and has had unique visitors.

Wishful Thinking

Here I address the issue of biological differences between men and women and try to provide a simple lesson in what statistical data relating to sex differences mean and do not mean. Let us first look closer at truth and objectivity in searching for truth.

I will tell you a fascinating story. In the year 2005, Lawrence Summers, the president at Harvard University, gave a talk at an academic conference. He came upon the topic of why there are relatively few women at the top positions in science. Summers then did the unthinkable. He speculated that maybe it could be worth considering whether there are innate differences between men and women that make men slightly more likely to be working with science. I have not found the specific words, but he was saying that statistically, it could perhaps be so that men more often were suitable for scientific thinking. He also questioned whether it was really so common that women were actively discriminated at Universities. Summers was very careful when he presented this hypothesis since he knew it was a sensitive subject. He emphasized that this did not in any way mean that women should not be just as welcome to study science as men. His claims were well supported by references and rational thinking. But none of this helped…

The response was a global feministic fury, both because Summers claimed biology does matter and because he questioned whether discrimination was really so common. The worldwide outrage reached all the way to Sweden, where it attracted the attention of our type of feminists. They took it as an illustrative example of how terrible the world is, how women never will be free and all kinds of other things. Summers had to apologize officially, but it was not enough. People wanted more blood and the year after he had to leave his position as president at Harvard.

The psychologist Steven Pinker was in an interview for The Harvard Crimson regarding Summers. The reason was that he had expressed sympathy with the president. In the interview he was asked whether he found what Summers had said to be “offensive”. He replied:
Look, the truth cannot be offensive. Perhaps the hypothesis is wrong, but how would we ever find out whether it is wrong if it is “offensive” even to consider it?
Pinker also participated in a very interesting scientific debate on the subject with Elizabeth Spelke at Harvard.

This story exemplifies my first point regarding statistics. The truth about the differences between men and women is not necessarily what we want it to be. Feminists and gender researchers both think that it would be great if there were no biological differences between men and women. So they choose to believe it. Some time ago I thought people would agree that this is insanity, but now I am not so sure anymore. It seems nowadays one must actually point out that we cannot assume things are true because we wish them to be true. It is against the very nature of scientific thinking to do so. Furthermore, science must be free to test whatever hypotheses, whether they are “offensive” or not.

Another example is the Swedish professor of neurobiology and histology Annica Dahlström (now retired) who knows quite a lot about sex related differences in behavior as well as differences in the architecture of male and female brains. She published a popular science book in Swedish about her research. As a result she was terrorized in various ways, ranging from threatening phone calls to letters containing rat excrements! Her book was perhaps in part a response to the Swedish feminism trying to obscure evidence of any innate biological differences between the sexes. Official authorities had prevented her opinions from being published since they did not agree with the established “Swedish point of view”. (We are fortunate to have a government that decides what is true and false, soon we can ignore doing research altogether.)

Another academic in Sweden who has taken part in the debate is Germund Hesslow, professor of neuroscience. In public debates he has pointed out simple scientific facts about biological differences between the sexes. He has also commented on the fact that the wage differences between men and women cannot be explained by discrimination. As a result, people threatened Hesslow in various ways. Among other things, feministic bloggers wrote that they wanted to kill him and official representatives at Lund University considered banning him from teaching. (So much for academic freedom.)

Of course, all the examples given above illustrate how rage is a primary driving force behind feminism. Also, the feministic dominion has obviously taken over both our government and our universities. However, the point I wish to make here is that the truth is not necessarily what we want it to be. We cannot assume things are true because we wish it to be so. Personally, I do not think it is such a terrible fact that men and women are statistically different. I rather think it makes the world slightly more interesting when there are things we can consider truly female and truly male. But this is another story.

The Importance of Variance

Statistical data on sex differences are often presented in popular media using the simplified form “men are like this” or ”women do like this”. One then speaks of AVERAGE values. Strictly, there is nothing wrong with such statements, but there is additional information which is very important. How significant were the results? How does the study compare with other studies? Such questions relate to how trustworthy the conclusions are. Another very important parameter which is often ignored is the VARIANCE in the data.

Variance by itself says a lot about what we can expect from the observed differences between sexes. For instance, if we note that women on average have better sensitivity in their fingertips it could indicate that overall there will be more women doing knitting. However, it should also be noted that if women have higher variation in the fingertip sensitivity, this alone implies that the best knitters in the world are likely female. Of course, other factors play in as well when it comes to who likes knitting and my only point here is that variance is important to consider in general.

The figure below shows an example where variance is, in some sense, MORE important than the average value. Assume we want to find the persons in our society who smoke the most cigarettes. This is just an example I made up, but say we look at available statistical data and see that on average, men smoke more cigarettes than women. One would then guess that we should search among men to find the greatest smokers. However, we also see in the data that the variance among women is much greater than for men. The difference in variance between sexes is very large while the difference in average is relatively small and in the opposite direction. As a result, the persons that smoke the most are female as indicated by the statistical distributions.

Illustration of how variance can be more important than the average value. We want to find the individuals with the highest value of a certain property. The statistical data tells us that on average, men have a much higher value of the measured property. However, since the variance for women is higher than for men, we still find mostly women at the high end of the scale in this case.

