Private schooling appears to be associated with better outcomes but also with being from a richer household. Would you consider splitting out the results of upper-middle and rich to see if the benefits of private still hold?
Hmmmm if I were to homeschool my future kids, one of the primary motivations would be so they wouldn't be brainwashed into being trans... But this shows that that homeschooled kids are the most likely to go trans
Parents who choose different approaches to schooling are different, and their children thus differ, wherever you stand on the nature/nurture thing. Children also often have input, and children who push in different directions are different.
This is beyond confounded, and can't be fixed by tossing a couple of variables in the magic unconfounding jar.
Apologies if this comes across as rude, but this is the kind of social science statistical analysis that really winds me up.
As a voucher enthusiast I am surprised that private and home schools don’t have more of a lead over public schools in terms of satisfaction. Maybe it’s because of the high achieving selection of ACX readers? Kids who are smart are more supported by teachers and administrators?
I understand that there is a general skepticism about p value and statistical hypothesis testing, but without even a standard deviation those values are pretty much useless.
Gender non-conformity is quite strongly associated with poverty, so it is not a good measure of weirdness when your categories are correlated with poverty. Specifically, higher rate of gender non-conformity in home schooled population may be partially statistically explained by the higher rate of poverty.
The basic question is: what in the parents' status, education, living place, specific way they care about their children's wellbeing, you name it ... affects whether the child goes to public/private/religious school or is homeschooled/ unschooled. All of those things could directly affect the outcomes we see here (life satisfaction, being poorer or richer, ...).
I find the results interesting to read, and I don't disagree with Scott's conclusions. I think this type of presentations tempts to read more into those numbers than they actually tell you. Well, maybe ACX readers are exceptionally good at resisting such temptation?
The main problem here is that while most American public schools are similar, each home school is very different. It seems wrong to try to make a cohesive group out of homeschoolers.
How do you differentiate between home schooling and "unschooling but encouraged to learn unstructured"? It seems more like different pedagogies than different categories. And if you are truly unschooled i.e. no one bothered to give you an education, and you couldn't avail yourself to the public option, I would tend to assume a broken home...
Unstructured learning does sound interesting for some kind of students (highly intelligent AND highly curious AND highly self disciplined and possibly adverse to standard socialisation) but highly dangerous for the rest of us, even the weirdos/nerds amongst us. I know there's no way I would have put myself through any kind of complex scientific topic if I hadn't been forced to.
OTOH, it's not like my career or hobbies really involve or necessitate lots of scientific knowledge...
I did a quick ordinal regression looking at school satisfaction predicted by schooling * childhood class. I dropped unschooling due to low n across class categories, and similarly combined "rich" and "upper middle" for the same reason.
Childhood social class had a strong and consistent effect on school satisfaction, with OR of 1.69 for the second lowest social class category compared to the lowest, 2.56 for the next highest compared to the lowest and 3.31 for the highest compared to the lowest. All significant to p<0.001.
For school category, there was a "borderline significant" difference between home schooling compared to public schooling OR=3.01, p=0.56. Neither of the other two comparisons were close to significant, though trended in the direction of both being higher than public schooling (but the confidence intervals were extremely wide i.e. OR=0.69-3.6 for religious vs public school.
The average age was about 30, which is very similar to the average age from the 2014 survey (29) from 9 years ago. So it looks like the readers aren't ageing along with blog. Which suggest people tend to loose interest over their 30s and get replaced with new readers in their 20s. It definitely seems like this blog is a thing that specifically-guys-in-their-late-twenties enjoy.
It's reinforcing my suspicion that curiosity/mental acuity peaks around age 30 and drops off pretty strongly afterwards. It's when a lot of the great physicists/mathematicians made their discoveries (Newton, Einstein etc. Scott himself?).
I'd like to ask the 40 year old guys why they lost interest, but I guess they're not around to answer.
Have any older guys who do still read the blog experienced anything like a loss in intellectualism/curiosity?
I’m in my thirties, and I’m actually surprised today I don’t remember my SAT scores at all. Or even the GRE. I’m pretty sure I remembered them for years after, but at some point they vanished. Do me a favor and look at the most recent survey for people who left the SAT question blank, and maybe you can find the average age we forget at…
> Yes, these are the averages of hundreds or thousands of SAT scores in each cell. I *told you* that you were heavily-selected and weird. Given the level of selection effects, I’m not sure we can conclude much here.
I should call out another selection effect in those scores that is only obvious if you're familiar with SAT scores.
The SAT-math is much easier than the SAT verbal. Accordingly, a score of X doesn't mean the same thing on the math that it does on the verbal - it's more impressive as a verbal score than it is as a math score.
But the average SSC SAT scores are significantly higher on the verbal side than they are on the math side. This is not expected; a typical population would have equal ability on both sides, which would translate to the math scores being higher than the verbal scores. SSC readers appear to be noticeably bad at math (given their high IQs).
My main takeaway is that effect size on everything is surprisingly low. Is it just that almost everybody rates life satisfaction in the same fairly narrow range? Due to hedonic treadmill etc.
Your relationship question doesn't seem to ask whether the person was EVER married or in a LTR, but a current status. I realise that considering the huge bias towards youthfulness of your population it is less important, but unless you controlled for that, your "singles" will include a substantial % of people who HAD a LTR which ended. For example, I'm a 52 years old widow after a 20+ years of marriage (likely rare but increasing in probability with age) and undoubtedly there will be many divorced people out there. So "not YET" married seems like an overstretch.
It's a weird category because it's made up partly of families who chose it from the outset (therefore already social outliers), and partly of families whose children struggle in school (often because of traits that are bound to have other effects in life). It also covers children who are 'unschooled' at home and those who attend self-directed learning centres. Either way, I can only see the category growing in years to come – I'm very glad Scott is interested in it.
I'd be very cautions about generalising school type across countries; implications of school type will differ a lot between the US and Europe (and probably even more outside the West).
For example, in the UK a lot of schools, including a big chunk of government schools, and more-or-less all private schools, are "religious" in theory, and to an American would look unacceptably religious - they have chapels and mandatory school prayer (all schools are legally obliged to have weekly collective worship in England and Wales). In practice, these should still be rounded off as secular schools for all intents and purposes, as no-one takes the religion seriously.
In Belgium, there are quasi-private schools paid for by vouchers that are fairly common, but private schools that you pay for are banned for Belgian citizens. Again, some of them are Catholic but this isn't terribly important.
Several countries also have varying degrees of technical school at the secondary education level which focus more on practical/trade skills than academic learning, although these are generally being phased out.
Growing up homeschooled in the US, we learned that there were two groups of non-conventionally-schooling families: hippies who didn't want their kids locked inside and forced to recite the Pledge and conform (cf. The Teenage Liberation Handbook), and Christians who didn't want their kids taught to believe in evolution (cf. the loudest group of homeschooling boosters these days, having switched from "evolution" to "woke gender BLM ideology"). These groups interact only at arms' length and have typical pedagogy and ideology differing from the conventional school system in opposite directions, so I'd expect pooling them before analysis to give confusing results.
This isn't exactly captured by homeschooling vs. unschooling (we were hippie homeschoolers), but could be captured pretty well in this dataset by rescoring each "ReligiousBackground" as "hippie" (e.g. Unitarian, atheist) or "trad" (e.g. Catholic, Mormon, Jehovah's Witness), and combining homeschooling with unschooling. I'd be interested in doing this and looking at satisfaction, SAT scores, and adult counterculturalism when it's not the middle of the workweek.
