Science Matters

Sex ratios and scientific research

Boyce Rensberger

(1/2024) Science makes progress not by proving that theories are true but by testing them rigorously and being unable to prove them false. That's a crucial distinction to keep in mind when evaluating the claims and counterclaims that are hurled about so often on almost any issue.

An honest scientist, as I’ve written before, will almost always say a new discovery is "probably" true. A phony will usually make claims as if there could be no doubt in the world.

There are very few things in cutting-edge science that can be stated honestly with absolute certainty. I’m making a distinction here between textbook science (the established findings we were taught in school) and cutting-edge science (the ongoing research that scientists are doing now).

The closest that researchers usually get is to express a high degree of confidence that something is true. But they can reach that level of confidence only after rigorously testing the hypothesis. Or if experiments are not possible, as in astronomy, confidence comes only after repeated observations that point to the same conclusion.

Let's study this issue with a deceptively simple example. Let’s say that you, as the scientist, would like to know the chances that any given birth will produce a boy or a girl.

You could simply observe your own family. You might record that it has two girls and one boy. If you’re comfortable with that size of sample, your observation might lead you to propose a hypothesis that nature has a mechanism that dictates that twice as many girls are born as boys. That's fine. It's a perfectly valid hypothesis, based on) real-world observation, though extremely limited.

Should you believe it? Not yet. If you want to reduce your chance of being fooled, you should first follow the standard scientific approach and test your hypothesis.

How? One simple method is to make many observations and examine them as a group.

So, you look at another family and find that it has three boys. The third family has one boy and one girl. Gradually, as you widen your sample, you find the boy-girl ratio approaches 50-50. Then, a strange thing starts to happen—once your sample reaches a certain number of families, the ratio never again varies much. You may feel satisfied that you have looked far enough.

Now you have produced what seems to be a piece of scientific information—the chances of any birth producing a boy are about 50 percent, and the same goes for girls. If you are satisfied with your data, you are entitled to take the next step and ask why the sexes should be evenly matched.

The usual practice in science when faced with a "why" question is to think up a "because" statement. You can search your memory for something that you already know. Or you can make up something. It's okay; scientists make up explanations all the time. That's what a hypothesis is. Scientists just try not to believe them without testing them.

So, you might hypothesize that the sexes are equally matched because everybody needs somebody to love, and a 50-50 ratio makes this possible.

This is the kind of hypothesis that was popular among the ancient Greeks. Objects fell downward, Aristotle taught, because the ground was their rightful place. Since ancient times, science has developed a rather different approach -- testing and careful observation.

In the 1930s, the late philosopher Karl Popper recognized this tradition in science, analyzed its logic and named it "critical rationalism." Popper formalized the idea that a hypothesis can only be proven wrong, never proven right. Therefore, one of the rules of the scientific method is that the only good hypothesis is one that can, in fact, be tested. Our "somebody to love" hypothesis doesn’t seem very testable.

Because textbook science tells us about sex chromosomes, a hypothesis for our birth-ratio research project might be as follows: "The sexes are equally matched because parents can bequeath only one of two combinations of sex-determining chromosomes—XX or XY. And the odds that an embryo gets one or the other must necessarily be 50-50." (I’m glossing over the fact that in very rare cases, malfunctions of the chromosome-sorting mechanism inside cells can produce individuals with XYY, XXY and other combinations.)

We might imagine testing this by looking at other species with the same chromosome situation to see whether they produce half males and half females.

Let’s say it is a year or so later and that we have had a wonderful time on our field trips checking up on the sex ratios of lions and larks, termites and tuna. And we have looked up the reports of other scientists who have studied sex ratios in other species. We would find that not all are 50-50. Some are, but others are not. Our hypothesis needs work. It isn't as simple as we thought. We can't count on the Nobel this year.

In the meantime, one of our graduate students has been examining official records of human births and deaths and has made a surprising finding. If you look at the number of each sex at each age of life--and this is genuine information--the ratios change over time. Among newborns, there are more boys than girls. But male babies are more likely to die than female babies.

As we move through older age groups, at each age more males die than females. The ratio is close to 50-50 during the reproductive years. And among the oldest people, of course, there are many more women than men. This gives us one scientific conclusion: females really are the stronger sex.

This deeper analysis of the data has revealed a phenomenon far more profound than it seemed at the outset. We can now refine our hypothesis: "Nature has somehow arranged things so that the number of males who die before the age of parenthood is exactly offset by the larger number of males conceived and born in the first place."

So, the "somebody to love" hypothesis was not so far off. And rigorous testing of our hypotheses has led us to a far more complex problem. The arithmetic of sex-determining chromosomes would seem to dictate a 50-50 ratio. The fact that nature doesn't work that way shows that a more subtle phenomenon must be at work.

We need another grant.

Read past editions of Science Matters