Before this is about your group chat, it's about a psychology lab in Ohio, because the single most useful fact about how women report their sex lives was established there two decades ago, with a prop.
In 2003, researchers Michele Alexander and Terri Fisher had college students fill out questionnaires about their sexual attitudes and behavior under three conditions. Some believed a researcher might see their answers. Some answered anonymously. And some were hooked up to what they were told was a polygraph, a machine that could catch them lying. The machine was fake. The results were not. When participants thought someone might read their answers, the reported gap between men's and women's sexual behavior was at its widest, with women's numbers dropping on exactly the behaviors considered least acceptable for women. Under the fake lie detector, when lying seemed pointless, the gap between the sexes nearly vanished.
Read that again, because it is the whole article in one finding. The behavior didn't change between conditions. Only the audience did. The number these women reported was not a fact about their bodies. It was a fact about who was in the room.
Your group chat, in research terms, is an exposure threat condition. Everyone in it will read your answer, remember your answer, and reference your answer at a bachelorette in 2028. Of course nobody's filing accurate numbers in there. The thread is the least anonymous survey instrument ever designed, administered over mimosas, with a permanent written record. Alexander and Fisher would have thrown it out of the study.
The Reporting Conditions at Brunch
So the testimony adjusts, and it adjusts along lines any researcher could predict. The married contingent inflates, because somewhere along the way frequency got installed as the marriage's public vital sign, and nobody reports a weak pulse to a panel that will worry. "A few times a week" is the standard filing, and it is doing heroic work. The single contingent adjusts directionally by audience: the same quarter gets rounded down for the marrieds, who must not think it's chaos over here, and restated upward for the other single women, in whose company a drought reads as a defect. Same woman, same three months, two different earnings reports, and neither one is exactly a lie. They're audience calibrations, the household version of what the Ohio lab measured with electrodes.
The result is that every woman in the thread is benchmarking her real number against a data set with no honest entry in it. Which would merely be funny, except that it's the precise mechanism by which four separate women in one group chat have all privately concluded that they, specifically, are the one falling behind.
The Audited Numbers
Here is what the anonymous data actually says, and if you've been using the group chat as your benchmark, sit down, because you have been comparing yourself to a fiction.
The General Social Survey, the closest thing America has to an honest ledger on this, puts married couples at roughly 54 times a year, about once a week, and that figure has been falling for decades; in 1990 it was 73. The share of married couples reporting sex weekly or more dropped from over half in 1990 to under 40 percent in the most recent data. Demographers have been calling the broader trend the "sex recession" since the late 2010s, and it spans every group: married, single, young, old, religious, secular. Whatever number you've been carrying around as "normal," it was probably calibrated to a decade that no longer exists.
It goes further. By the standard research definition of a sexless marriage, ten times a year or fewer, some 15 to 20 percent of American married couples qualify. Not new marriages in crisis. Marriages, generally, at a rate approaching one in five, which means the group chat statistically contains at least one, and you would never in a hundred years guess which one, because her filings are immaculate. And roughly a quarter of married couples report their frequency runs below what they'd prefer, which is perhaps the most humanizing statistic in the whole literature: an enormous share of the population is quietly wishing for more and reporting more than they're having, simultaneously, to different audiences.
Then there's the finding that should end the competition outright. A study of more than 30,000 people, run by researcher Amy Muise and colleagues, looked at the relationship between frequency and happiness and found that wellbeing rises with frequency up to about once a week, and then flattens. Past weekly, more sex does not correlate with happier couples. The couples posting the numbers your thread pretends to are not, measurably, happier than the couples who'd be embarrassed to say theirs out loud. The scoreboard everyone is quietly losing on doesn't even track a real prize.
What the Number Was Never Measuring
Put the lab study and the survey data side by side and something clarifying happens. The science says the reported number moves with the audience. The ledger says the real numbers are modest, declining, and clustered tightly enough that your entire friend group is probably within a rounding error of one another. Which raises the only question worth asking: if the number is this unreliable as testimony and this unremarkable as data, what exactly has it been doing with all that power?
It's been standing in. Frequency is easy to count, so it gets deputized to measure things that are hard to count: whether you're still wanted, whether the marriage is alive, whether you're doing this era of your life correctly. But a count can't hold any of that. It has no column for wanting, which is the actual vital sign. It can't distinguish the couple keeping a streak alive out of duty from the couple whose rare nights are the best thing in the building. It doesn't adjust for the newborn, the trial, the medication, or the grief. As evidence of anything beyond itself, the number was always inadmissible. The lab proved people won't report it straight; the surveys prove it doesn't predict happiness anyway. It's a bad witness testifying to the wrong question.
So the next time the subject circulates, at the dinner, in the thread, you now know what you're watching: an exposure threat condition producing exactly the data such conditions produce. You can enjoy it as theater, because it is, performed by women you love for reasons the literature has documented since 2003. What you can no longer do, in good conscience, is enter your real life into evidence against it.
The researchers needed a fake polygraph to get the truth. You don't. You just need to notice that the only audience whose reading of your number ever mattered is the person you share it with, and possibly not even them, since the one statistic that actually tracks happiness isn't how often. It's whether the two people involved are satisfied, a figure that has never once come up at brunch, because it can't be ranked.
Your number is fine. It was always fine. It just kept getting read in the wrong room.
