Have you ever read about a study, perhaps on this blog even, and thought to yourself, "Well those results are interesting in the lab, but they have absolutely no implications for life outside of the lab?" I remember quite clearly thinking exactly that when I was told about the name-letter effect several years ago. The name-letter effect is the entirely unsurprising finding, first reported (as far as I know) by Nuttin1, that people prefer letters in their names, especially their first and last initials, over other letters. They also prefer numbers in their birth date over other numbers. Wow! Who'd a thunk it? I mean, people write their names and birth dates thousand of times over a lifetime, so it's entirely unsurprising that people would prefer letters and numbers in them over other letters and numbers. But who cares? It's not like that affects anything people do outside of rating letters and numbers in a psychology experiment, right?
Or does it? Soon after I was told about the name-letter effect, the same person told me about research a mutual acquaintance of ours was conducting. That research (later published in the Journal of Consumer Research2) ultimately involved four studies, in which they found that if asked to name a product they desired (e.g., a food while hungry or a drink while thirsty), people were more likely to choose a brand name that began with the same letter as their first name than one that didn't; that when given two unknown brands of tea (rated equally tasty by a separate set of participants), people were more likely to select the brand that begins with the same name as their own first names; and when choosing among familiar brands, they will prefer those that begin with the same letter as their first name. So the name-letter effect does have potential real-world implications! If your name is Chris, you might prefer Coke, but if your name is Penelope, you'll likely prefer Pepsi.
But ya know what? Big deal. Coke and Pepsi are pretty much the same thing anyway, so it doesn't really matter which one you choose. And the same goes for most brands of most types of products.. The first letter in your name influencing your product choices will have little effect on your life. This "so what?" attitude was reinforced when I read a paper by Hodson and Olson3 replicating the name-letter effect with brand names, but finding no name-letter effect for the names of food types, animal names, leisure activities, or national groups. That is, the first letter of the names of those sorts of things didn't affect how much people liked them, probably because there are much bigger differences between food, animal, leisure activities, and national groups than there are between different soda brands.
But then I was pointed to a paper titled "Why Susie Sells Seashells by the Seashore: Implicit Egotism and Major Life Decisions," and my incredulity came crashing down. Or I became intrigued, at least. That paper, by Pelham et al.4, really has to be read to be believed, so I recommend you click on the link and do so. I'll continue when you're done.
Done? OK, let's summarize what Pelham et al. did. Here's their description of their first study:
In Study 1, we first identified the 40 largest cities in the United States. We then consulted the 1990 census to identify all of the common (top 100) male and female first names that shared a minimum of their first three letters with any of these city names. Because the popularity of different first names varies with age, we selected the two qualifying European American female names that we could match most closely for age (based on the relative proportion of deceased people born in 1900 and 1950 who had these first names). We then repeated this procedure to produce the two most closely age-matched male names. The resulting name-city combinations for women were Mildred-Milwaukee and Virginia-Virginia Beach. The resulting combinations for men were Jack-Jacksonville and Philip- Philadelphia. To generate the data for Study 1, we created two separate 2 X 2 name-city tables, one for women and one for men. We expected that the women and men whose first names resembled the name of a specific city would be over-represented among the deceased residents of that specific city. (p. 471)
Here's their table (Table 1, p. 471) presenting the results of Study 1:
Pelham et al. first computed the expected number of people with each name in each city (based on the number of people with that name in the population and the size of the city), and compared that number to the actual number of people in a city. As you can see in the table, there were more Mildred's in Milwaukee than we would expect, but fewer than we would expect in Virginia Beach. On the other hand, there were more Virginia's in Virginia Beach than we would expect, and fewer in Milwaukee. The same pattern occurred for the male names (Jack in Jacksonville and Philip in Philadelphia).