> Wow! Home schooling is even more religious than religious schools!
Well, yes, that's what home schooling is. It's not bespoke tutoring for future Isaac Newtons. It's a way for religious people to keep their kids out of the clutches of the godless state. The energy behind legislation for making it easier to home school your kids comes from religious people.
I have some familiarity with this area because I took my kid out of the public school system during the pandemic, which meant that technically I was a home schooler (we hired a teacher with some other families). The two things you immediately discover when you do this is that 1) it is shockingly easy to take your kid out of public school and 2) you need to look around to find the other secular folks.
For #1, you barely have to do anything at all. Did you imagine that there would be elaborate forms to process, maybe some kind of proof or at least attestation that you are providing a semblance of education to your kid? Perhaps a chat with a state social worker? Nope. Just let the school know, "Hey, we're not sending our kid next year."
Regarding #2, there are tons of home school class materials and curricula available, but a lot of them are things like, "Math, but with Jesus" and "Spare the rod, invite trans Satan into your home."
In case it sounds like I'm anti-home schooling: I'm not. I'm sure there are some parents who do a terrible job home schooling their kids -- it would be shocking if that weren't that case -- but in the main I suspect it works out fine for most families, and of course people are free to instill religious values and beliefs in their children, for better or for worse. But this is very much what home schooling is.
This mostly jives with my experience (homeschool, religious upbringing). Your readership might skew in the direction of liking homeschooling more, though - I liked it mostly because it gave free rein to my curiosity. Some of my friends and siblings who were less intellectually self-driven had a lot harder time with it.
Socially, I'm highly extroverted but didn’t learn that until later in life. I would say that I had a stunted social experience growing up that in turn made me the typical homeschool type: awkward and anxious, etc. I know other kids who were similar. But it took about two years of college to get over 90% of that. In my observation, homeschoolers start off weird, but they figure it out. By my late 20s, I rarely thought about anything that happened in high school and could barely tell the difference between friends who were homeschooled or public schooled.
Looking across my friends and acquaintances, I'm persuaded by the Harvard study idea that loving and supportive parents are much more influential than structural differences in upbringing. I have watched people with cruel, controlling, or highly ideological parents go in much different directions than those whose parents were present, reasonable, and loving - whatever their schooling background. Some of the worst homeschooling outcomes I know aren't really from the homeschooling itself - they are kids whose parents homeschooled them because of their own controlling and narcissistic instincts. One friend in particular is a huge queer anti-homeschooling activist now. Her argument is basically: "if I was in school there would have been other adults who could catch what was going on."
If you're an educated person thinking about homeschooling your kids, I would start with your own motivations, not a theory of education. Why do you really want to do this? If the answer is something along the lines of "I think this would be good for my child," then your kid will probably be fine as long as you make an effort. If it's something like "I want to have full power to craft a culture war weapon that I can then unleash upon the world to validate my ideas," then I guarantee your kid will be all kinds of messed up (and will probably hate you). I'm less familiar with secular private schools, but I imagine the same dynamic exists. If you're sending them to Braxley-Sheaton-Hogwarts because you want them to have the best experience possible, then they'll probably be fine (but they would probably be fine in a government school too). If you're sending them there to try to ensure they carry on the family legacy and make you proud, then they probably won't.
In future analyses, can you put confidence intervals around your means (or give some other measure of spread like standard error)? It’s really easy to do, and it’s hard to eyeball-interpret the differences between means without them.
I'm a little surprised at the surprise of how religious home-schoolers are. I consider myself religious, attending a church in my community regularly. But there is a group that is far more religious than my congregation on average, and that is the subset of the community that embraces home schooling. I think they are mostly trying to avoid the 'evil influences' of public education or something to that effect. And I think this is at least common across the whole region where I live. Rather than surprise, I would hypothesize that home-schoolers across the nation are even more religious than the home-schooled ACT readers, and yes, more religious than the average religious school attendees. Keep in mind that many religious schools are Catholic. I'm not trying to disparage the Catholic faith, but it strikes me as one of the least religious religions.
Devil's advocate position on my own subgroup: I wonder if homeschoolers (not specifically those in your sample, but all of us) might not be slightly conditioned to answer higher on "are you happy, was homeschooling good" type surveys. There's a thing where a lot of people will ask you those questions in an adversarial way - i.e. "Isn't it hard to be homeschooled, considering you are a friendless loser freak?" or "What do you think about homeschooling your kids, considering it will ruin them forever and ever?"
When that happens, I think one of the normal human reactions to that is to be like "No, it's great, in pretty much every way", because if you say admit to any flaws at all in any tone at all, sometimes the reaction you get is "See, I knew it was a horrible practice in every way and that I can now smugly look down on you for it".
I've been trying to figure out a way this kind of effect could even be quantified, if it exists. Were gay marriages more likely to report happiness shortly after national legalization because they knew it was a politically relevant question? What about transitioners? What about the religious in general?
Thanks for this, Scott! Quick question about the original survey - was there any way to control for responders who fit in multiple schooling categories (e.g., public school for awhile, then homeschooled)?
In my experience (which may not be generalizable), lots of homeschooling families start off sending their kids to public, get disillusioned, then try homeschooling, realize it's too hard as the kid hits the upper grades, and then fork over the money for private or religious school. This lines up with my own background as well: public school until I had a major medical condition, homeschooling during the medical condition and for a couple years after, then religious school for grades 9-12. How would a person like me get diced up in the survey? :)
> Wow! Home schooling is even more religious than religious schools!
I wonder how this correlates with age, but it doesn't surprise me. Most Catholic schools, in my experience, do an awful lot to undermine the faith -- teaching it badly, hiring teachers or ultra-heterodox nuns that vocally dissent from the faith on nearly every core belief (the real presence, abortion, contraception, the incarnation, the immaculate conception, Jesus being male, extra ecclesiam nulla salus, etc.). I go to a traditionalist parish, and a lot of families here homeschool because the religious instruction and formation is either lacking or wrong in most of the local Catholic schools.
To publish this results without publishing standard deviations, should be embarrassing to someone who is attempting to be serious and claiming a "rationalist" moniker.
You were surprised that home schoolers are even more religious than people attending private religious school?
Maybe it's because my son is in catholic school and maybe it's because I know several people who home school their kids - but i knew this from experience.
People who don't like the libs and are vaguely religious send their kids to religious private school.
People who are super duper committed to their religion home school their kids.
Unschool responder here! One confounding factor is my very religious parents would have said homeschooling. I started in 6th grade because the public school was ‘too worldly’ and my small town didn’t really have other options. My mom had enthusiasm for the task at the start but that very quickly dissolved into an obsession with tennis and other personal interests. I was vaguely encouraged to read, but it was really my choice- assuming I stuck to Christian books. There was no curriculum. There were no taught lessons. She did buy me math textbooks. I devoured them and even gave myself math tests which I graded myself. I think this all says a lot more about me than it does the quality of the education. For science I was once supplied with an anti evolution trac. I don’t think it is fair to homeschoolers to call what i experienced homeschooling, but this isn’t a regulated industry. Quality varies widely and there may be less of a distinction than you are thinking. And, as for satisfaction with their education, the less satisfied homeschoolers are perhaps, like me, inclined to label it unschooling.
