This kind of info is not actually useful when weighing it up. The homeschooling is only as good as you want to make it and are capable of making it. You can te

This kind of info is not actually useful when weighing it up. The homeschooling is only as good as you want to make it and are capable of making it. You can teach math and care about it? Might go great for your kids. You can't add but think it's ok if you just pray lots? Not so much. So many hard-to-identify confounding factors. Randomised trials these are not. The correlations here just don't inform your decision. You want something like: Probability of some particular kind of success given your reasons for doing it, resources and ability. This is an elusive thing to estimate even if the data were very much higher quality. The other thing worth noting about these studies is they're old. 2007 is /ancient/ compared to the resources you have at your disposal now. The Khan Academy problem bank alone is likely to assist homeshcooling math score averages significantly. That one thing alone is a hell of a resource for homeschool teachers to learn what they need and for students. There are a stack of others too. Darwinian selection will likely pick a few winners that are just great too. I have only pandemic homeschooled children. I have some reasons you don't care about for not doing it. People considering going that way can probably take a mountain of salt to the claim "Homeschooling is really bad for STEM..." made by username90. When dudul points out "The data ... [is] not super compelling" That seems like a very fair observation to me. But nobody here is getting papers published, citatation metrics enhanced and so on. In the absence of clear and compelling data, your prejudice about decisions concerning your family are better than those of some random stranger regardless of the robes of authority they wear.