What the...? OK, let's pause for a minute. There are perfectly reasonable explanations for this. People's parent's choose their first names, and it may be that parents are more likely to choose names that sound like the names of the cities in which they live. So this isn't really very good evidence that the name-letter affect influences where people live. Recognizing this, Pelham et al. conducted a second study using last names instead of first names. People usually don't choose their last names, so if they find similar results with last names, it would be pretty strong evidence that the name-letter effect is influencing where some people choose to live. In the second study, they looked at the state rather than city level, as well, because some of the last names weren't common enough to produce sufficient data at the city level. Once again, they computed the expected number of people in a state with a given name, and compared that to the actual number of people in that state with that name. Here's their data (Table 2, p. 472):
In 5 out of 8 of the states, the last names they chose occurred more often than expected, and overall, names beginning with the same letters as a state were significantly more likely to occur in that state than expected. In a third study, they obtained the same pattern of data using surnames and Canadian cities, with 7 out of 8 of the name-city combinations occurring more often than expected. Study found the name-letter effect in state-name combination for male and female first names in four southern states each (Florida-Florence, Georgia-Georgia, Louise-Louisiana, and Virginia-Virginia for the females; George-Georgia, Kenneth-Kentucky, Louis-Louisiana, and Virgil-Virginia for the males). They found that name-state combinations occurred more frequently than expected for 7 out of 8 of the combinations (Virgil's don't like to live in Virginia, apparently).
In their final look at the name-letter effects influence on where people live, they looked at people's birth dates and cities that have numbers in them (e.g., Five Points, Alabama). This time they computed the expected number of people living in a city who were born on a certain data, and compared it to the actual number. Here's the data (Table 9, p. 478):
This time, they found that in 7 out of 7 cases, people were more likely to live in cities with a number in the name that matches their birth date than expected by chance.
At this point, your mind is probably swimming with questions and objections, but hold on to those for a minute, because after study 6, things get even weirder, starting with Study 7. I'm going to let them describe this study for you, because I'm not sure you'd believe it if you only heard it from me:
We searched for dentists and lawyers by consulting the official Web pages of the American Dental Association (http://www.ada.org/directory/ dentistsearchform.html) and the American Bar Association (http://lawyers. martindale.com/aba). These sources proved to be very useful because they provided comprehensive national directories of the official members of these two professional organizations (though they only allowed searches on a state by state basis). Each of these directories also allowed for searches based solely on people's first names, and neither directory placed limits on the number of hits they allowed for a specific search. We began this search by consulting 1990 census records. Using these records, we attempted to identify the four most common male and female first names that shared a minimum of their first three letters with the names of each of these two occupations. However, we had to relax our three-letter criterion from Law to La for all of the female names (and for three of the four male names) because there were no names that qualified using the stricter criterion. The 16 names we generated in this fashion included the female names Denise, Dena, Denice, Denna, Laura, Lauren, Laurie, and Laverne and the male names Dennis, Denis, Denny, Denver, Lawrence, Larry, Lance, and Laurence. We expecte that people with names such as Dennis or Denise would be overrepresented among dentists, and people with names such as Lawrence or Laura would be overrepresented among lawyers. Because the lawyer search engine produced a great number of false alarms involving last or middle as opposed to first names, we carefully cleaned the data generated by this search engine to limit hits to true first name hits (no such problem occurred for the dentist search engine). Finally, to make manageable the task of cleaning the lawyer data, we limited both searches to the eight most populous U.S. states (California, Florida, Illinois, Michigan, New York, Ohio, Pennsylvania, and Texas). (p. 479)
If you're like me, you're thinking, "No way!" Way. As with the location-name combinations, they found that for both dentists and lawyers, occupation-name combinations occurred more often than expected when the occupation and name shared the first few letters, but not when they didn't. To provide further evidence for this version of the effect, in their eighth study they compared the number of published geoscientists with names that began with the letter "G" to the number with names beginning with the letter "T" (a first-letter that occurs about as often as "G" in the general population). Consistent with the results of Study 7, geoscientists were more likely to have names that began with "G" than with "T." And study 9 found that hardware store owners were more likely to have names that begin with H than R, while roofing store owners were more likely to have names that begin with R.
OK, what's going on? Pelham et al. argue that their results are evidence for a type of "implicit egoism." Implicit egoism refers to the idea that "people gravitate toward people, places, and things that resemble the self"5. According to Pelham et al., implicit egoism is so strong that a resemblance as trivial as common first letters can influence major life decisions like where we live and what we do for a living. But there are reasons for doubt. Using Pelham et al.'s data, Marcello Gallucci6 attempted to replicate their findings using expanded analysis. He failed to do so, for the most part, and concluded that Pelham et al.'s findings were essentially statistical artifacts -- mere byproducts of the properties of distributions and the flaws in the analyses Pelham et al. conducted on them.