This makes me (very very slightly due to not thinking these results say much of anything at all) feel better about sending my daughter to public school. Aside from things that could easily be the result of confounders, there's not enough difference here to make me think schooling type has huge implications on life outcomes, at least for folks who read this blog.
This was really interesting - thankyou! As the parent of an 18 year old boy and 15 old girl, I have asked myself many of these questions - especially since the covid lock-downs. Your findings make me feel like maybe I HAVEN'T made any huge mistakes in their educations thus far.
I had some fun playing with the data; a couple general points:
- agree that you generally want to work with interaction variables of childhood SES x school type
- I would be skeptical of applying many results from the overall survey data to girls and to average/below average test scorers because the sample is so skewed
Result #1: American men age 30 or older whose childhood was "middle class" were less likely to currently describe themselves as "rich" or "upper middle class" if they were homeschooled than if they went to public school (p=0.013). Noting that the coefficient for private school was positive but insignificant (p=0.140).
Result #2: American men who had indicated their education was complete were more likely to have an advanced degree (mostly thinking of this as another way of asking "how much did you like/excel at school?" and the high SAT score of the sample and similar average across modes of education might actually be a plus here) if their childhood was one of the richer-than-middle-class categories, and less likely if their childhood was one of the poorer-than-middle-class categories. The only school type or interaction variable that was significant was children from a poorer-than-middle-class background x homeschooled being particularly less likely to have an advanced degree (p=.021). Meanwhile, while "private x poorer" and "religious x poorer" weren't statistically significant, their coefficients were positive and of greater magnitude than the negative coefficient for the poorer-than-middle-class variable alone. Given how underpowered these groups are, it supports a theory that smart kids from poor backgrounds like school better in those two environments than in public school.
Might try to think of more things to look at later.
Speaking for myself, my son is autistic and has ADHD; we made the large financial sacrifice to send him to private school, because he was falling apart academically and socially in middle school. You definitely need to include learning differences and social dysfunction in the questionnaire.
Calling out once again that your survey didn’t ask people where they grew up or did most of their schooling, it asked which country they most “identify with”. (Whatever that means.)
There’s likely some fraction of people who answered “United States” for the “country” question, but are immigrants who grew up somewhere else. The design of your survey makes it impossible to tell how big that group of people are and might lead to inaccuracies in generalizing about the US (or any particular country).
If you say the non-US numbers aren’t too different here, then maybe it didn’t affect this specific analysis too much, but I’d be on guard for this general class of errors.
On homeschooling and religiosity: one obvious factor:
If you are religious family, homeschooling allows you the scheduling flexibility to attend more than one church service per week-- this is a bigger thing for the more "high church" denominations that are likely to have morning services on weekdays, when most kids are in school. IIRC there is a strong correlation between number of services attended per week as a kid, and religiosity as an adult. Our own church surveys indicate that kids who only attended Sunday services (i.e. once a week) have a lousy retention rate as adults-- once a week isn't enough.
One of the supposed benefits of unschooling is that it can help people develop greater creativity. It would be interesting to look for this in the data. Is there anything in the data that could serve as a proxy for creativity?
Regarding the high religiosity of home-schooled subscribers, this should not be surprise, because the home-schooling movement was strongly pushed by fundamentalist Christians.
The following story is worth approximately 0¢, statistically, but I find it amusing, so I'll tell it anyway. The son of a friend of mine was kicked out of HS for being a disruptive cut-up. He also committed pranks verging on the criminal toward a teacher he hated. He was taken to court and the settlement required the family to go to counseling for 6 months. My friend thought this was a draconian sentence, but afterwards, when I asked him how it went, he said that actually, it went well. One outcome was that his son, an artist who stayed up till 4am painting, should be home-schooled. His parents (who are ardent atheists, though they love to do a traditional Xmas, in which all the tree ornaments are hand-made by the family), let their son live on whatever schedule he liked, but they did give him assignments in the required state subjects. He applied to a very prestigious art academy and got 800s on his SATs. He was admitted and got to every class on time, whereas many of his fellow students stayed up all night and didn't necessarily go to class. I asked him about that and he said something like "Well, I already did that, but now I'm at a place I love, doing what I love."
I noticed that some respondents put in their full SAT scores (e.g. 1550) instead of the subsection scores when answering the survey. How did you exclude these when taking the average scores for the verbal and math questions?
I am not sure the difference between home schooling vs. unschooling is robust - i.e. that people
who studied at home interpreted the question uniformly enough to answer one or the other according to the same criteria. In other words, two people who received the same home education may have given different answers to this one. How do you account for that?
It would be worth running some of these controlling for social class. For instance, none of the rich either home schooled or unschooled and none of the upper class unschooled, so if richer people are more satisfied with life or more likely to be married or ... that could make home schooled and unschooled look worse on those measures even if, controlling for social cost, they were better.
I wonder what results you would get if you dis-aggregated public schools based on whether the school was a magnet school, or some other measure of school quality.
I was told by a psych PhD who studies giftedness that this school is easily the best school available for profoundly gifted kids: https://www.davidsonacademy.unr.edu/ They call themselves a public school, but they hardly seem like a typical public school (e.g. "Our accredited classes group students by ability rather than age", "Our unique approach to learning has attracted families from across the country and globe")
Hey @Scott, I was confused about one thing in the Survey. I'm retired, which I said early on. But then, much later, you asked things about "working" It wasn't clear to me whether I should answer that or not, and I I think I did answer.
Though retired, I do things that I consider "work", but not for remuneration. For example, one year, I was Treasurer for a non-profit org, and that was definitely work. But I also have some individual projects that I work on.
So I wish you had defined work more explicitly; For example, "activities associated with a job for which you are paid", if that's the definition you meant, whjch would have led me to pass up the question I referred to above.
Were the SAT averages calculated using corrections for the time period they were presumably taken, at least to address the major shifts? If so, what equivalence tables were used?
This seems like it adds up to weak argument for sending your kids to public school. All the effects are basically negligible, private school is expensive, and homeschooling is a ton of work.
My barely middle-class family scraped enough cash together to keep us in Catholic schools in the 1950s. We weren't particularly religious, but for some reason our folks decided parochial schools were better. The nuns were effective and exacting, and that ten years of education was one of the greatest gifts our parents gave us. (Serious religious women ran the schools; priests or brothers were rarely seen.) We didn't know what 'diversity', 'inclusion', or 'integration' was; classes were always a broad mix of 'races' or cultures with a smattering of Protestants, Jews, or nonreligious students.
Some non-denominational, evangelical schools seem to have adopted the more rigorous academic aspects of the old Catholic schools, but infused them with the more populist doctrines of American evangelical Christianity and appear to be doing quite well. But they are expensive, certainly unavailable to most households making less than six figures.
Yet, regardless of current arguments regarding rigor, regardless of attempts to inject white guilt and gender ideology into curricula, our town still has an Immaculate Conception School still run by nuns still proud to send most of its graduates to college.
This actually remarkable how much a difference schooling makes. Surely somewhere between 60 and 70% of public school people (far more than 8%) were raised with a household religious belief, yet for public school belief waned to 8.4%. Compare to home and religious school where one gets much better "retention" as it were. This seems very important for thinking about things like ideology in schools and whether it matters. In the religious context it does. Contra Caplan et al. I'd surprised if ideology in schools didn't matter.
