Cognitive Daily

One of the most common “icebreaker” conversation topics is music preferences. We ask friends what they’re listening to on their iPods, bloggers post playlists on their sidebars, and one of the most popular websites on the planet (MySpace) is built around sharing music. The assumption is that musical preferences can tell us something beyond what someone likes to listen to — we believe we can judge a person’s personality, fashion preferences, and more based just on the style of music they prefer.

For me, it’s difficult not to form a mental picture of a person when I hear what music they listen to. A heavy metal fan evokes an entirely different mental image for me compared to a classical music buff or someone who likes religious music. But are these mental pictures accurate? Can we really make reliable judgments based just on music preferences? Until recently, very little research had been conducted on the subject.

We’ve discussed one such study here on Cognitive Daily, by Peter Rentfrow and Samuel Gosling, which suggested that music preferences do correlate with some personality traits. Now Rentfrow and Gosling have conducted a new study which explores a wider range of musical styles, and focuses more on stereotypes than individual preferences.

They quizzed over 200 college students on their stereotypes of fans of one of 14 different musical genres: blues, classical, folk, jazz, alternative, heavy metal, rock, country, pop, religious, soundtracks, electronic, rap, and soul. For most genres, the judges were largely in agreement as to what the typical fan was like.

The graph below shows how the judges rated five personality traits of fans of four of the genres:

For each of these genres, the judges were largely in agreement. But when Rentfrow and Gosling asked 85 actual music fans to rate their own personalities, their own ratings agreed with the judges for just two of these genres: rock and religious. In fact, judges’ personality ratings correlated with actual ratings for only seven of the fourteen genres studied.

For personal qualities such as attractiveness, intelligence, religiousness, and conservatism, correlations between judges’ ratings and actual ratings were even rarer — they occurred in only four genres.

The researchers also asked both groups about values such as ambition, love, salvation, friendship, and courage, and again found that the stereotypes did not always match the actual ratings. In only two genres — jazz and religious music — did the stereotype correlate with the actual ratings for personality, personal qualities, and values.

Rentfrow and Gosling asked judges to stereotype music fans for which recreational drugs they thought they preferred, but because of privacy concerns, they didn’t ask actual music fans about it. Still, it’s interesting to see how the stereotypes pan out:

Maybe we could do a Casual Friday study to see if the stereotype matches up with actual drug preferences. Do you think there are ethical issues with such a study? While our survey software does collect the IP addresses of those who fill in the survey, it would be impossible for us to identify individual respondents.

Rentfrow and Gosling point out one additional limitation to their study: They only studied college students. Perhaps older adults have different stereotypes and actual preferences. Again, perhaps this is something we could address on a Casual Friday. What do you think?

Rentfrow, P.J., Gosling, S.D. (2007). The content and validity of music-genre stereotypes among college students. Psychology of Music, 35(2), 306-326.

Comments

#1 more of a tech guy
February 20, 2008

I think in order to have any relation to reality we need a double blind study which this was not. Self-ratings don’t mean much. For instance a fan of religious music might be self-conscious about religious stereotypes and compensate by giving him/herself a high rating for openness.

The ratings need to be based on the ratings of others, perhaps after a simple unrelated group task is completed, as a way of giving them interaction and a basis for rating. The ratings can then be performed without any prior knowledge of each other’s musical preference.
#2 Bobby
February 20, 2008

“Perhaps older adults have different stereotypes and actual preferences.”

Yes, I believe so. It would be great to see this revisited on a Casual Friday.

Did this study take into account people who are fans of more than one genre? I consider myself a music nerd and in a typical day I will actively listen to rock, jazz, rap, classical, and metal – depending on my mood and task.
#3 agnostic
February 20, 2008

There’s a big problem with this — the average undergraduate at U of T Austin (the participants) has little to no experience with fans of almost all of those 14 genres. It’s not surprising that we’re bad at stereotyping a group we have no contact with, however indirectly.

If I wanted to get the conclusion that people are bad at stereotyping music fans, I would have included 20 or 30 genres, mostly unknown to the participants, and shown how in “only” 7 of them were people good at stereotyping.

So, what they need to do is have students also provide some indicator of how confident they are — fairly confident, have a hunch, or shot in the dark guess. Or maybe have the indicator reflect how much exposure they’ve had to the target fans (whether first-hand, seeing lots of heavy metal documentaries, or whatever).

The same thing comes up in stereotypes about national character — in fact, people are incredibly good at guessing the direction of the average personality traits in various countries / ethnic groups, although they tend to exaggerate the magnitude. But they perform poorly when asked about the average person in a country they’ve never heard of or know nothing about. Duh.

The authors tried to spin this as showing that popular stereotypes about national character were wrong, but of course that’s wrong.
#4 Sarcastro
February 20, 2008

Given the results in regards to marijuana, I take it they aren’t considering reggae to be religious music.
#5 zeferino
February 20, 2008

I’m also suspicious of self-reporting. I would expect people who listen to religious music and, perhaps, classical to have a bit of a ‘moral identity’ which would cause them to report they are better than they really are, as I’ve noticed happen in real life. (same goes for Jazz, too, btw. I can definitely tell you that after being a music major.)

