This article on home schooling and studies thereof is rather interesting. A few comments:
The author tentatively suggests that perhaps a weakness of home schooling is that it doesn't sufficiently prepare students for the "soul-destroying" regimentation of life outside the home. She specifies two areas where this might be the case--deadlines and pointless (or apparently pointless) rules.
I would say that these are different. On the one hand, if parents are not forcing their children to meet deadlines, that's a problem. They should. I always point this out as a caution to parents who are about to start home schooling: Be sure that your child has projects that he has to turn in by a particular time. You might do a lot more of this in high school than in grade school. That's entirely appropriate. But make sure that he does it. Home schooling needn't mean schooling without any discipline. The same goes for making your students finish tests on time. Force them to learn how to do timed testing, even if this means having them do some standardized testing (or preparation using prep books) when this isn't strictly necessary. It will be good for them.
Pointless or apparently pointless rules are a bit different and do present a conundrum. The dilemma is this: You, as a parent, want to be a good boss, a good ruler. Good bosses or rulers don't make up stupid, counter-productive, or pointless rules just to haze their employees or subjects. (At least, I would argue so.) But in the world your kids will sometimes have to work for people who do demand a kind of mindless conformity. Do you need to pretend to be such a boss in order to prepare your students to work for such a boss?
I could not bring myself to do that, because it would seem to me to be wrong. My entire goal is to be as Solon-like as possible in my interactions with my children, including my interactions with my children as my students. I can't set that aside and deliberately be a jerk or a tyrant for a few days (or weeks, or months) just to teach them to work for jerks or tyrants.
But by the same token, I'm disinclined to say that it would be better for them to be in a school environment where they have to obey rules that I truly think to be poor rules just to "learn to do it."
Now, of course, there's a big grey area where rules might be poor in a home environment but understandable in a school environment. For example, at home you let your children get their own meals, but at school they aren't going to walk into the school kitchen and scrounge up a lunch. There's nothing at all wrong with school uniforms, in fact they can be a good idea, but it would be silly to make your children wear them at home just to get used to the idea.
It may indeed be the case that home schooled children take less well to regimentation when they graduate than school-schooled children. I'm not entirely sure this is a bad thing. Some people are indeed going to be suited better by environment and training for regimented positions than other people. Therefore, different groups will supply the world with different types of workers. Moreover, considering the anarchic state of too many public schools these days, the peaceful "lack of regimentation" in a home schooled environment seems extremely orderly and disciplined by contrast.
It's not a bad idea for home schooling parents to be reminded to subject their children to discipline in their work--to deadlines, high or even nit-picky academic standards, and timers. We shouldn't turn into functional un-schoolers. But as for regimentation more generally, I'm pretty comfortable with letting someone else supply the world with young people who submit to it tamely and instinctively.
A few comments on studies on home schooling: Sometimes it's argued that such studies show nothing of value because they don't include a random cross-section of students. This is of course the nature of the beast. You can't design a study by randomly assigning some cross-section of American (say) children to be home schooled, simply forcing their parents to home school them regardless of their own inclinations, situation, or capability. But we could argue that by the same token, neither can we (nor should we) force some randomly selected cross-section of parents to send their children to public school or private Catholic school or any other school for the sake of science. By the very nature of the case, the type of schooling a child is already receiving will have to some extent been selected by the parents. The fact that public schooling is (perceived as) free and that many parents treat it as a default setting does give us bigger and broader samples of American children in the public schools, but it is still by no means a full-scale cross-section of American children. (As witnessed by the fact that many high-income parents who are no friends to home schooling nonetheless send their kids to private schools.)
I fully admit: Studies of home schooling aren't going to be studies of "what any kid would be like if he were home schooled," because not any kid can be home schooled. Home schooling requires at least one available and willing parent, which many children don't have.
But it doesn't follow that studies of home schooled children tell us nothing. For one thing, many of the people who currently home school were hesitant about doing it originally or were unsure that they could do so. To the extent that their children's outcomes, or the outcomes of children in family situations in some sense "like theirs," are measured by studies of home schoolers, they may find their decision vindicated. This point extends to people currently considering home schooling. In particular, studies that seem to show that the educational level of parents is less important than might have been believed to the child's educational outcomes (this is mentioned in the article) are rather surprising and should be noted by parents considering home schooling. In fact, parents with lower incomes or lower educational levels who home school their children appear to get better outcomes than parents with comparable income levels or educational levels who send them to public school. If this finding is in fact well-supported (I don't have time to dig into the original research and give it an independent evaluation), it is extremely important. Sometimes someone will learn that I have a PhD and home school and will then make a disparaging comment about home schooling that exempts me. "Oh, well, it's okay for you. You and your husband both have PhDs, but what about other people?" My own anecdotal experience is that this is misguided. A PhD is by no means necessary to be a good home schooler. For that matter, considering the way that college education is going today, a BA is not clearly necessary either, though it might help to keep busy-bodies off your case to have one. It is especially true that home schooled children who grow up and home school their own children may well have received a high school education equivalent to or better than many college educations and may therefore be well equipped to home school without a BA. The snobbery of degrees should be shelved, and if well-designed social science studies can help to do that, all the better.