The results indicate a strong positive for public schools because many of the factors associated with poor outcomes end up relegated to the public school group: working single mothers, messy divorces, financial insecurity, children raised by foster parents, parents who don't think much about education.
Meanwhile anyone who homeschools their kids is by definition putting enormous resources and effort into their children's education. The homeschool group will be composed of stable two parent households who are making huge sacrifices in order to do what they think is best for their children. Of course private school families are dumping a lot of money into education, so they have means and value their kids' learning.
Seeds of Science (ACX grant winner), a journal publishing speculative and non-traditional scientific articles, would like to offer itself as a peer-reviewed publishing platform for any analyses using the ACX reader survey data. Visit the website (theseedsofscience.org to learn more or contact us at info@theseedsofscience.org.
The unschooling results make sense to me; a decent proportion of the unschooling families I've known over the years appear to have opted for unschooling simply as a term to justify their severe educational neglect of their kids, often accompanied by other types of neglect as well.
Sorting that out from people who have ideological convictions about unschooling and do provide learning opportunities to their kids consistent w/the kids interests and needs would be complicatey.
My three children all unschool. It’s a second generation thing for them. I went to school till fifth-grade, homeschooled with a curriculum in sixth-grade, then unschooled after that.
Unschooling for me is definitely a way of life, and one that puts us far outside the societal norm. We’ve made seven moves since we’ve had kids, three times across state-lines, and not having to worry about “finding a good school” is a tremendous advantage to mobility. Even on a day-to-day basis, not having the rhythms of our family life determined by school schedules gives tremendous freedom to engage in the world. It’s very easy to go see a super-bloom, stay up late to look at the stars, take time when grandparents visit, or whatever else.
Our kids all take their education seriously. For what would be high-school, our oldest daughter volunteered in a university lab as a worker. She helped on several projects, ran equipment, performed bench work, and was even named as (an admittedly minor) co-author on a published paper. She’s now a sophomore at a UC school and began working in a lab her first term of college. Our two younger have been at least as aggressive in their own chosen domains.
While we definitely let the kids follow their own interests, we do insist on basic literacy and numeracy. While it hasn’t been a problem for us, were our kids not to be able to read or multiply and divide I’m sure we would have insisted they learn. But as I said, it just wasn’t an issue. My position on this is probably guided by my own family experience. My father has a B.S. in math and never uses it. My brother, who also unschooled, got into college not knowing much more than the most basic pre-algebra. But there, he fell in love with math, got a B.S. in math, and uses higher math on a daily basis.
While we mostly let our kids follow their interests, we do not desert them to figure things out themselves. And sometimes we do push them in certain directions. For instance my 11yo son wants to be an engineer and he’s making games in Unity. Earlier this month, and over his objections, I insisted he start using Co-pilot or Ghostwriter.
As i said in the beginning, our way of life puts us far outside the norm in ways that can be dangerous. For instance, about a decade ago, when we moved to a new state, we put the kids in school as an experiment. My wife had found a small horse-farm on the edge of town, and the kids were volunteering there a couple days a week to learn to care for the animals. We’d also take days to explore the new town, hike, visit museums etc., without too much regard to the school schedule. Within less than a month we got a call from the school asking about our kids unexcused absences. We explained the absences were perfectly fine, and that the public school was just one of several experiences we were affording our children, that we weren’t concerned with grades or any particular curriculum. The school told us if we had a handful more absences (I believe it was three) they’d need to call child protective services. We, of course, pulled the kids from school immediately to avoid the danger.
The point is, our natural way of living can very easily lead to severe consequences. For that reason, attending community college or college is important for the kids. Not so much for the actual content of the classes, but for the training in how to navigate and avoid conflict with the bureaucracies of our society.
People are frequently concerned about the social costs of homeschooling or unschooling. Certainly, it is more difficult as a non-schooler to meet kids your own age. And most of those kids, being schoolers, have their lives and realities dominated by a bureaucracy that is entirely alien to the non-schooler. On the other hand, the schooled environment is traumatic for many. And, as a non-schooler, i feel there is something abusive in sorting children into tightly-bound age cohorts, forcing them to sit still in rows of desks all day, at the mercy of an authority figure they’ve had no part in choosing, and to whom they must petition for even the right to use the bathroom.
But even if I were to accept for the sake of argument that the schooled environment is a healthy one for socialization, between sports, online communities, and homeschool groups, there are plenty of opportunities to socialize. The main social problem I see is finding romantic partners of the appropriate age. There’s really no good way to reliably meet your high-school aged crushes or boyfriend(s) as a non-schooler. So that mostly has to wait till college or the workplace.
That said I did have one girlfriend when high-school aged, don’t feel in retrospect I had more than the normal amount of awkwardness in my brief career of college dating, and ended up meeting my wife in my junior year. And, as a mid-forties man, I don’t feel the lack of socialization in public school kept me from developing personal friends or professional relationships.
A final thought on “unschoolers have low self-reported life-satisfaction.” Putting aside the small sample size and overall tightness of the reported values, I doubt the question compares apples-to-apples. Life satisfaction is essentially a measure of existential angst, and the unschooler’s existential angst is very different from the schooler’s. The unschooler is always forced to figure out for themselves what they want from life, whereas the schooler can allow themselves to be led by their peers, the system, or authorities. And so, especially at the younger ages answering the survey, it seems to me the schooled may be more likely to feel satisfied with their life because of external affirmations - graduating college, having a job, etc. Whereas the unschooler is more likely to judge their progress by their own standards. These seem like different enough yardsticks to me that we cannot reliably compare the readings between the two.
And a final finally, lest I’ve seemed to hate on public education or educators, I certainly recognize the tremendous social benefit the public school system offers in its breakfast and lunch programs. Many otherwise abandoned children obtain their only physical and emotional sustenance from the adults running their schools. Unschooling worked well for me and mine, but I really won’t confidently generalize its utility for children who are not natively curious, or lack parents who are devotedly solicitous of their care.
Many of these comparisons scream for a regression, instead of other adjustments. I feel bad saying this when I don't have time to run them myself and publish the results, but just noting it would make things more reliable and easier to communicate.
Also, should turn some of these tables (especially the 2d ones) into charts.
Private schooling appears to be associated with better outcomes but also with being from a richer household. Would you consider splitting out the results of upper-middle and rich to see if the benefits of private still hold?
Hmmmm if I were to homeschool my future kids, one of the primary motivations would be so they wouldn't be brainwashed into being trans... But this shows that that homeschooled kids are the most likely to go trans
Makes me think twice about it
Parents who choose different approaches to schooling are different, and their children thus differ, wherever you stand on the nature/nurture thing. Children also often have input, and children who push in different directions are different.
This is beyond confounded, and can't be fixed by tossing a couple of variables in the magic unconfounding jar.
Apologies if this comes across as rude, but this is the kind of social science statistical analysis that really winds me up.
Absolutely cracking first article on the survey results for this year - keep em coming!!
I'm very curious now how those 2020 SSC results relate to the 2023 ACX results.
Also, can someone translate SAT scores for me?
As a voucher enthusiast I am surprised that private and home schools don’t have more of a lead over public schools in terms of satisfaction. Maybe it’s because of the high achieving selection of ACX readers? Kids who are smart are more supported by teachers and administrators?
