serious motivational and educational disabilities students develop in communities with high concentrations of poverty and violence. I'd argue that the latter problem is the one that has serious societal consequences if we don't get it fixed, sooner than later.
- Domenic_S 12y ago
  > A second, largely unrelated problem is the under-funding (in CA for sure, perhaps elsewhere) of public schools I hear this often. I'm curious, do you have a dollar figure in mind that would be appropriate?
  - aaroninsf 12y ago
    CA spends less than half as much per student as New York: http://www.usnews.com/news/blogs/data-mine/2015/01/29/how-states-are-spending-money-in-education http://www.usnews.com/news/blogs/data-mine/2015/01/29/how-st... What the 'right' value is subject to debate, but as a parent in SF having negotiated the tragicomic SFUSD system for my own children, I'd wager it's a lot closer to New York's 19K per student than CA's $9K. This is pure speculation, but my subjective experience is that CA funds schools at close to the minimum possible level, and that there is a lot of headroom in which every dollar extra spend causes an immediate improvement in service and richness of experience. I say this because one of the differentiating factors of the most-desired schools in the SFUSD is the large amount of extra money parents bring in via fundraising. Yet even the most successful schools, which raise on the order of half a million dollars, are only adding something like ~$1K per student. Yet what they offer above and beyond comparable schools without that funding bump is very evident: all the 'enriching' staples my own generation took for granted as par of public education: art classes, PE, green space, staff with expertise in supporting kids with differing needs (at both ends of the performance/ability spectrum)... etc etc etc. I have also observed that affluent parents who privilege education also both prioritize and have the luxury of donating their time -- no small thing in a city in which a single-income household is mostly unthinkable. So that extra $1K may be multiplied by donated time, as well. But any way you slice it, a few extra K per student, judiciously spent, would make a hell of a difference in my parts. That said, a dollar raised and allocated by the parents, in accord with parents' priorities, and subject to their audit, is probably much better spent than one squeezed through the SFUSD. QED non-profit charters (of which there are several) are currently a very strong bet in SFUSD...
pbreit 12y ago
Is there any evidence that private schools make much of a difference in the overall education landscape or is it just yet another option for rich people? If you really want to make a difference you need to address the broader audience and do something like charter schools.

87 points by ghosh 12y ago

Almost every education article posted to HN that boasts a "new approach" is invariably a school that re-purposing Montessori. In this case it's using technology and data to a higher degree to streamline the BS teachers have to deal with so they can get on with teaching. That's cool and definitely a step above 99% of schools out there, but not a revolution in the Montessori sense. I wish more people would grok the basic principles of Montessori and just focus on doing those right - because when it's done well its perfect and really doesn't take much structure. My wife runs a Montessori school from our home and has been a teacher for almost 12 years now. She has found that the more hands on she is (and other teachers are) with the students, the worse the outcomes are. As practitioners we see a lot of junk that claims to be "Montessori" but doesn't really stick to the principles. These are primarily: self directed learning and building a community for exploration and creativity. The best Montessori schools have agreements with local warehouses/mills/shops etc... where children can do internships and apprenticeships. Having "curricula" like the article states is generally antithetical to that and kills a large portion of what makes Montessori so successful - which is pupils taking prepared "vocational" paths which are liberal in educational scope. It's simpler in the long run but takes much more up-front work and I think the technology is there to make that up-front work much easier. The real challenge is fitting these types of schools into the larger "schooling" world of standardized tests etc... One thing we found was that many students leave Montessori because the clearest path into college is through the testing system and Monetssori doesn't play well with that. That is what needs real reform.

