I worked in an MRI lab (academic) for a few years (as a tech/programmer/statistician) and honestly these kind of sex difference findings were really common. We

I worked in an MRI lab (academic) for a few years (as a tech/programmer/statistician) and honestly these kind of sex difference findings were really common. We never did fetal scans, but we were also studying resting state fMRIs connectivity in adults. I'm semi ashamed to admit it, but a lot of times, if your main hypothesis didn't come to fruition in the data, you could usually just fall back on some sex-difference finding. You wouldn't even need to fish very hard for it. There is probably no other single 2-way split you could make that would more reliable give you large magnitude differences. I don't know how else to present that idea though. It's not completely unreasonable that there might be some inherent bias in the processing. There are really a handful of tools that everyone uses. It's possible that there is some undiscovered bias somewhere. But if that is not the case, then I am pretty convinced from my time there that the differences are real, and they are usually of large magnitude. At least larger than the more interesting findings researchers look for. Pretty crazy that they did fetal scans though. We had a lot of trouble normalizing even toddler brains on one study. I can't even imagine fetal ones.