I understand that there is a general skepticism about p value and statistical hypothesis testing, but without even a standard deviation those values are pretty much useless.
Gender non-conformity is quite strongly associated with poverty, so it is not a good measure of weirdness when your categories are correlated with poverty. Specifically, higher rate of gender non-conformity in home schooled population may be partially statistically explained by the higher rate of poverty.
The basic question is: what in the parents' status, education, living place, specific way they care about their children's wellbeing, you name it ... affects whether the child goes to public/private/religious school or is homeschooled/ unschooled. All of those things could directly affect the outcomes we see here (life satisfaction, being poorer or richer, ...).
I find the results interesting to read, and I don't disagree with Scott's conclusions. I think this type of presentations tempts to read more into those numbers than they actually tell you. Well, maybe ACX readers are exceptionally good at resisting such temptation?
The main problem here is that while most American public schools are similar, each home school is very different. It seems wrong to try to make a cohesive group out of homeschoolers.
How do you differentiate between home schooling and "unschooling but encouraged to learn unstructured"? It seems more like different pedagogies than different categories. And if you are truly unschooled i.e. no one bothered to give you an education, and you couldn't avail yourself to the public option, I would tend to assume a broken home...
Unstructured learning does sound interesting for some kind of students (highly intelligent AND highly curious AND highly self disciplined and possibly adverse to standard socialisation) but highly dangerous for the rest of us, even the weirdos/nerds amongst us. I know there's no way I would have put myself through any kind of complex scientific topic if I hadn't been forced to.
OTOH, it's not like my career or hobbies really involve or necessitate lots of scientific knowledge...
I did a quick ordinal regression looking at school satisfaction predicted by schooling * childhood class. I dropped unschooling due to low n across class categories, and similarly combined "rich" and "upper middle" for the same reason.
Childhood social class had a strong and consistent effect on school satisfaction, with OR of 1.69 for the second lowest social class category compared to the lowest, 2.56 for the next highest compared to the lowest and 3.31 for the highest compared to the lowest. All significant to p<0.001.
For school category, there was a "borderline significant" difference between home schooling compared to public schooling OR=3.01, p=0.56. Neither of the other two comparisons were close to significant, though trended in the direction of both being higher than public schooling (but the confidence intervals were extremely wide i.e. OR=0.69-3.6 for religious vs public school.
The average age was about 30, which is very similar to the average age from the 2014 survey (29) from 9 years ago. So it looks like the readers aren't ageing along with blog. Which suggest people tend to loose interest over their 30s and get replaced with new readers in their 20s. It definitely seems like this blog is a thing that specifically-guys-in-their-late-twenties enjoy.
It's reinforcing my suspicion that curiosity/mental acuity peaks around age 30 and drops off pretty strongly afterwards. It's when a lot of the great physicists/mathematicians made their discoveries (Newton, Einstein etc. Scott himself?).
I'd like to ask the 40 year old guys why they lost interest, but I guess they're not around to answer.
Have any older guys who do still read the blog experienced anything like a loss in intellectualism/curiosity?
I’m in my thirties, and I’m actually surprised today I don’t remember my SAT scores at all. Or even the GRE. I’m pretty sure I remembered them for years after, but at some point they vanished. Do me a favor and look at the most recent survey for people who left the SAT question blank, and maybe you can find the average age we forget at…
Need the confidence interval, Scott. Or at least the standard dev.
I'm admiring that I managed to take three of those five options. Four if you want to stretch Unschooling beyond reason.
> Yes, these are the averages of hundreds or thousands of SAT scores in each cell. I *told you* that you were heavily-selected and weird. Given the level of selection effects, I’m not sure we can conclude much here.
I should call out another selection effect in those scores that is only obvious if you're familiar with SAT scores.
The SAT-math is much easier than the SAT verbal. Accordingly, a score of X doesn't mean the same thing on the math that it does on the verbal - it's more impressive as a verbal score than it is as a math score.
But the average SSC SAT scores are significantly higher on the verbal side than they are on the math side. This is not expected; a typical population would have equal ability on both sides, which would translate to the math scores being higher than the verbal scores. SSC readers appear to be noticeably bad at math (given their high IQs).
My main takeaway is that effect size on everything is surprisingly low. Is it just that almost everybody rates life satisfaction in the same fairly narrow range? Due to hedonic treadmill etc.
Your relationship question doesn't seem to ask whether the person was EVER married or in a LTR, but a current status. I realise that considering the huge bias towards youthfulness of your population it is less important, but unless you controlled for that, your "singles" will include a substantial % of people who HAD a LTR which ended. For example, I'm a 52 years old widow after a 20+ years of marriage (likely rare but increasing in probability with age) and undoubtedly there will be many divorced people out there. So "not YET" married seems like an overstretch.
I wrote about our part-compelled, part-chosen entry into unschooling here: https://nickasbury.substack.com/p/the-opposite-of-school
It's a weird category because it's made up partly of families who chose it from the outset (therefore already social outliers), and partly of families whose children struggle in school (often because of traits that are bound to have other effects in life). It also covers children who are 'unschooled' at home and those who attend self-directed learning centres. Either way, I can only see the category growing in years to come – I'm very glad Scott is interested in it.
I'd be very cautions about generalising school type across countries; implications of school type will differ a lot between the US and Europe (and probably even more outside the West).
For example, in the UK a lot of schools, including a big chunk of government schools, and more-or-less all private schools, are "religious" in theory, and to an American would look unacceptably religious - they have chapels and mandatory school prayer (all schools are legally obliged to have weekly collective worship in England and Wales). In practice, these should still be rounded off as secular schools for all intents and purposes, as no-one takes the religion seriously.
In Belgium, there are quasi-private schools paid for by vouchers that are fairly common, but private schools that you pay for are banned for Belgian citizens. Again, some of them are Catholic but this isn't terribly important.
Several countries also have varying degrees of technical school at the secondary education level which focus more on practical/trade skills than academic learning, although these are generally being phased out.
The one that surprised me was that people who were unschooled were strongly less likely to be in a relationship. Any theories?
Most of the others seemed to imply that type of schooling didn't make much difference.
Growing up homeschooled in the US, we learned that there were two groups of non-conventionally-schooling families: hippies who didn't want their kids locked inside and forced to recite the Pledge and conform (cf. The Teenage Liberation Handbook), and Christians who didn't want their kids taught to believe in evolution (cf. the loudest group of homeschooling boosters these days, having switched from "evolution" to "woke gender BLM ideology"). These groups interact only at arms' length and have typical pedagogy and ideology differing from the conventional school system in opposite directions, so I'd expect pooling them before analysis to give confusing results.
This isn't exactly captured by homeschooling vs. unschooling (we were hippie homeschoolers), but could be captured pretty well in this dataset by rescoring each "ReligiousBackground" as "hippie" (e.g. Unitarian, atheist) or "trad" (e.g. Catholic, Mormon, Jehovah's Witness), and combining homeschooling with unschooling. I'd be interested in doing this and looking at satisfaction, SAT scores, and adult counterculturalism when it's not the middle of the workweek.
> Wow! Home schooling is even more religious than religious schools!
Well, yes, that's what home schooling is. It's not bespoke tutoring for future Isaac Newtons. It's a way for religious people to keep their kids out of the clutches of the godless state. The energy behind legislation for making it easier to home school your kids comes from religious people.
