more on science

http://www.newscientist.com/article/mg20026851.800-the-secret-signals-in-human-sweat.html
"The Secret Signals in Human Sweat"
by Caroline Williams
December 3, 2008
NewScientist

This is an article about human pheremones. The author writes:

"Brain-imaging studies are also providing tantalising evidence that humans release and respond to alarm pheromones. These have been less well studied than those thought to be involved in attraction, but a handful of psychological studies have claimed to show that humans can detect the "scent of fear". In 1999, Denise Chen, a psychologist now at Rice University in Houston, Texas, asked a group of volunteers to sniff sweat from people who had watched either funny or scary film clips. More than half of the volunteers successfully identified a sample of fearful sweat despite not being able to consciously smell any difference in the samples."

Let's pause a moment and analyze this. "Over half" - how far over 50%? As far as we're told, subjects were just given a binary choice. And as we know, in such a situation, just picking randomly leaves you with a 50% chance of being right. So how many people did they conduct this study on? And what, exactly, was the percent correct? Did they test each person once - or is the study rather saying that 50% (or so) of people can reliably identify fearful sweat? (Was it a double-blind experiment? What about control sweat?)

Of course, what is nice is that the author immediately points out that these studies were done badly: leading questions were asked, and small sample sizes were used. They then turn to another study. In this study, subjects were placed in an fMRI and sprayed with what, unbeknownst to them, was fearful sweat. The fMRI detected reactions in their fear centers.

But there are questions here as well: maybe these test subjects were just freaked out because they were placed in a big, loud machine, for unknown (?) purposes, and then sprayed with some random aerosol. In other words, this study should have used several controls - sprayind subjects with water, with heat sweat, with happy sweat, with nothing at all - and then talked about those results. (Another question: how many of the subjects had been in an fMRI before?)

While the article doesn't mention these, it does discuss one other problem:

"Lundstrom says there is still a missing piece of the puzzle. "To my mind, activation of brain is not enough," he says. "I use brain imaging in my work but I like to see that there is a behavioural response - and to see it consistently. Not just once, but every time.""

This is important. Consider the following article:

http://www.newscientist.com/article/mg20026845.000-memories-may-be-stored-on-your-dna.html
"Memories May Be Stored on your DNA"
by Devin Powell
December 2, 2008
NewScientist

This article discusses new research that indicates that DNA methylation is responsible for the storage of memories, rather than neural structures (although it hints that the mechanism by which methylation stores memories is by influencing neural structure). The reason I mention this is to back up Lundstrom's point: in the end, it is human behavior we're interested in, and not the mechanisms that underly it. That is, in order for what was detected by the fMRI to be interesting (from the point of view of understanding human psychology and sociology), it needs to be shown how it links up with behavior.