Ignoring branding: the double-blind test

What is “really” the difference in quality between cheap and expensive wines, restaurants, hotels, medications, jeans, stereo equipment, computer software, etc? I’ve always been fascinated by this question, because I enjoy quality but do not enjoy paying top dollar needlessly, and am aware that price does not necessarily correlate with quality. In particular, brand names and other marketing tricks easily imbue a product or service with an artificial aura specifically for the purpose of luring consumers toward it, away from a practically similar competitors’.

Do the differences exist, do they matter, and what can we learn from the results of double-blind tests?

How to act when there are differences

I am not denying that there exist differences, or that sometimes it takes a real connoisseur to notice or appreciate the differences between, say, something of “high” quality (at 90th percentile, say) rather than of “extremely high” quality (at 99th percentile, say). But even if I notice a difference, I may not want to pay the price to get the higher quality. Suppose I could buy audio speakers for $50 that to me sounded 90th percentile, and I could get another set for $5000 that to me sounded 99th percentile. I would buy the $50 speakers. On the other hand, suppose I could buy running shoes for $50 that are 90th percentile, or shoes for $200 for 99th percentile. I would not hesitate to get the $200 shoes, because of the importance in my everyday quality of life in maximizing my running experience.

Coca-Cola and Pepsi and experts

Lately there has been a lot of interest in objective tests to determine whether people can actually tell the difference between wines, soft drinks, medication, musical performers, etc. Probably the most famous in the past decade has been the Coca-Cola versus Pepsi soft drink wars. We all know the results: when taste tests were blind, people had different preferences than when not blind. This shows the power of advertising to influence our perceptions and preferences. And most recently, we’ve all been hearing about how people cannot tell the difference between cheap wines and expensive wines.

When first hearing about these results, a lot of people get very angry and defensive, because of the cognitive dissonance. One objection they sometimes raise (without even reading the details of the tests) is that yes, some undiscerning individuals might have been fooled, but real “experts” would not be (and by implication, they view themselves as real experts, in accordance with the Lake Wobegon effect). But in fact, these tests are done with experts as well as novices, and sometimes the experts even do worse than novices (this effect calls for its own blog post entirely).

The placebo effect

These matters are not just trivia for amusement. For example, it is an industry secret that the placebo effect when it comes to medication is extremely strong, and even getting stronger. There is controversy over the use or non-use of placebos in medical care. Let’s face it, medical patients are often very overbearing and demand medication from their physicians, who often give them some placebo to satisfy them. Since this actually “works”, is it wrong for doctors to lie to their patients (who seem to inadvertently create the incentive for such lies)?

In the area of music, some years ago there was a big flap over a fraud of a classical pianist that went undetected for a long time: an aging Joyce Hatto passed off recorded performances of obscure pianists as her own, and became lauded for the revival of her recording career. Let’s face it: people like a heartwarming “human interest” story. We love stories about old pianists or athletes making a comeback.

Can you identify the Stradivarius?

Today I saw an article about a double-blind violin test in which listeners heard a segment of music performed on two different violins and were asked to identify the Stradivarius. Apparently people did poorly, including expert violinists.

This article has already attracted over a hundred comments since yesterday, testifying to the interest people have in questions of branding and quality.

My experience with the test

I am not a violinist. Although I would like to learn the violin sometime, I have never handled one. Furthermore, although I have listened to quite a lot of violin playing, I have never been interested in the Stradivarius story or mystique (or that of competing vintage violins). To put it bluntly: I don’t actually know what a Stradivarius sounds like. When I listen to a CD or go to a concert and hear a violin concerto or something, I don’t read the program notes, and therefore I have no idea what kind of violin someone is playing.

I got the correct answer in the listening test.

What does that even mean, though, given that I don’t know what a Stradivarius sounds like and don’t really care?

The two violin tracks sounded quite different to me. The best I could do was choose which one I liked better. If it wasn’t the Stradivarius, no big deal; if it was, no big deal.

One violin sounded rough and raspy, and the other one was smooth. That’s all I could go on. I don’t know if a rough or smooth sound is actually considered desirable by aficionados. I don’t know what the Stradivarius is supposed to sound like. And who knows whether the violinist just happened to be doing very different things physically on the two tracks presented? It would have been better if there had been more samples.

Deliberate ignorance?

In order to shield myself from what I consider extraneous noise, sometimes I enjoy deliberately being ignorant. Much of the music I like, I originally encountered without knowing its composer, performer, or title. I don’t really care if something was written by Mozart or whether he plagiarized it from someone else. If it is interesting, that’s fine, and if not, that’s also fine.

I recognize the virtues of tradition, of authority, but the branding effect privileges the old over the new in ways that the old probably would not have enjoyed. There is no way Mozart would have enjoyed having his early childhood music all collected and recorded as though it were as valuable as his later masterpieces. Unpublished drafts by dead novelists get more attention than published work by new writers. Nostalgia for the old, in the name of “authenticity”, always sells well. Older religions get more respect than newer ones, which get called cults for a while just because they are newer. Older political parties attract a following because of their branding.


We are not as clever or discerning as we tend to think we are, and are influenced in our perceptions and decision making by branding, sentimentality, framing, contexts.

Still, we can’t live without shortcuts for quality and consistency. It’s not possible to spend vast amounts of time doing one’s own blind tests on everything, and it’s not worth it either. We are human and should respect our limitations and imperfections. Humility and respect is the best we can do. Therefore, I don’t make fun of those who are “tricked” in double-blind tests. That’s us, that’s me, in some context or other.

comments powered by Disqus