I have some familiarity with this area because I took my kid out of the public school system during the pandemic, which meant that technically I was a home schooler (we hired a teacher with some other families). The two things you immediately discover when you do this is that 1) it is shockingly easy to take your kid out of public school and 2) you need to look around to find the other secular folks.
For #1, you barely have to do anything at all. Did you imagine that there would be elaborate forms to process, maybe some kind of proof or at least attestation that you are providing a semblance of education to your kid? Perhaps a chat with a state social worker? Nope. Just let the school know, "Hey, we're not sending our kid next year."
Regarding #2, there are tons of home school class materials and curricula available, but a lot of them are things like, "Math, but with Jesus" and "Spare the rod, invite trans Satan into your home."
In case it sounds like I'm anti-home schooling: I'm not. I'm sure there are some parents who do a terrible job home schooling their kids -- it would be shocking if that weren't that case -- but in the main I suspect it works out fine for most families, and of course people are free to instill religious values and beliefs in their children, for better or for worse. But this is very much what home schooling is.
This mostly jives with my experience (homeschool, religious upbringing). Your readership might skew in the direction of liking homeschooling more, though - I liked it mostly because it gave free rein to my curiosity. Some of my friends and siblings who were less intellectually self-driven had a lot harder time with it.
Socially, I'm highly extroverted but didn’t learn that until later in life. I would say that I had a stunted social experience growing up that in turn made me the typical homeschool type: awkward and anxious, etc. I know other kids who were similar. But it took about two years of college to get over 90% of that. In my observation, homeschoolers start off weird, but they figure it out. By my late 20s, I rarely thought about anything that happened in high school and could barely tell the difference between friends who were homeschooled or public schooled.
Looking across my friends and acquaintances, I'm persuaded by the Harvard study idea that loving and supportive parents are much more influential than structural differences in upbringing. I have watched people with cruel, controlling, or highly ideological parents go in much different directions than those whose parents were present, reasonable, and loving - whatever their schooling background. Some of the worst homeschooling outcomes I know aren't really from the homeschooling itself - they are kids whose parents homeschooled them because of their own controlling and narcissistic instincts. One friend in particular is a huge queer anti-homeschooling activist now. Her argument is basically: "if I was in school there would have been other adults who could catch what was going on."
If you're an educated person thinking about homeschooling your kids, I would start with your own motivations, not a theory of education. Why do you really want to do this? If the answer is something along the lines of "I think this would be good for my child," then your kid will probably be fine as long as you make an effort. If it's something like "I want to have full power to craft a culture war weapon that I can then unleash upon the world to validate my ideas," then I guarantee your kid will be all kinds of messed up (and will probably hate you). I'm less familiar with secular private schools, but I imagine the same dynamic exists. If you're sending them to Braxley-Sheaton-Hogwarts because you want them to have the best experience possible, then they'll probably be fine (but they would probably be fine in a government school too). If you're sending them there to try to ensure they carry on the family legacy and make you proud, then they probably won't.
I'm just here to say that school is still nothing more than Child Prison to me.
Unschooling is necessarily going to be hard to characterize because it's more about something that is NOT happening (schooling).
In future analyses, can you put confidence intervals around your means (or give some other measure of spread like standard error)? It’s really easy to do, and it’s hard to eyeball-interpret the differences between means without them.
I'm a little surprised at the surprise of how religious home-schoolers are. I consider myself religious, attending a church in my community regularly. But there is a group that is far more religious than my congregation on average, and that is the subset of the community that embraces home schooling. I think they are mostly trying to avoid the 'evil influences' of public education or something to that effect. And I think this is at least common across the whole region where I live. Rather than surprise, I would hypothesize that home-schoolers across the nation are even more religious than the home-schooled ACT readers, and yes, more religious than the average religious school attendees. Keep in mind that many religious schools are Catholic. I'm not trying to disparage the Catholic faith, but it strikes me as one of the least religious religions.
Devil's advocate position on my own subgroup: I wonder if homeschoolers (not specifically those in your sample, but all of us) might not be slightly conditioned to answer higher on "are you happy, was homeschooling good" type surveys. There's a thing where a lot of people will ask you those questions in an adversarial way - i.e. "Isn't it hard to be homeschooled, considering you are a friendless loser freak?" or "What do you think about homeschooling your kids, considering it will ruin them forever and ever?"
When that happens, I think one of the normal human reactions to that is to be like "No, it's great, in pretty much every way", because if you say admit to any flaws at all in any tone at all, sometimes the reaction you get is "See, I knew it was a horrible practice in every way and that I can now smugly look down on you for it".
I've been trying to figure out a way this kind of effect could even be quantified, if it exists. Were gay marriages more likely to report happiness shortly after national legalization because they knew it was a politically relevant question? What about transitioners? What about the religious in general?
Thanks for this, Scott! Quick question about the original survey - was there any way to control for responders who fit in multiple schooling categories (e.g., public school for awhile, then homeschooled)?
In my experience (which may not be generalizable), lots of homeschooling families start off sending their kids to public, get disillusioned, then try homeschooling, realize it's too hard as the kid hits the upper grades, and then fork over the money for private or religious school. This lines up with my own background as well: public school until I had a major medical condition, homeschooling during the medical condition and for a couple years after, then religious school for grades 9-12. How would a person like me get diced up in the survey? :)
> Wow! Home schooling is even more religious than religious schools!
I wonder how this correlates with age, but it doesn't surprise me. Most Catholic schools, in my experience, do an awful lot to undermine the faith -- teaching it badly, hiring teachers or ultra-heterodox nuns that vocally dissent from the faith on nearly every core belief (the real presence, abortion, contraception, the incarnation, the immaculate conception, Jesus being male, extra ecclesiam nulla salus, etc.). I go to a traditionalist parish, and a lot of families here homeschool because the religious instruction and formation is either lacking or wrong in most of the local Catholic schools.
I'm applying for the position of village idiot at ACX. I don't recall exactly but my verbal SAT was something like 600. Be kind to the village idiot.
To publish this results without publishing standard deviations, should be embarrassing to someone who is attempting to be serious and claiming a "rationalist" moniker.
You were surprised that home schoolers are even more religious than people attending private religious school?
Maybe it's because my son is in catholic school and maybe it's because I know several people who home school their kids - but i knew this from experience.
People who don't like the libs and are vaguely religious send their kids to religious private school.
People who are super duper committed to their religion home school their kids.
Unschool responder here! One confounding factor is my very religious parents would have said homeschooling. I started in 6th grade because the public school was ‘too worldly’ and my small town didn’t really have other options. My mom had enthusiasm for the task at the start but that very quickly dissolved into an obsession with tennis and other personal interests. I was vaguely encouraged to read, but it was really my choice- assuming I stuck to Christian books. There was no curriculum. There were no taught lessons. She did buy me math textbooks. I devoured them and even gave myself math tests which I graded myself. I think this all says a lot more about me than it does the quality of the education. For science I was once supplied with an anti evolution trac. I don’t think it is fair to homeschoolers to call what i experienced homeschooling, but this isn’t a regulated industry. Quality varies widely and there may be less of a distinction than you are thinking. And, as for satisfaction with their education, the less satisfied homeschoolers are perhaps, like me, inclined to label it unschooling.
