Monday, November 2, 2009

Cranking the Bubble Machine: “The Smell of Virtue”, Pt. 2

In reviewing Katie Liljenquist’s study of Windex and “the smell of virtue”, I pointed out some of its weaker points—the lack of control conditions using other odors, and its glib treatment of moral concepts. Other aspects of the study had the alarm on FirstNerve Bogosity Meter bleating like an angry goat, but I thought it unseemly for my critique to run longer than her entire scientific article. So I scrawled a few notes on my copy of the manuscript and set it aside. 

(Yes, manuscript. Although news stories based on a BYU press release are fully dispersed on the Internet, an electronic pre-print is not yet available from Psychological Science. Dr. Liljenquist was kind enough to send me the manuscript. It is also available here.)

One of my marginal notes was “What does it smell like?” The description of “citrus-scented Windex” puzzled me—I still think of Windex as the blue liquid in the spray bottle, the Glass Cleaner with Ammonia-D®. One thing we’ve learned at great cost over the years in the study of smell is that you have to be specific about the scents you use, if for no other reason than it allows others to replicate your work. Another lesson: any one scientist’s description of a test smell doesn’t cut it. For example, in the Liljenquist study there is no independent evidence that volunteers found the Windex product to smell “clean” or “citrus-like”.

What sent me back to my notes were comments on my post by Minette, who blogs at Minette also takes issue with the study and in doing so riffs on the olfactory character of Windex. She challenges Liljenquist’s characterization of it as “unobtrusive” and asks why other types of cleanser odors weren’t used. She also said she couldn’t find the product on the Windex website.

Well, that did it: time to sniff for myself. I drove over to the A&P and checked out the chemical aisle. There was the blue stuff: Original Windex With Streak Free Shine! Next to it was Windex Antibacterial Multi-Surface cleanser: Kills 99.9% of Bacteria in Seconds! Windex AMS is a yellow liquid and has a sliced lemon on the label. This, evidently, is what was used in the Liljenquist study.

I brought the yellow AMS home and rummaged around under the sink for my bottle of Original. Spraying each on a blotter card—the preferred method of fragrance evaluation for household cleaning products—I could now compare the two. Windex Original has that familiar soapy note that takes the edge off the ammonia smell. Windex AMS smells like lemons: there’s a sweet, lemon gum drop note but also something harsher and terpenic beneath it. It’s no wonder it doesn’t smell like ammonia: the label proclaims it Ammonia Free!

So if you—like Minette—were imagining the iconic ammonia smell of Original Windex and having a hard time squaring it with the smell of virtue—well, the problem isn’t with you. 

But where could we have gotten the idea that the Windex in question was a glass cleaner and not a hard-surface cleanser?

Oh, that’s right. From the BYU press release:

Exit question: Shouldn’t three business school professors respect the difference between product segments, brand extensions, and flankers?

P.S. An academic buddy of mine told me I should send a link of my post to the editors at Psychological Science. Nice idea. When I looked up the journal’s masthead who did I find on the editorial board? Why, Adam D. Galinsky, one of the authors.

As the Church Lady used to say, “How conveeeeeenient.”

P.P.S. Full disclosure: S.C. Johnson is a former client of mine; I belong to the Association for Psychological Science and subscribe to its journal; and I think Dr. Liljenquist has great squeegee technique.

UPDATE November 16, 2009

Looks like I'm not the only one who’s wearing his cranky pants.

UPDATE November 19, 2009

Michael O’Brien at Human Resource Executive Online uses the First Nerve critique to nicely balance his story on the Liljenquist study.


Anonymous said...

great follow-up. and awesome detective work, going to the store and finding the citrus-scented windex. i read the descriptions for all of the products, and none of them mentions citrus - the AMS doesn't mention an aroma at all! that's why i thought the citrus-scented windex product did not exist.

now that's it's being linked with this study, they oughta consider changing the web copy to promote its citrus magic.

oh, wait, that's another product. i love citrus magic - citrus oils that can cover up even the most potent of cat-box smells. (smells like you've just sliced open fresh oranges and lemons.) and it's all natural.


Nathan Branch said...

Fascinating dissection of what appears to be a half-a**ed study all dressed up for maximum blog exposure.

The study itself would have been far more interesting to read about if they had bothered to include information on all the points you addressed. As it is, the conclusion comes across as shallow and glib -- cleanliness promotes virtue.

Just another brick for the nanny to throw.

Avery Gilbert said...


I'm always about the smell.

Which is why I have little patience with olfactory science (or olfactory art for that matter) which doesn't fully connect with the phenomenology of odor.

Avery Gilbert said...

Nathan Branch:

"Brief Communications" have their place in scientific publication, but brevity shouldn't be an excuse for gliding over important questions of experimental design and interpretation.

Remember "Designer Imposters"? They'd smell like Giorgio or whatever for about two minutes then leave a lingering bad impression . . .