h1

science (or possibly not-science) in context…

May 27, 2011

When I teach Research Methods, we spend the first week talking science – what is “science”, what distinguishes it from “not science”, how it works (or doesn’t work). A recent controversy captured a few of the points we encounter quite nicely.

First, here is a recent posting (the Google cache actually, as the original post was removed) at Psychology Today. It is written by Dr. Satoshi Kanazawa, an evolutionary psychologist who holds a position at the London School of Economics and Political Science. In the posting, Dr. Kanazawa presents a series of analyses that he uses to support his claim that black women are less attractive than women of other races. As you might imagine, this is a controversial claim. Soon after it was posted, Psychology Today deleted the post , but a petition campaign emerged that is pressuring Psychology Today to both drop Dr. Kanazawa as a contributor and develop policies to avoid this type of posting in the future. That is the short story – it may seem simple on the surface, but this raises some very complex and difficult issues.

The first issue is whether the posting represented actual (good) science…

  • If you took a look at the cached version of the posting, it sure looks like science. There are graphs based on data. It is written by a fellow with a Ph.D.. The dependent measure of attractiveness is “objective”. However, the criticisms that have arisen have pointed out several problems: the measure is actually a subjective rating provided by the person administering the Add Health survey (the source of the data), Dr. Kanazawa only used some of the ratings (apparently of younger participants, not adult participants), and another analysis of the same data found no difference between the rated attractiveness of women of different races. These are all problems that severely limit the claims Dr. Kanazawa is making. The issues related to data analyses can be tricky (as we cover in Research Methods – different ways of summarizing and then analyzing a set of data can lead to different conclusions), but I am appalled that Dr. Kanazawa tries to sell an obviously subjective measure (one person’s rating of another person’s attractiveness) as an objective measure. If all of his analyses were correct, Dr. Kanazawa should really just say that other people think black women are less attractive. The source of this bias can then be examined. Personally, I consider attractiveness to be a social construct, so it is foolish to assert any measure of attractiveness can be objective. We can objectively measure aspects that might be related to attractiveness (e.g. symmetry or the relationships that exist between facial features), but I have never seen a measure of a construct like attractiveness that is free from bias.
  • Is this science? What is the theoretical framework that supports the hypothesis being tested in the study? Oh wait, is there even a hypothesis? Dr. Kanazawa begins his piece with the non-illuminating lead-in, “There are marked race differences in physical attractiveness among women, but not among men.  Why?” After this question, he presents his data. After he presents his evidence to support the difference in attractiveness, he begins conjecture. He decides it has to do with testosterone levels – black women have higher testosterone levels, so they are less feminine is his conclusion. Ok, so we have no hypothesis. A conjecture about what might be the cause of an analysis that is questionable in itself. No independent evidence about hormone levels (remember the only data is about rated attractiveness). This is not good science, and could possibly be classified in this case as pseudo-science – the posting looks like science, has the trappings of science, but is not grounded in the principles that guide science. Dr. Kanazawa has published numerous articles in a number of well-respected journals, he has an advanced degree and an academic position (although it is not in psychology department), but he is not maintaining adequate scientific rigor in this case.

I also was interested in this controversy because Dr. Kanazawa is a well-known evolutionary psychologist, and I have been trying to get a better handle on the scope and usefulness of this approach to understanding human behavior.

  • Dr. Kanazawa operates with a clear theoretical framework in place. He is an evolutionary psychologist that advocates for what has been called the “savannah principle” – human behavior is shaped largely by the experiences of our ancestors in the African Savannah hundreds of thousands of years ago. Ok, that idea has been used to “explain” numerous behaviors such as our desire for fatty foods and why we prefer people who cheer for the same sports teams as we do (in-group bias), but what does it have to do with the study in question? I don’t know. I do think that some of his work has been problematic and suffers from some of weaknesses that plague much of the work being done in evolutionary psychology (for a lucid and interesting critique of evolutionary psychology check out Evolutionary Psychology as Maladapted Psychology by Robert Richardson).

Finally, it is easy to criticize Dr. Kanazawa for publishing this post without proper scientific oversight (e.g. peer review – having other experts evaluate the work before it is made public), but this situation raises some complex and difficult issues.

  • Science does not occur in a vacuum (a point I’ve noted in earlier posts). As mentioned above, there is a petition for Psychology Today to remove Dr. Kanazawa from its contributors. There is also a petition for his removal from the faculty at the London School of Economics and Political Science. This episode is not exclusively an academic debate (although many of Dr. Kanazawa’s critics point to the lack of scientific rigor in this work). Psychology Today is a “respected” (although not my personal opinion) public resource for information about psychological issues, and many of his critics are upset that by allowing the article to be posted, Psychology Today is giving weight and credence to his potentially hurtful conclusions. This is not the first time Dr. Kanazawa has drawn public ire – he has published work proposing that problems faced by African nations reflect the lower intelligence of their populations, and he has another blog post entitled “Are all women essentially prostitutes?” – and he does not shy away from advocating that science has to sometimes ask difficult questions (here is an article entitled “10 politically incorrect truths about human nature” that provides some insight into his views – note that again it is published in Psychology Today, so there is not a fully independent critique and evaluation of the claims he makes). So, should he lose his position and essentially be banished from academia because his ideas have upset a lot of people? I’d say no. Science has to proceed independently from social mandates. However, I strongly believe that his position, and the weight given to his ideas, should be called into question because of the seemingly poor quality of some of his work and his lack of respect for the responsibility he has as a scientist. We deal with this in Research Methods – the interplay between societal concerns and scientific inquiry – and it is not an easy topic to grapple with because there is no easy answer. Science has to be unfettered, but scientists have to be aware of and sensitive to the social and ethical implications of what they do. There should be research exploring issues like intelligence and attractiveness and how groups of people who differ along various dimensions (gender, race, education, socio-economic status) differ (or don’t differ) with regard to those constructs. However, because these topics can have very real implications for how we think about ourselves and each other, they have to be studied with extra rigor and thoroughness. Poorly executed analyses with questionable conclusions being posted to Psychology Today blog is exactly the wrong thing to do.

I think it is interesting that one of Dr. Kanazawa’s first publications [Kanazawa, S. (1992) Outcome or expectancy?: Antecedent of spontaneous causal attribution. Personality and Social Psychology Bulletin, 18, 659-668] explores the notion of causal attribution. For someone who has studied how people reason about the world, some of his research shows a surprising lack of sophistication in how he chooses to interpret the data. Just saying…

Advertisements

One comment

  1. I totally agree on this one. Well said!



Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: