It was just one of those silly “fluff” stories they put on CNN to fill up the hours between commercials and bits of real news. The anchors were reporting on a study that surveyed the most common CEO names. With straight faces, they reported that the most common men’s CEO names were short single-syllable, mostly four-letter names like Jack, Fred, Bill, Peter, Bob, and Bruce. The most common women’s CEO names were Deborah, Debra, Sally, Cynthia, and Carolyn. Their tiny bit of “analysis” of this undigested factoid by an “onomastics specialist” (a scholar who specializes in names) was that “shortened versions of given names are often used to denote a sense of friendliness and openness. Female CEOs, on the other hand, use their full name to project a more professional image.” And, like most other stories on Headline News, CNN, USA Today, and other media which consist of no more than a few sentences of information and some superficial comment or analysis, that was it. No further discussion about why those names tend to be most common among CEOs.
It was such a jaw-droppingly stupid piece of tabloid journalism that I was intrigued to look it up and see what was really behind the story. But the story on the Linked-In site is not much longer or detailed than the reporting on CNN. Many other media on the web picked up the same story and gave it the same superficial level of discussion. All the site did was a simple analysis of correlations between occupations and first names, based on their data set of Linked-In members of professionals who join the site for networking purposes (which skews the result right there—it’s not based on a random sample of names from census data or any large, rigorously collected data set). But is it true, as they imply, that the name that parents give their child somehow determines what kind of profession they might enter? Does a decision most parents make before we are born lock us into certain types of jobs? Are we fated, like the boy named Sue in the Johnny Cash song, to follow a certain path because a parental choices?
No, of course, not! Sadly, not one of the media reporting the story bothered to dig any deeper, or ask any tougher questions about what this apparent correlation might mean. They made the tacit but false assumption that “correlation equal causation” without any real attempt to examine what might have caused the correlation in the first place. The first questions they should have asked is: how different is that frequency of first names compared to the frequency of popular boys’ or girls’ names for the cohort of people born at the same time as most CEOs today? It’s easy to go to the U.S. Census website, type in a year, and find out what the top 10, top 20, or top 100 most popular boys’ and girls’ names were for any given year. Let’s assume that your average CEO today is 55 (you can pick any age for CEOs, who might be expected to be in the late 40s to early 60s now). In 1956, the top 20 boys’ names were all the traditional short ones popular with the parents of the Baby Boomers: Michael, James, Robert, David, John, William, Richard, Mark, Charles, Steven, Charles, Joseph, and Gary make up the top 13 most popular names, and Peter, Fred, and Bruce are all in the top 40. The same goes for girls’ names, with traditional Biblical and Anglo-Saxon names dominating the list, as they do for boys: Mary, Debra, Linda, Deborah, Susan, Patricia, Karen, Cynthia, Donna, Barbara, Pamela, Nancy, and Sharon making up the top 13. Note that both Debra and Deborah were in the top 5 that year, and that Sally, Cynthia, and Carolyn all made the top 40. What would have been much more useful than just a list of the top 5 names would be the raw statistics to see how these names ranked in popularity, or whether if you looked at the top 20 names, you ‘d get a list nearly identical to the ones that the U.S. Census would have for that cohort. No such data are provided. If they had been, one could do a simple statistical analysis to determine whether the frequency of names on their survey is significantly different from a random sample of the most popular names at the time CEOs were born. Now that would have been a meaningful comparison, and if the null hypothesis were falsified (the list of CEO names is statistically distinct from a random sample of the most popular names at that time), then you would have something worth explaining and analyzing. As it stands in the Linked-In website, it’s a factoid without any real meaning, and no conclusions can be drawn from it whatsoever—except that the Linked-In people have no training in statistics or social science studies like this.
The site has some other quirks that reveal a lot about the sampling methodology. cultural biases, and very little about the meaning of name rankings. For example, the top names in the restaurant business were Thierry, Phillippe, and Laurent. Hmmmm… Do parents give their boys French-sounding names to turn them into chefs (as the website seems to imply)? OR is it due to the fact that the high-end restaurant business is dominated by French cuisine, so most chefs are French in cultural heritage? The most popular name in engineering is Rajesh. Do parents not from India pick that name to turn their son into an engineer, or could it be that there are a lot of people of south Asian origins in engineering? In athletics, the five most popular names are Ryan, Matt, Matthew, Jessica, and Jason. Given that athletes would tend to be in their 20s or 30s, wouldn’t you guess that these are among the most popular names among kids born in the 1970s and 1980s? Sure enough, that’s what the U.S. Census site shows.
And then there are real head-scratchers that raise a red flag and make me doubt that this data set has any validity at all. The site reports that the top names in sales are Chip, Todd, and Trey! What? I don’t think I’ve ever met a “Chip” in my life, and the only time I’ve heard of a “Trey” was the character played by Kyle MacLachlan on the Sex and the City HBO series, or South Park animator Trey Parker (whose real name is Randolph Severn Parker III, hence the nickname “Trey” or “third”). Those are really the most popular names in sales today? Sounds more like a suspicious data set to me.