Likewise, those that listen to Rap & Metal might want to be percieved as more reckless and wild than they are. If they believe that music preference puts forth an image, they might be just as inclined to skew their answers to uphold that image just as much as a religious music fan with a moral identity would.

Just my thoughts…
#6 Cletus McFly
February 20, 2008

They forgot the two best kinds of music, country AND western.
#7 Dunc
February 21, 2008

What if you class yourself as a fan of at least 8 out of those 14 categories of music? It’s not unknown for me to listen to (for example) Miles Davis, Black Sabbath, Gordon Lightfoot and Fairport Convention in the same listening session.

There are only 2 kinds of music – good and bad.
#8 Apium
February 21, 2008

Haven’t there been studies showing that people are incredibly poor at self-rating? Their perception of their own qualities was contradicted when friends were asked? And even something where people could evaluate personality traits based on a five-minute tour of their living space? Perhaps a better way of evalutating musical stereotypes would be to recruit friends and/or family and weight the other’s comments higher?
#9 Hessian Nerd
February 21, 2008

Like Dunc, I listen to several of the aforementioned genres. I were to choose one, however, I would first and foremost be a ‘heavy metal’ listener. As a fan of all sorts of Metal I consistently run up against wild and inaccurate stereotypes. People really tend to get put off by my primary taste in music. My personal experience with the metal culture has been a very positive one. I have met lots of honest, generous, and intelligent people.

Please read the following article and think twice the next time you see a long haired guy in all black sporting a SLAYER T-shirt.

http://www.telegraph.co.uk/connected/main.jhtml?xml=/connected/2007/03/21/nmetal21.xml
#10 JET
February 22, 2008

One major flaw here; “Religious” is in no way a musical genre – except in that most religious popular music tends to be sub-standard because lyrical content and creed of the musicians is as important as the actual musical qualities (or more so).

Almost as serious a flaw is that Rock is so all-encompassing a term as to be virtually pointless, and – seriously now – where is the line between Rock, Pop and Alternative? Classical ranges from Vivaldi to Stockhausen, Electronic goes from Daft Punk to Squarepusher, Jazz from Kenny G to Miles Davis, etc. Everyone I know listens to at least six or seven of those genres on a regular basis, but only select acts within those genres. So, is Emo alternative? What about Goth? Post-Hardcore? How about Post-Punk? Hey, how about Punk?

It’s an intriguing test to perform, but I think it needs to be redone with more detail and a more global genre distribution. And without “Religious” as a genre.
#11 Marc
February 23, 2008

I have the same opinion about this as many of the commentors above: the self-report aspect of this study makes the results questionable. Almost every one of the qualities (conscientiousness, attractiveness, intelligence, emotional stability, etc.) are considered valuable to possess. Even the least conscientiousness or stable person taking part in the survey is going to be inclined to rate themselves higher in such categories than they really are. And the musical categories are much too broad. I personally listen to mostly indie rock, but the label “rock” brings to my mind hard rock music, which has a very diffferent fan base.
I do like the intent of the study, however. I feel that many people who listen predominantly to one type of music end up assimilating a little more into the culture of that music, or a culture has the effect of introducing them to music more in line with itself. We come to consider our musical tastes part of our identity, we often put them on display, and use them as ways to relate to other people. Thus, I think they both shape, and are shaped, by our personalities.
#12 Samantha Vimes
February 24, 2008

Don’t most people like more than one type of music?
#13 James
February 25, 2008

The study would have to also state whether they are doing the tests with musicians or with non musicians. As a musician, or maybe not for that reason, I am a fan of a very wide variety of music, I can get into classical as much as I can get into metal as I can get into rap as I can get into 0 beat ambient.
I’m not sure whether religious music can be considered a genre, as, excuse the bias, they may be indoctorined into listening to it, or feeling obligated to listen to it. However, that been said, peers can pressure each other to listen to specific genres of music.
#14 Yenzo
February 28, 2008

JET is absolutely right, there is waaay to much variance in each of these genres. It might make a lot more sense to ask for favorite bands and then categorize those bands into genres yourself. But first, of course, you’d have to establish a list of at least 1,000 or something band names to choose from… pretty complicated research, if you ask me.

What astounds me in the above results is that Rock ranks higher in the marijuana stereotype than Rap (probably not significantly, but still). Marijuana use is promoted a lot more in Rap songs than in Rock songs, and there are more mellow Rap songs that you might probably want to listen to when you’re high. All that might not influence the actual users’s behavior, but it should influence the cliché.
#15 Utbildning
February 28, 2008

I guess that one major problem with this kind of research is that kids nowadays are listening to not only one (as in the old days) but three, four and even more different genres at the same time. Furthermore, the genres were much more clear cut during the seventies.
#16 Nick
February 29, 2008

I don’t think there’s something wrong with self-report per se. How, exactly, did the music fans in this study self-report? If they were just asked to rate themselves on a given personality domain from 1 to 7 (which appears to be what was done), then this study is probably problematic.If, however, the study employed personality inventories whose reliability and validity were already established for this study’s population, there shouldn’t be a problem with the data being self-reported.
#17 Kerry
October 13, 2009

this is the dumbest thing i have ever witnessed.. i am ashamed to say that i have even come across this and actually read it.. this is just plain ridiculous.. there are way more music categories than this.. where did you get your information from? this is stupid.. flat out ridiculous..