This makes me (very very slightly due to not thinking these results say much of anything at all) feel better about sending my daughter to public school. Aside from things that could easily be the result of confounders, there's not enough difference here to make me think schooling type has huge implications on life outcomes, at least for folks who read this blog.
This was really interesting - thankyou! As the parent of an 18 year old boy and 15 old girl, I have asked myself many of these questions - especially since the covid lock-downs. Your findings make me feel like maybe I HAVEN'T made any huge mistakes in their educations thus far.
I had some fun playing with the data; a couple general points:
- agree that you generally want to work with interaction variables of childhood SES x school type
- I would be skeptical of applying many results from the overall survey data to girls and to average/below average test scorers because the sample is so skewed
Result #1: American men age 30 or older whose childhood was "middle class" were less likely to currently describe themselves as "rich" or "upper middle class" if they were homeschooled than if they went to public school (p=0.013). Noting that the coefficient for private school was positive but insignificant (p=0.140).
Result #2: American men who had indicated their education was complete were more likely to have an advanced degree (mostly thinking of this as another way of asking "how much did you like/excel at school?" and the high SAT score of the sample and similar average across modes of education might actually be a plus here) if their childhood was one of the richer-than-middle-class categories, and less likely if their childhood was one of the poorer-than-middle-class categories. The only school type or interaction variable that was significant was children from a poorer-than-middle-class background x homeschooled being particularly less likely to have an advanced degree (p=.021). Meanwhile, while "private x poorer" and "religious x poorer" weren't statistically significant, their coefficients were positive and of greater magnitude than the negative coefficient for the poorer-than-middle-class variable alone. Given how underpowered these groups are, it supports a theory that smart kids from poor backgrounds like school better in those two environments than in public school.
Might try to think of more things to look at later.
Speaking for myself, my son is autistic and has ADHD; we made the large financial sacrifice to send him to private school, because he was falling apart academically and socially in middle school. You definitely need to include learning differences and social dysfunction in the questionnaire.
Calling out once again that your survey didn’t ask people where they grew up or did most of their schooling, it asked which country they most “identify with”. (Whatever that means.)
There’s likely some fraction of people who answered “United States” for the “country” question, but are immigrants who grew up somewhere else. The design of your survey makes it impossible to tell how big that group of people are and might lead to inaccuracies in generalizing about the US (or any particular country).
If you say the non-US numbers aren’t too different here, then maybe it didn’t affect this specific analysis too much, but I’d be on guard for this general class of errors.
On homeschooling and religiosity: one obvious factor:
If you are religious family, homeschooling allows you the scheduling flexibility to attend more than one church service per week-- this is a bigger thing for the more "high church" denominations that are likely to have morning services on weekdays, when most kids are in school. IIRC there is a strong correlation between number of services attended per week as a kid, and religiosity as an adult. Our own church surveys indicate that kids who only attended Sunday services (i.e. once a week) have a lousy retention rate as adults-- once a week isn't enough.
One of the supposed benefits of unschooling is that it can help people develop greater creativity. It would be interesting to look for this in the data. Is there anything in the data that could serve as a proxy for creativity?
Biggest obvious confounded not mentioned: having parent(s) who want their child in private/homeschool/unschooling
That’s going to correlate with a ton of upstream features of the parents (probably reducing to genes and small environmental factors)
Or in local terms: my priors on any of these outcomes being impacted by schooling type vs genes are unmoved and still in the 99% genetic factors camp.
I would be interested to see results on single-sex vs coed schools, although I know that single-sex schools are very rare in the US.
Maybe this is a little off topic, but it surprises me how few second generation homeschoolers there are.
I would expect to see it a lot, but I am the only homeschooling parent who was homeschooled in the homeschool social groups my kids are a part of.
Regarding the high religiosity of home-schooled subscribers, this should not be surprise, because the home-schooling movement was strongly pushed by fundamentalist Christians.
The following story is worth approximately 0¢, statistically, but I find it amusing, so I'll tell it anyway. The son of a friend of mine was kicked out of HS for being a disruptive cut-up. He also committed pranks verging on the criminal toward a teacher he hated. He was taken to court and the settlement required the family to go to counseling for 6 months. My friend thought this was a draconian sentence, but afterwards, when I asked him how it went, he said that actually, it went well. One outcome was that his son, an artist who stayed up till 4am painting, should be home-schooled. His parents (who are ardent atheists, though they love to do a traditional Xmas, in which all the tree ornaments are hand-made by the family), let their son live on whatever schedule he liked, but they did give him assignments in the required state subjects. He applied to a very prestigious art academy and got 800s on his SATs. He was admitted and got to every class on time, whereas many of his fellow students stayed up all night and didn't necessarily go to class. I asked him about that and he said something like "Well, I already did that, but now I'm at a place I love, doing what I love."
I noticed that some respondents put in their full SAT scores (e.g. 1550) instead of the subsection scores when answering the survey. How did you exclude these when taking the average scores for the verbal and math questions?
I am not sure the difference between home schooling vs. unschooling is robust - i.e. that people
who studied at home interpreted the question uniformly enough to answer one or the other according to the same criteria. In other words, two people who received the same home education may have given different answers to this one. How do you account for that?
It would be worth running some of these controlling for social class. For instance, none of the rich either home schooled or unschooled and none of the upper class unschooled, so if richer people are more satisfied with life or more likely to be married or ... that could make home schooled and unschooled look worse on those measures even if, controlling for social cost, they were better.
(( Subscription settings question ))
How do I stay subscribed but turn off emails? Other substacks we're easy to config, with this one I can't figure it out. Thanks
A lot of people who enroll their kids in religious private schools aren't particularly religious, they just don't want their kids being beaten up.
I wonder what results you would get if you dis-aggregated public schools based on whether the school was a magnet school, or some other measure of school quality.
I was told by a psych PhD who studies giftedness that this school is easily the best school available for profoundly gifted kids: https://www.davidsonacademy.unr.edu/ They call themselves a public school, but they hardly seem like a typical public school (e.g. "Our accredited classes group students by ability rather than age", "Our unique approach to learning has attracted families from across the country and globe")
Hey @Scott, I was confused about one thing in the Survey. I'm retired, which I said early on. But then, much later, you asked things about "working" It wasn't clear to me whether I should answer that or not, and I I think I did answer.
Though retired, I do things that I consider "work", but not for remuneration. For example, one year, I was Treasurer for a non-profit org, and that was definitely work. But I also have some individual projects that I work on.
So I wish you had defined work more explicitly; For example, "activities associated with a job for which you are paid", if that's the definition you meant, whjch would have led me to pass up the question I referred to above.
It was unclear to me
Were the SAT averages calculated using corrections for the time period they were presumably taken, at least to address the major shifts? If so, what equivalence tables were used?
This seems like it adds up to weak argument for sending your kids to public school. All the effects are basically negligible, private school is expensive, and homeschooling is a ton of work.
My barely middle-class family scraped enough cash together to keep us in Catholic schools in the 1950s. We weren't particularly religious, but for some reason our folks decided parochial schools were better. The nuns were effective and exacting, and that ten years of education was one of the greatest gifts our parents gave us. (Serious religious women ran the schools; priests or brothers were rarely seen.) We didn't know what 'diversity', 'inclusion', or 'integration' was; classes were always a broad mix of 'races' or cultures with a smattering of Protestants, Jews, or nonreligious students.