Sounds like the end of the story, and it's a disappointing ending, right? Not so fast. In a reply to Gallucci, Pelham et al.7 admit mistakes in their first paper, but report that when they analyzed the data using more conservative methods, they found the same pattern of results. They also found new support for their implicit egoism hypothesis in several new studies on name-location pairs. They conclude that their findings can't be dismissed as statistical artifacts, but that in some cases -- more than anyone would expect -- people really are making major life decisions based on shared first letters.
Currently, the debate over the data continues to rage, and I wouldn't be surprised if some really interesting papers were published in the next few years. For now, I think it's safe to say that I need to suspend my initial dismissal of the name-letter effect as irrelevant. It could turn out that I was right, but given Pelham et al.'s data, and the insufficiency of Gallucci's criticisms in the face of Pelham et al.'s reply, it's just as likely that I'm wrong. At the very least, though, this makes for some good reading, and further evidence that people are weird!
1Nuttin, J. M. (1985). Narcissism beyond gestalt and awareness: The name
letter effect. European Journal of Social Psychology, 15, 353-361.
2Brendl, C.M>, Chattopadhyay, A., Pelham, B.W., & Carvall, M. (2005). Name letter branding: Valence transfers when product specific needs are active. Journal of Consumer Research, 32, 405-415.
3Hodson, G., & Olson, J.M. (2005). Testing the generality of the name letter effect: Name initials and everyday attitudes. Personality and Social Psychology Bulletin, 31, 1099-1111.
4Pelham, B.W., Mirenberg, M.C., & Jones, J.T. (2002). Why Susie sells seashells by the seashore: Implicit egotism and major life decisions. Journal of Personality and Social Psychology, 82(4), 469-487.
5Pelham, B.W., Carvallo, M., & Jones, J.T. (2005). Implicit egoism. Current Directions in Psychological Science, 14(2), 106-110.
6Gallucci, M. (2003). I sell seashells by the seashore and my name is Jack: Comment on Pelham, Mirenberg, and Jones (2002). Journal of Personality and Social Psychology, 85(5), 789-799.
7Pelham, B.W., Carvallo, M., DeHart, T., Jones, J.T. (2003). Assessing the validity of implicit egotism: A reply to Gallucci (2003). Journal of Personality and Social Psychology, 85(5), 800-807.
This is fascinating, despite it not applying to me. (Wait! I'm a scholar!) I wonder, though, if it correlates with specific letters or on a phonemic level -- the "sco" of my name sounds like the "scho" of "scholar," after all.
This is crazy. I love these bits of research on hidden biases.
I'm confused about Table 9. It looks to me like the expected values are all equal to or lower than the actual values; i.e. if anything, people are less likely to live in cities with numbers that match numbers in their birth dates.
Brent, the expected values are in parentheses. The actual values are those in bold next to them. All of the actual values are equal to or greater than the expected values.
Note that the expected values are what you would expect if the name-letter effect were not having an influence. If the actual values are higher, then that's evidence that the name-letter effect is in fact influencing the decisions.
As a kid I definitely wanted to try (and subsequently liked and consumed for many years thereafter) the cereal "Special K" because my name was Katherine.
I can't really see how this applies to me. I guess if you stretched it really far, you could almost say I work with RF technologies....
Anyway, I think you went back 500-600 years, you might see an even bigger correlation, occupation-wise anyway. (ie: Smith, Hunter, Taylor, etc.) Seems we've come full circle.
Sounds like thundering zebras to me. The numbers in second study are small enough to be counted as artifacts due to a founder effect. My surname is shared by about 3,200 people in the U.S. About half of those live in North Dakota an Minnesota just because that's where the family landed in 1904. The pair matching methodology sounds hinky to me too. What about all those other letters. More H's may own hardware stores than expected, but maybe so could C's, L's and Q's. In fact, maybe half the letters are above expectation and half are below. Maybe the lawyer-dentist study found some more significant numbers, but I would want to see some more broadly based populations before I would attribute much significance to this. I would make a wager, for example, that men named Ole would be most likely live in states with an O in the last half of the name. Not because of any psychological reason, but because Scandinavians generally settled in the Dakota's, Minnesota and Wisconsin and many later moved to Washington in the 50's and 60's or retired to Arizona.
justawriter, you could be right. Check out the paper, and look at some of the other analyses they run. See what you think. Like I said, I don't really know what to make of it. I'm skeptical, but I think it's cool enough to pay attention to.