Indeed, among humans there is always great variation almost regardless of what property we measure. In other words, as soon as we meet a person, we should not presume to know anything about the behavior or preferences of this individual just because it is a man or a woman. Prejudice based on sex is thus wrong both from a moral and a scientific standpoint. When it comes to men and women, a common expression is that the differences between the sexes are smaller than the differences within each sex. This may be true when comparing standard deviations and average values, but unfortunately people seem to take this as evidence for the differences being insignificant. This is simply not correct! No matter how large the variance is, the difference in average value is still there! It simply means that we may need to collect a large set of data to clearly observe them. Indeed, this is what we see in meta analysis of large amounts of scientific data.

Show the Evidence!

So what kind of evidence do we have that there actually are biological differences between the sexes when it comes to other things than physical appearance? First of all, I would like to make it clear that there is nothing strange about this. Our behavior and mental abilities are not independent of the physical world. Let me give a simple example: If I hit you hard in the head, your behavior will change. We also know that behavior changes as a result of e.g. drugs. Add to this the fact that we know men and women are different biologically. We have different brains, different genes in action inside our bodies and so on. Since we are differently composed biochemically, It is EXPECTED that we behave different.

Today we have a very great amount of statistical data available where we can identify all sorts of differences between men and women. Men more often have homepages. Men kill other people more often. Women are more careful when driving a car. Women buy more clothes. The list of things that separates the sexes in terms of behavior is practically endless. Now my question is: How can it be reasonable to assume that biology has nothing to do with this? Still, this is the trend we see today. It is generally accepted, all the way up to the Swedish government, that male and female behavior is only about social factors.

It is of course difficult to prove that a certain type of behavior is purely biological or purely social. There are usually reasons why both factions can provide reasonable explanations. This is exactly why I emphasize that we cannot ignore any of them! Biology and sociology both have their part in explaining male and female behavior. Often neither can provide a full explanation since we see the product of them both. It is also important to note that neither explanation is deterministic, especially so for biology. If a type of behavior has its roots in biology, it does not mean we cannot change it by social influence.

In some cases it really can be shown beyond reasonable doubt that it is biology which plays the major role. I will give two examples. The first is infant behavior. Newborn baby boys behave differently than baby girls in some respects and it obviously does not make sense to explain this by referring to environmental influence. The second example is hormones. We know that we can change the behavior of both men and women in certain ways by giving them hormones. Since we also know that men and women have different hormone levels naturally, it follows that they will have innate differences in behavior.

Interpretation of Results

Unfortunately, a lot of people seem to think that if one believes men and women have innate biological difference, one must also believe that men and women should somehow not have the same rights in our society. For instance, if I say I think women are generally better at handling small children, people accuse me for wanting to stop men from going on parental leave. This is wrong in several ways, in part because of the importance of variance as explained above, but primarily because we cannot derive moral truths from scientific observations in such a direct manner. In the same way, we all agree that it is wrong to kill handicapped people, even though the theory of evolution say they would not have survived the natural selection.

It is exactly in the type of statistical data used in the debate about feminism, coming from large populations, where we can expect to see a manifestation of the natural differences between men and women. Still, it is these very same data that feminists are upset about. Typical examples are the number of women at top ranked positions in society and the differences in income. There are perfectly natural explanations for these differences and they do not in any way prove discrimination. It does not matter if we use biology or social factors to explain them. In fact, when one takes into account other factors, typically those related to the types of choices in life that men and women often make, the SAME data indicates that discrimination related to salary does NOT occur to any significant extent. Check out the book "Why Men Earn More" by Warren Farrell.

This relates to a common flaw in the interpretation of statistics. If A correlates with B, it does not mean that A causes B. For instance, I think one could find a correlation between people dying in heart related disease and whether or not they have a football in their home. Let us say those that have a football in their basement die less often from heart attacks. No one in their right mind would say that the football lying down in the basement cures their medical condition. The explanation is of course that the football indicates that those people do sports and get more physical exercise, which we all agree that it does help preventing heart attacks in the long run. This kind of misunderstanding is illustrated in the figure below.

A common misunderstanding in statistics. We observe that A and B are correlated (blue arrow), but this does not mean that A causes B. In fact, there may be something else which follows from A, that causes B. This fact is here denoted C (red arrows indicate causality).

Greater than Zero

I wish to emphasize again that I believe social factors are more important than biology. I just want to show that the contribution from biology is greater than zero! As with all other things, assuming extreme standpoints is not very wise. Still, this is what modern feminism does when it rejects anything that indicates sex differences are innate. It suppose it could be that feminists who do believe biology matters are not so few, but it is certainly not the voices of those people that we hear in the debate.

A common counterargument is that there are just as many investigations showing NO correlation between sex and cognition/behavior as those that show correlation. It must then be noted that all studies obviously are different and diverse aspects of behavior are investigated. One cannot rule out biology OVERALL just because one particular study of a certain type of behavior showed no difference. As I am trying to say, biology plays a role for some types of behavior and not for others.