hullo 12y ago
If this were still May Day I would think that sending our seven year olds off to the mill would be a more obvious troll, but as things stand I'm not so sure.
- brlewis 12y ago
  The Montessori school my oldest went to did week-long apprenticeshps in 7th and 8th grade. I'd be surprised if any school had 2nd grade apprenticeships.
- saraid216 12y ago
  They're not expected to go to the mill to help support their family.
- notahacker 12y ago
  I recall that after two weeks of "work experience" on a print-shop production line as a teenager I couldn't wait to get back to the relative blissful liberation of an organised lesson programme and essay tasks based on the exam curriculum...
- AndrewKemendo 12y ago
  Apprenticeship =/ child labor.
ironchef 12y ago
Montessori isn't the end-all be-all of "student centered learning". There are a LOT more types of schooling out there which have similarity as they seem to have roots in Piaget, Steiner, etc. This seems to be another take on montessori + some reggio emilia and other bits. Also, as an fyi, one of the better montessoris in our area (montessori academy of chicago) has taken a big bite of the standardization area by aligning common core with the classic montessori didactic system.
- AndrewKemendo 12y ago
  >Montessori isn't the end-all be-all of "student centered learning" I agree in theory, but have yet to see another system as flexible or robust with any actual, measurable history of good outcomes at scale. All other systems seem to be on either extreme - either too much "freedom" and the kids never learn, like Summerhill, or are too focused around some oddly specific dogma, like Waldorf.
  - ironchef 11y ago
    I didn't mean to suggest that Montessori wasn't good or that the other systems were better. All I was suggesting was that there are a multitude of student centered learning frameworks (some of which are montessori, waldorf, reggio, and summerhill). I think based on the concepts of differentiation, some of these may be more optimal for certain learners than others.
germinalphrase 12y ago
Hopefully we will see the influence of standardized testing companies diminish as more robust in-class assessment tools are developed. Our trust in standardized testing is dependent on the expectation that teachers cannot accurately assess the skills and competencies of their students during regular lessons. To free schools to approach education however they want we just need to find a way to reliably assess students without the need for a centralized gatekeeper. If the process of developing/proctoring/critiquing assessments was partially standardized (to allow for distribution and certification of assessment elements) and linked to student performance data then we could both reliably integrate this kind of assessment into regular lessons while also mitigating the need for expensive standardized tests. There are some hints of this in the article; as a HS teacher I'd love to see assessment overhauled in this direction.
- aidenn0 12y ago
  Standardized curricula (which is admittedly distinct from standardized testing) is useful as prior to statewide standards, it was not unusual for a move of just 10 miles to put your children in a school district with completely different curriculum for at least one subject. The most positive thing I've heard about Common Core is that a federal standard will mean out-of-state transfer students will be more likely to be on the same page as current students.
  - brlewis 12y ago
    Alternatively, take an educational approach where it's not essential that all students be on the same page.

hmate9 12y ago

Glad to see technology used to enhance education. I like their approach and hope they succeed (with Peter Thiel on their side, it should almost be a guarantee).

diminoten 12y ago

If a 5th grader "decides" to learn some algebra, those are the cards the teacher will start preparing for that student, right? But to "learn" algebra, one traditionally learns a set of concepts and practices those concepts in the form of large quantities of practice problems. Every math teacher I've ever met has said the key to mastery of math concepts is practice and repetition. So, if we go back to our 5th grader, and algebra comes up for him/her in the customized lesson plan, does that mean for the next year, algebra cards are going to appear, with a set of problems to solve, and links to... what, wikipedia pages that explain how to solve them? What happens when the kid decides he/she doesn't want to "explore" math anymore? Does that kid just not get a math education, at that time? Are we going to have 16 yr olds who never figured out that "exploring math" is going to be a key to being a competent adult? I'm sure what I wrote is quite commonly brought up and an answer exists, but I do wonder what that answer is -- to sum: what do you do when a kid doesn't want to "explore" an area of knowledge that we as adults know is going to be critical to functioning as one of us in the world?

Marcus316 12y ago
Not sure where you got the idea that it is entirely the child's choice what he learns. What I got out of the article is that the child's interests, strengths and weaknesses are taken into account to produce an individualized curriculum. If a child is weak in mathematics, the curriculum is adjusted for _that child_ to emphasize mathematics. It's not as simple as "Johnny wants to learn poetry and only poetry, so that's what we let him learn." At least, that's what I got out of the explanation. It's chaotic in that every student needs to be guided differently, and AltSchool is using technology to make that guidance easier for the teachers to apply across individuals. The idea is not to lose students in the crowd, but also not to let them lose themselves in the chaos.

anonbanker 12y ago

This school looks like a terrifying extension of Snow Crash's "living-in-a-storage-container" lifestyle, combined with training on how to be a perfect worker drone for the new tech/service-based economy (they're even teaching the kids to clock in and out!). If this is where private/public school is headed, I'll stick to homeschooling/unschooling my kids, thanks.

ryoshu 12y ago
Seemed a bit more Diamond Age to me.

glup 12y ago

A surveillance-state Montessori with a big data fetish and a bottom line doesn’t count as reform, sorry AltSchool. I was particularly struck by: > “It’s not hard to model language acquisition if you can listen to every word a person is saying,” This is patently false big data hubris, which I’m afraid bodes poorly for the rest of the rest of their approach. I study early language acquisition and have worked on some very large datasets (e.g. Speechome, http://www.ted.com/talks/deb_roy_the_birth_of_a_word http://www.ted.com/talks/deb_roy_the_birth_of_a_word), and there’s a lot you don’t know about the language system of a child even with a complete record of utterances and videos of everything that happens. Higher density/quality data is fantastic, but to understand language acquisition or to understand the progress of a single child you have to work out a complex system of latent variables. Education is similar— the biggest thing is that there is s different mixture of explicit pedagogy vs. observation. I’m all for school reform that leverages technology but this misunderstanding of data, in conjunction with the for-profit aspect, leaves me pretty convinced that this isn’t the right direction. The other thing that caught me off guard was: > “and wearable devices that help keep track of students off campus” I did a double-take. Why does this school track the kids off campus? The complete lack of criticism eliminated the last shred of credibility Wired had in my mind. I don’t care if you think this is a good idea or not, tell me who doesn’t think it’s a good idea and why.

pbreit 12y ago
> Why does this school track the kids off campus? So they don't get lost on field trips?