Some non-denominational, evangelical schools seem to have adopted the more rigorous academic aspects of the old Catholic schools, but infused them with the more populist doctrines of American evangelical Christianity and appear to be doing quite well. But they are expensive, certainly unavailable to most households making less than six figures.
Yet, regardless of current arguments regarding rigor, regardless of attempts to inject white guilt and gender ideology into curricula, our town still has an Immaculate Conception School still run by nuns still proud to send most of its graduates to college.
To me the obvious headline finding is theism.
Committed theists:
Public: 8.4%
Home: 28.8%
Religious: 20.1%
Private: 6.4%
Unschooling: 5.7%
This actually remarkable how much a difference schooling makes. Surely somewhere between 60 and 70% of public school people (far more than 8%) were raised with a household religious belief, yet for public school belief waned to 8.4%. Compare to home and religious school where one gets much better "retention" as it were. This seems very important for thinking about things like ideology in schools and whether it matters. In the religious context it does. Contra Caplan et al. I'd surprised if ideology in schools didn't matter.
The results indicate a strong positive for public schools because many of the factors associated with poor outcomes end up relegated to the public school group: working single mothers, messy divorces, financial insecurity, children raised by foster parents, parents who don't think much about education.
Meanwhile anyone who homeschools their kids is by definition putting enormous resources and effort into their children's education. The homeschool group will be composed of stable two parent households who are making huge sacrifices in order to do what they think is best for their children. Of course private school families are dumping a lot of money into education, so they have means and value their kids' learning.
Seeds of Science (ACX grant winner), a journal publishing speculative and non-traditional scientific articles, would like to offer itself as a peer-reviewed publishing platform for any analyses using the ACX reader survey data. Visit the website (theseedsofscience.org to learn more or contact us at info@theseedsofscience.org.
This is great
Funny thing is the missing government-run/funded religious schools. Like the Catholic school system in Canada.
The unschooling results make sense to me; a decent proportion of the unschooling families I've known over the years appear to have opted for unschooling simply as a term to justify their severe educational neglect of their kids, often accompanied by other types of neglect as well.
Sorting that out from people who have ideological convictions about unschooling and do provide learning opportunities to their kids consistent w/the kids interests and needs would be complicatey.
My three children all unschool. It’s a second generation thing for them. I went to school till fifth-grade, homeschooled with a curriculum in sixth-grade, then unschooled after that.
Unschooling for me is definitely a way of life, and one that puts us far outside the societal norm. We’ve made seven moves since we’ve had kids, three times across state-lines, and not having to worry about “finding a good school” is a tremendous advantage to mobility. Even on a day-to-day basis, not having the rhythms of our family life determined by school schedules gives tremendous freedom to engage in the world. It’s very easy to go see a super-bloom, stay up late to look at the stars, take time when grandparents visit, or whatever else.
Our kids all take their education seriously. For what would be high-school, our oldest daughter volunteered in a university lab as a worker. She helped on several projects, ran equipment, performed bench work, and was even named as (an admittedly minor) co-author on a published paper. She’s now a sophomore at a UC school and began working in a lab her first term of college. Our two younger have been at least as aggressive in their own chosen domains.
While we definitely let the kids follow their own interests, we do insist on basic literacy and numeracy. While it hasn’t been a problem for us, were our kids not to be able to read or multiply and divide I’m sure we would have insisted they learn. But as I said, it just wasn’t an issue. My position on this is probably guided by my own family experience. My father has a B.S. in math and never uses it. My brother, who also unschooled, got into college not knowing much more than the most basic pre-algebra. But there, he fell in love with math, got a B.S. in math, and uses higher math on a daily basis.
While we mostly let our kids follow their interests, we do not desert them to figure things out themselves. And sometimes we do push them in certain directions. For instance my 11yo son wants to be an engineer and he’s making games in Unity. Earlier this month, and over his objections, I insisted he start using Co-pilot or Ghostwriter.
As i said in the beginning, our way of life puts us far outside the norm in ways that can be dangerous. For instance, about a decade ago, when we moved to a new state, we put the kids in school as an experiment. My wife had found a small horse-farm on the edge of town, and the kids were volunteering there a couple days a week to learn to care for the animals. We’d also take days to explore the new town, hike, visit museums etc., without too much regard to the school schedule. Within less than a month we got a call from the school asking about our kids unexcused absences. We explained the absences were perfectly fine, and that the public school was just one of several experiences we were affording our children, that we weren’t concerned with grades or any particular curriculum. The school told us if we had a handful more absences (I believe it was three) they’d need to call child protective services. We, of course, pulled the kids from school immediately to avoid the danger.
The point is, our natural way of living can very easily lead to severe consequences. For that reason, attending community college or college is important for the kids. Not so much for the actual content of the classes, but for the training in how to navigate and avoid conflict with the bureaucracies of our society.
People are frequently concerned about the social costs of homeschooling or unschooling. Certainly, it is more difficult as a non-schooler to meet kids your own age. And most of those kids, being schoolers, have their lives and realities dominated by a bureaucracy that is entirely alien to the non-schooler. On the other hand, the schooled environment is traumatic for many. And, as a non-schooler, i feel there is something abusive in sorting children into tightly-bound age cohorts, forcing them to sit still in rows of desks all day, at the mercy of an authority figure they’ve had no part in choosing, and to whom they must petition for even the right to use the bathroom.
But even if I were to accept for the sake of argument that the schooled environment is a healthy one for socialization, between sports, online communities, and homeschool groups, there are plenty of opportunities to socialize. The main social problem I see is finding romantic partners of the appropriate age. There’s really no good way to reliably meet your high-school aged crushes or boyfriend(s) as a non-schooler. So that mostly has to wait till college or the workplace.
That said I did have one girlfriend when high-school aged, don’t feel in retrospect I had more than the normal amount of awkwardness in my brief career of college dating, and ended up meeting my wife in my junior year. And, as a mid-forties man, I don’t feel the lack of socialization in public school kept me from developing personal friends or professional relationships.
A final thought on “unschoolers have low self-reported life-satisfaction.” Putting aside the small sample size and overall tightness of the reported values, I doubt the question compares apples-to-apples. Life satisfaction is essentially a measure of existential angst, and the unschooler’s existential angst is very different from the schooler’s. The unschooler is always forced to figure out for themselves what they want from life, whereas the schooler can allow themselves to be led by their peers, the system, or authorities. And so, especially at the younger ages answering the survey, it seems to me the schooled may be more likely to feel satisfied with their life because of external affirmations - graduating college, having a job, etc. Whereas the unschooler is more likely to judge their progress by their own standards. These seem like different enough yardsticks to me that we cannot reliably compare the readings between the two.
And a final finally, lest I’ve seemed to hate on public education or educators, I certainly recognize the tremendous social benefit the public school system offers in its breakfast and lunch programs. Many otherwise abandoned children obtain their only physical and emotional sustenance from the adults running their schools. Unschooling worked well for me and mine, but I really won’t confidently generalize its utility for children who are not natively curious, or lack parents who are devotedly solicitous of their care.
Many of these comparisons scream for a regression, instead of other adjustments. I feel bad saying this when I don't have time to run them myself and publish the results, but just noting it would make things more reliable and easier to communicate.
Also, should turn some of these tables (especially the 2d ones) into charts.