Saturday, May 8, 2010

Face to Face with Perfume: Dr. Theresa L. White

Theresa White is an Associate Professor of Psychology at Le Moyne College in Syracuse, New York. One of her research interests is sensory psychology. At the recent AChemS smell and taste conference in St. Petersburg, Florida, she gave a poster called “Perfume Masculinity/Femininity Affects Face Gender Judgments.” We caught up with her afterwards and talked to her about the study.

FN: Tell us about this work.

White: We thought that in a social situation where many cues were ambiguous, people would be apt to use olfactory information to disambiguate the situation. So we ran a number of pilot studies. The first one we ran was to see if we could find odors that were marketed as masculine or feminine perfumes and that people did, in fact, perceive as masculine or feminine. We found two: one was Caesars Man and the other was Shania by Stetson.

The other thing we did was to go through a lot of pictures of faces that had been cropped so you couldn’t see the hair, you couldn’t see the mouth, and that were emotionally neutral. We used ones Steven Nosek published in a paper[*] in 2002.

FN: So these were a previously used set of faces?

White: Yes. He sent us twelve: three African American females, three African-American males, three Caucasian females, three Caucasian males. We tested people with that set and found the pictures they were most reliably able to say were male or female. We picked one from each category so that we had four.

Then we took all twelve photos and morphed across sex but not across race. So, one African-American male was morphed with three African-American females, so that every possible combination was morphed. 

I don’t know how much you know about morphing but you can go to a wide number of steps. We went to 100 steps, took the middle 40 to 60, and had people look at those and say whether they were male or female. Then, from the most ambiguous ones, we selected two African-American and two Caucasian. So altogether we had four ambiguous and four distinctive faces.

We divided people into two groups—nine in each group—and they either smelled the masculine perfume or they smelled the feminine perfume while they looked at the faces. They saw each face twice in a random order and their job was simply to say if it was male or female. We anticipated that ambiguous faces would be more likely to be seen as male or female depending on the perfume the person was smelling 

FN: So the perfume would put a spin on their interpretation of an ambiguous face?

White: That’s exactly what we were hoping to see. . . . We didn’t see it. [laughs] 

FN: That’s weird.

White: What we did see was that if you wanted to call that ambiguous face something that was not the perfume that you were smelling it took you longer. So if you were to label a face with a gender that was inconsistent with the perfume you were smelling, it would be slower for you. 

Dr. Theresa L. White

FN: So I’m looking at this ambiguous face and I’m going to say it’s female, but if I’m smelling a masculine fragrance it takes me longer to say so.

White: You got it.

FN: The incongruence between the gender of the fragrance and my bet about the gender of the face slows me down.

White: Right. What we think is going on is some sort of priming effect from the odor. There are a couple of things that could improve this study. We let people look at the face as long as they wanted; they could take their time, “it might be male, it might be female, I’m not really sure”. That’s kind of a problem. I’d like to push people so they have a shorter amount of time to see the face and make their decision.

FN: So flash it at them and then force them to make a call.

White: Right. And another thing: with the distinctive stimuli people were very good; there were almost no mistakes. The mistakes they were made were in the direction of the perfume that they were smelling. So if you were smelling a female perfume and looking at a male face, you’d error by calling it a woman. There were hardly any mistakes at all with distinctive faces, but the ones there we found were in the right direction. So that gives me a hint that if I speed things up, I could induce more errors on the distinctive faces and on the ambiguous faces to the point were I’d get more of an effect.

The other thing I’d really like to do is run the study as a within-subjects experiment because the “good subject tendency” could be at work. In other words, if people know they’re smelling a masculine perfume, they may answer “male” to those stimuli that they’re not sure about because they can guess that’s what the experimenter wants them to do. 

FN: So the way you actually ran it, a person only smelled one perfume but looked at many faces?

White: Right.

FN: And you’d like to run each fragrance against every type of face, all scrambled.

White: That would be my ideal way to do it.

FN: So when you had people rate the gender of the perfumes . . .

White: We did that afterwards.

FN: . . . that was done by the people in the experiment but afterwards. And were those pretty clear results?

White: It was clear that the perfumes we thought were masculine they thought were masculine; the ones we thought were feminine, they thought were feminine.

FN: So you basically asked people “how masculine is this perfume” or “how feminine is this perfume”, using rating scales?

White: You got it.

FN: And people can do that with no problem?

White: No problem.

FN: Some people claim that the gender of perfume is all a function of marketing.

