Sergeant, M.J.T., Davies, M.N.O., Dickens, T.E., Griffiths, M.D. (2005). The selfreported importance of olfaction during human mate choice. Sexualities, Evolution, & Gender, 7, 199-213.
Publications by Mark Sergeant et al., are well worth reading by those who want to know more about human pheromones, but do not turn up these works via typical search strategies.
Mark also reviewed the updated 2002 version of my book in 2003 for The Human Nature Review–here’s the URL:
I am highly skeptical of this study. The reason why is that there is no measure for the concentration of the ‘body odor’ on the t-shirts. Men who live with women or are around women -have lower testosterone. I imagine that most homosexual men do not live with or around women.
Therefore it follows that homosexual men have higher testosterone on average AND testosterone affects concentrations of ‘male body odor’.
Even if concentration of ‘body odor’ was measured on the t-shirts, without measurement of individual olfactory thresholds, the measurement wouldn’t have much meaning. Have you read the article in its entirety? I agree, and so does Sergeant et al., that there are many confounds. But I also add, quoting the paper that: " In all cases, hedonic evaluations were independent of odor strength (i.e., stronger smelling odors were not perceived as being more unpleasant, while weaker odors were not perceived as pleasant)." Do you have a better idea of how best to measure the odor concentration? In this type of study, hedonic ratings are often used and that is probably what I would have used. In any case, studies are limited by how cost effective the measurements might be–and you never know until you analyze the results. So, the issue of funding is inevitable–as is the “hard” science (e.g., chemistry) versus “soft” science (e.g., psychological measurements) approach. I’m hoping to get more ideas on the “hard” science, if possible–but any ideas are welcome.
It would take a very long article to fully address the ramifications of testosterone levels that may or may not vary with adult sexual orientation, and living arrangements.
Unfortunately, I am unwilling to pay the $32 for the full text from Springer-Verlag.
I realize that this is ‘cutting edge’ science but when designing an experiment like this -wouldn’t it make sense to gather other ‘unknowns’ to look for correlations? Again, I don’t have access to the original study so I don’t know what measures are there –I am merely speculating.
Most major breakthroughs are accidents. For example Fleming discovered penicillin while cleaning his dishes. Teflon, TNT, and bubblegum were all accidents. But how does one go about finding an accident? Someone makes an observation and then infers something –which is then tested.
The problem is that in the absence of knowledge –it is very hard to make these inferences. What if, all the gay guys in the study (all 10 of them) were communal hippies –and they didn’t bathe. Would this affect the results? What if 6 of the hetero men (n=9) were older than 70? Having small sample sizes of each test group is problematic because anomalies can spring up.
Nevertheless, collecting as much data as possible (even more data than reasonable) is always the best solution. One can not make inferences in a vacuum. If I were trying to form a hypothesis on this study –many questions would form. How close to ovulation were the women who smelled the shirts? Were the women closer to ovulation more apt to think that a shirt smell was sexier? Who were the men in the study living with? What did they do for a living? What is the difference between the body odors of the homosexual and the heterosexual men?
This last question is thoroughly pertinent, because here on the forum we think that it’s pheromones. The million dollar question is: which pheromone is responsible for a higher ‘hedonic rating’.
You can almost always get a copy from the author, though email addresses are sometime difficult to find.
Depends on what you’re looking for. If you examine too many things, some soft scientists will tell you that each finding is statistically less significant because you looked for more than that thing. PE Meehl has argued that, statistically, everything correleates–and that also it does not–and that soft scientists might better pursue careers in shoe sales.
Most researchers try to level the playing field from the start because of small sample sizes. That’s why most researchers also read the full text of pertinent articles–to make sure that this was satisfactorily done. To be published in a peer reviewed journal, the study design and statistical analysis must meet the criteria of at least one (and usually more reviewers). That’s why publication is important! But it is not the final word on any subject.
The published study addresses such things in its introduction to tell others why they undertook the study. What they did to control conditions is included in another section of the publication.
Researchers are rapidly getting closer to answering this question, and this is why I hope others will take an interest in research. My next publication may tell everyone which pheromone is responsible–if the submission passes peer review. What will set it apart from other individual publications is that the study design and execution were based on a “hard” science model for which I have won “hard” science and “soft” science awards. Some product marketers don’t want to present both sides. It is nice to see that, so far, Pherotalk is promoting this type of discussion. Hopefully, it will continue.