Certainly food for thought despite the issues raised with the methodology. FWIW I currently work in Paid Search and live in Portobello following a degree in Philosophy and a Post-grad in Publishing. Favourite food is pasta with pesto (seriously!).
Of course I asked colleagues about this and almost no-one else had any name-letter relationships in their home towns, occupations, etc. Looks like I'm the most egotistical in the room, no surprises there :-)
This is a very interesting post Chris! My name is Avery Archer and I attended Andrews University (USA) as an undergrad, and I'm currently an M.Phil student at St Andrews (Scotland). A few minutes ago I was reading the blog of a fellow-philosopher by the name of Brian Berkey who is PhD student at...wait for it, wait for it, Berkeley! Anecdotal, but still mildly disturbing (particularly in my own case).
I thought I had wondered into Effect Measure or one of the math blogs but OMG! That is statistically significant wierdness.
This is quite interesting stuff, though like Chris I'm not quite sure what to make of it. I can say, just as a biographical remark, that I'm less disturbed by the similarity between my last name and my current home institution than Avery seems to be; Berkeley was by far the best department that I was accepted to, so there was no real decision to make, and hence no sense in which the name similarity affected where I ended up. Of course one might wonder if the similarity could have affected those who read my application, making it more likely that I would be accepted than if my name were different. Of course I'd like to think that this wasn't the case, but if Pelham is on the right track, then who knows.
I wonder if there is any internal echoing as well? I'm thinking of the ch/k in Chris and Psychologist.
There is no doubt that this borders on "statistically significant weirdness". It even borders on being able to propose other dangerously radical ideas: that our lives are "oriented" or guided by certain combinations of letters similar to our own personal name (!). For example, that if there are more people with letter-oriented predispositions towards a particular letter combinations (for living in certain places or having certain jobs or even voting certain presidential candidate names), this could have massive societal impacts. Say, vote according to letter-oriented chains of ideas where C's and H's mean more than B's and O's. Who knows. It seems like we start going too far such that people are no longer capable of actual choices based on thoughtful reflection. I'm still skeptical towards demographical research but interested enough to continue reading about data trends that "require" correlating to one or several causal factors.
Great Post. Mark
I wonder if there has been any research into correlation between these things and city of birth. I was born in a city beginning with K and have noticed over time that I have many friends whose names also begin with K. Also, I am now living in another country beginning with the same letter.
Also, I wonder if there's any research into whether married women seem to have preferences correlating more to their maiden name or their married name.
What if you are attracted to people and things that begin with the same first letter? This has nothing to do with my name, but i recently realized that most of the singers i enjoy listening to begin with the letter J. I have dated 3 guys who begin with the letter J. I never even realized this until recently when I dicrovered that I am attraceted to my boy friends enitials almost as mucha as i am attracted to him. Would this have anything to do with this study?
I know there will always be exceptions to a trend, and I guess I'm one of them in this case. I don't prefer any of the numbers in my birthdate over other numbers (in no particular order, those numbers are 1, 2, 3, and 9, and 0 if you feel like writing in dd/mm/yyyy format). I like all kinds of numbers, but none of them are in my birthdate, except maybe for 2.
As for my initials (SJ), I don't prefer those letters either, at least if the names I give to my fictional characters are any indication. I have very few characters whose names start with S or J, compared with other letters. One story is an exception â the protagonist's name starts with S (though I don't actually like the name and don't think it suits her), her father's name starts with J (and it's a name I like so much I also gave it to a character in another story), and two antagonist's names start with J, and one character's nickname starts with "S". It might have something to do with the characters being French â some names I normally wouldn't use sound better in French. XD
Most of my characters' names start with K, C, L, I, H, M, A, or T. I really like the letter K, and that's not to be found anywhere in my name or any of my relatives' names. In one of the stories I'm writing, there's an entire family whose first names start with K, and their last name starts with K too. I also have characters in other stories with K names, and they're among my favourite names.
And the Japanese name I'm going to use if I ever become a citizen of Japan has the initials "NM" (my username here is a shortening of it) â no relation to my real name or initials. I just like the meaning and how it sounds.
Oh God, I just realized, my first name starts with an S, and I'm in college to become a SOCIAL WORKER!!!! (Except that I want to go into psychology after that, and I also want to be a writer, so I guess that doesn't really count.)
And I like Skittles and Starburst, but not nearly as much as other junk food. (I'm a chocoholic.) I almost never buy them. I like soy milk but that's because I've drunk it since I was a little kid. :P