So I found this study called "The length of human pregnancy as calculated by ultrasonographic measurement of the fetal biparietal diameter" (H. Kieler, et al) and on page 2 there was a treasure trove of studies and the various average gestation lengths they found, WITH study size! It included the Swedish study I'd found of over 400,000 women, they found the mean to be 281 days. It included a 1962 study of 1,970 uncomplicated spontaneous births, they found the mean to be 279 days.
And then there was the Mittendorf study that everyone LOVES to link to the abstract was because they found the median to be 288 days for first time moms... 41 W, 1 D. I always thought that made no sense for a median and wondered how they found it. Well it turns out their case study size was 114 women.
Being reminded of it again, I had to do more searching, I had to find more on this study. Eventually I did find a page where I could read the whole thing, FINALLY. And you know what? This study that everyone loves so much consisted of 31 first time moms (primiparas). 31. You can study the entire population of Sweden, or 31.
You know how much shit those hens in pregnant gave me for "ignoring scientific data" when I made a chart showing the average pregnancy ending at 280 days?
And that's not the only weirdness with The Length of Uncomplicated Human Gestation... they base all their numbers on ovulation, not LMP, even if only cycle length was known. Like, if a woman had a 24 day cycle instead of 28, and gave birth 280 days after her last menstrual period, the study uses that to support adding 268 days to ovulation. And then the internet people are using that to say "Well that's 282 days for a normal woman so 280 days is obviously wrong."
Whatever. I'm still running my survey, and have recently completed all new pages of data analysis using charts generated by the open source jQuery library, flot. The first page of the survey is an introduction that has a list of studies I've come across. I'm still trying to find published data with daily breakdowns of the likelihood of giving birth spontaneously on a given day, but plenty of studies have held up the idea that the average is somewhere between 278 and 282 days. That's 40 weeks.
So let me conclude with this: if you are pregnant for 288 days, over a week past your trusty Naegele's rule due date, you are not "average", okay? You deserve a medal! You are a gestation goddess. Do not let anyone half-read an abstract on the internet and tell you otherwise!