Does success depend on your surname?
When I read about the recent developments in the European Parliament about Buttiglione, one remarkable fact is that all people directly involved have the ‘B’ as the first letter in their surname: candidate chairman of the European Commission Barroso, the two Italians Buttiglione and Berlusconi, and the current leader of the European Union, Prime Minister Balkenende (and Dutch foreign minister Bot, too).
This reminded me of an article I read lately in a pretty lousy Dutch magazine called “Psychologie Magazine” (Psychology Magazine), which claimed that having a surname with a letter that comes early in the alfabet increases your chances for success in life (which would not be so good for me with my ‘V’-surname…). As an example, they mentioned the fact that 6 out of 7 G7 member leaders have names in the first half of the alfabet. A possible explanation would be that when they attended school, people with ’small’ surnames would on the top of alfabetical lists, thus getting more attention.
So, is this a statistically feasible conclusion? There are at least three problems with the example mentioned above. First, 7 people is a much too small group to draw conclusions from. Second, this is an obvious case of data snooping: the example is used to form the thought that there might be a correlation, so it cannot be used to prove the correlation anymore: otherwise, pretty much anything could be proven. Third, there is no default group: if you don’t know what the normal chances are of having a surname with first letter from the first half of the alfabet, you can’t say whether a 6/7 score is good of bad. You need to keep in mind here that the second half of the alfabet contains some seldom-used letters such as q, x, y and z.
So, I decided to do a little statistical test myself. I formulate the following thesis:
There is a significant correlation between being a part of group X and having a first surname letter in the first half of the alfabet.
For the group X, I examined two groups that could be seen as groups of succesful people: inspired by the G7 example, I chose the 150 members of the Dutch parliament, and the Top-96 of richest people in the Netherlands from the Quote 500 list.
With these groups, I did a binomial test to see whether there was a significant correlation (i.e. the chance of having a small surname when in group X is significantly larger than the default chance). To calculate the default chance, I counted the names (well, pages of them, anyway 
of the telephone directory of Groningen. Here’s the results:
| Group | Test results | Ratio |
|---|---|---|
| Groningen | 204.75/344.25 | 0.5977 |
| Dutch Parliament | 93/150 | 0.62 |
| Quote 500 ‘01 | 56/96 | 0.5833 |
As can be seen, the ratio in the Quote list is smaller than in the default group, so in that case, we can directly reject our thesis.
For the Dutch Parliament group, we run a Binomial Test: we calculate the chance of getting the given test results supposing the true ratio is 0.5977. A common way of testing is to reject the thesis if this chance is larger than 5%: the test result is not unlikely enough to reject the default hypothesis. In our case, the chance of having 93 or more MPs with ’small’ surnames is 29.405%, so this is by far not significant enough for our hypothesis.
In conclusion, these test results, fortunately for me, give no indication about assuming there is a correlation between success and having a surname that begins with a letter from the first half of the alfabet. Since the Psychologie Magazine did not point to serious research proving otherwise, I have to suspect the correlation is just a cock-and-bull story.
In the mind-staggering “Currently Listening To”-series, today we feature the