White: Well, there may be an element of that, an element of learning what constitutes male scent and what constitutes female scent within a specific culture. People carry around culture-specific schema if you will, ideas of what a man smells like, what a woman smells like. And so these perfumes fit this ideal schema better for our culture. 

FN: Do you think these cultural schema are purely arbitrary? Could there be a culture where light florals are masculine and heavy woody notes are feminine?

White: I suppose it’s possible. I don’t know of any. [laughs] 

FN: Are there any biological boundaries to what’s possible in the olfactory assignment of gender?

White: You know, that’s outside the realm of any data.

FN: Can you speculate?

White: Like I said, I don’t know of any culture where light florals are masculine.

FN: Come back to the facial stimuli. What is it that makes a face masculine and easy to identify? Are there facial features that people use?

White: Yes, there are definitely features that make a face look more masculine or feminine: brow size, nose size, eye orientation, and puffiness of cheeks, for example.

FN: There are guys who look very feminine and women who look very masculine and yet we don’t deny that are males and females.

White: No. But there are times when you’re in a social situation where you just don’t know. So you look for additional information such as what this person is wearing. If fragrance is something they’re wearing, it’s something you can use to disambiguate an unusual situation.

FN: Let’s talk about possible applications of your work. What if a woman has a very feminine look, roundy eyes and everything you associate with feminine appearance; would it be a good idea or a bad idea for her to wear a gender-ambiguous scent?

White: I think odors help to disambiguate a situation where our more “reliable” sense—vision—is not reliable. At the sensory integration seminar here at AChemS, there was a fellow talking about how we weight sensory information: we weight highly reliable senses more heavily than the more variable senses, and olfaction is an incredibly variable sense. So I really believe we give a lot more credence to what our eyes say than what our nose says most of the time. But in situation where the eyes aren’t doing it for you, you’re going to make use of any other information to help clear things up. 

FN: So a smell could put you over the edge in an ambiguous situation?

White: Yes.

FN: Has anybody done anything with ambiguous voices?

White: Nope.

FN: That would be cool, wouldn’t it?

White: Yup. That would be way cool.

FN: Who was looking at the pictures in your study? Men and women?

White: Only men.

FN: Why only men?

White: Because preferences for masculinity of faces varies with the menstrual cycle and I wanted a less noisy data set. I would have felt obligated to control for phase of menstrual cycle and other factors if I’d used women, and that would have complicated the study considerably.

FN: So what’s the next step in this project?

White: I’m going to do it again as a within-subjects experiment with time limited exposure to the faces. But I have a number of different faces and I’ve morphed across emotions and that’s another direction I’m hoping to go with this. 

I bought Paul Ekman’s set of facial expressions and the ones I was trying to use were pleasantness and disgust because I though disgust was the most olfaction-related expression of emotion.

FN: I once got Paul Ekman to try to flare only one of his nostrils at a time. It drove him nuts for twenty minutes at a psychology faculty party.

White: [laughs]

FN: Thank you very much.

White: Thank you.

[*Nosek, B.A., Banaji, M.R., & Greenwald, A.G. (2002). Harvesting group attitudes and stereotypes from a demonstration website. Group Dynamics 6(1):101-115.]


La Bonne Vivante said...

this is REALLY interesting! It really flies in the face of the idea that gendered fragrance is a marketing ploy. I kind of like it. It means that when I wear my favorite 'masculines' i am confusing those around me in really intriguing ways... Thanks so much for this informative post!

carmencanada /Grain de Musc said...

As serendipity would have I, I was just writing a piece on my first olfactory gender-bending experiment, way back in the punk era... I was quite amused by the selection of scents, and would have been interested to know their composition (this stuff is most emphatically *not* sold in France).

~x~ said...

way to throw in the voice gender ambiguity.

and i'll be trying the one nostril thing til i get it.

Avery Gilbert said...

La Bonne Vivante:

The gender boundaries implicit in fragrance formulas are even more detailed than I what Dr. White and I spoke about. Using the same kind of rating scales, psychologist Debra Zellner found that unisex fragrances (e.g., cK Be are rated as neither masculine nor feminine but in somewhere in between. (Plus, her subjects associated gender-appropriate colors to the scents depending on their gender-based perceptions.)

Gender bending can be hot--like when your girlfriend wears your Brooks Bros. shirt (and nothing else).

Avery Gilbert said...

Carmencanada/Grain de Musc:

Damn! I forget to ask about that. They certainly were some off-beat choices. Let's see if I can get her to dial in here . . .

Avery Gilbert said...


Keep trying. You may discover the only reason in the world to want to get a unilateral case of Bell's palsy.