Aloisius 12y ago

I'm all for doing whatever we can to improve education, but it is extremely difficult to compare the quality of a education between public and private schools because of selection bias. Private schools tend to pick and choose the best kids and have significantly higher suspension and expulsion rates eliminating the "problem" children. Public schools on the other hand, have to accept everyone from the hyperactive to those with learning disabilities. When we try to control for these effects, often there is little to no difference between these custom approaches and standard ones. Montessori for instance does customization and outcomes are effectively the same as public schools. I suppose though even if the end result is a cheaper and highly replicable education system, it could become quite successful as a company.

pbreit 12y ago
Isn't the main point of private schools to pay for the selection bias?
brlewis 12y ago
Please cite a source for "Montessori...outcomes are effectively the same as public schools". Here's a study that found the opposite: http://www.scientificamerican.com/article/students-prosper-with-mon/ http://www.scientificamerican.com/article/students-prosper-w...
- Retric 12y ago
  At the study level sure, you can support any argument, a meta analysis like: http://www.irp.wisc.edu/publications/dps/pdfs/dp114197.pdf http://www.irp.wisc.edu/publications/dps/pdfs/dp114197.pdf Generally shows little to no benefit and often points out flaws in individual studies. Indeed, in our data, previous authors’ specifications rarely explain a substantial portion of the selection into the relevant privateschool sector, raising the possibility that the weak-instrument critique may hold for this literature. This point is corroborated in work by Ludwig (1997), who also finds compelling evidence that prior authors’ instruments are weak.
  - 12y ago

coldcode 12y ago

Everything can't be improved by adding Silicon Valley to it.

michaeledwards 12y ago
But everything CAN be improved by only adding silicon to it!
- crpatino 12y ago
  Or silicone...

danans 12y ago

There are several distinct problems in the public education system. There is the problem where motivated and largely well-off parents find that the public education system doesn't suit the unique interests and learning styles of their children. This sort of school seems like a wonderful experiment to address this sort of problem. A second, largely unrelated problem is the under-funding (in CA for sure, perhaps elsewhere) of public schools, given the expectation that these school systems rectify the serious motivational and educational disabilities students develop in communities with high concentrations of poverty and violence. I'd argue that the latter problem is the one that has serious societal consequences if we don't get it fixed, sooner than later.

Domenic_S 12y ago
> A second, largely unrelated problem is the under-funding (in CA for sure, perhaps elsewhere) of public schools I hear this often. I'm curious, do you have a dollar figure in mind that would be appropriate?
- aaroninsf 12y ago
  CA spends less than half as much per student as New York: http://www.usnews.com/news/blogs/data-mine/2015/01/29/how-states-are-spending-money-in-education http://www.usnews.com/news/blogs/data-mine/2015/01/29/how-st... What the 'right' value is subject to debate, but as a parent in SF having negotiated the tragicomic SFUSD system for my own children, I'd wager it's a lot closer to New York's 19K per student than CA's $9K. This is pure speculation, but my subjective experience is that CA funds schools at close to the minimum possible level, and that there is a lot of headroom in which every dollar extra spend causes an immediate improvement in service and richness of experience. I say this because one of the differentiating factors of the most-desired schools in the SFUSD is the large amount of extra money parents bring in via fundraising. Yet even the most successful schools, which raise on the order of half a million dollars, are only adding something like ~$1K per student. Yet what they offer above and beyond comparable schools without that funding bump is very evident: all the 'enriching' staples my own generation took for granted as par of public education: art classes, PE, green space, staff with expertise in supporting kids with differing needs (at both ends of the performance/ability spectrum)... etc etc etc. I have also observed that affluent parents who privilege education also both prioritize and have the luxury of donating their time -- no small thing in a city in which a single-income household is mostly unthinkable. So that extra $1K may be multiplied by donated time, as well. But any way you slice it, a few extra K per student, judiciously spent, would make a hell of a difference in my parts. That said, a dollar raised and allocated by the parents, in accord with parents' priorities, and subject to their audit, is probably much better spent than one squeezed through the SFUSD. QED non-profit charters (of which there are several) are currently a very strong bet in SFUSD...

pbreit 12y ago

Is there any evidence that private schools make much of a difference in the overall education landscape or is it just yet another option for rich people? If you really want to make a difference you need to address the broader audience and do something like charter schools.