Kevin Rarely Gets “Très Bien”


The graph is from French sociologist Baptiste Coulmont. It shows the proportion of students with a given name who received a “très bien” (meaning “very good”) on their baccalaureat, which is an exam French students take at the end of high school to qualify for university studies. The vertical axis plots the frequency of the name in the data (click on graph to enlarge).

What jumps out is the high frequency of English language names on the left hand side of the picture. People named Kevin, Anthony, Jordan, Cindy, or Dylan were much less likely to receive high scores. Although there is some evidence that names can affect how children perform in school, this more likely reflects naming preferences: parents in lower social classes are more likely to name their kids after characters from American tv shows or music groups than parents from higher social classes. Here is a response from one Kevin (in French) who is just fine.


12 Responses to Kevin Rarely Gets “Très Bien”

  1. Baptiste July 10, 2013 at 7:32 am #

    I’ll be happy to answer any questions about the data and/or the methodology.

  2. Andrew Gelman July 10, 2013 at 9:20 am #

    Nice graph. Maybe it would make sense to show the boy and girl names in different colors? Also, I’m wondering if we could make more use of the y-axis. Right now the y-axis is name popularity, but that information could be conveyed via font (or, perhaps more cleanly, by breaking the plot into three graphs corresponding to common, moderate, and rare names). What other info is available on the prenoms that could be used for the y-axis?

  3. Total July 10, 2013 at 9:47 am #

    Do the graders know the names of the student whose exam they are grading?

  4. Baptiste July 10, 2013 at 10:34 am #

    @Total : The graders do not know the name of the students. Nor do they know the highschool in which the Bac was prepared.

    @Andrew : Boys and girls in different colors : that could be done (Camille is a androgynous name, but the other ones are not).
    The database is structured thusly :
    LASTNAME Firstname1 Firstname2 Firstname3, YearOfBirth, MonthofBirth, Highschool, Type of Bac (S,L,ES,STG…), Grade (“Second chance”, “OK”, “somewhat good”, “good”, “very good”)

    I have the same data for 2012 (N=340000) and 2011 (without the YearOfBirth and MonthOfBirth) : I could use the rank difference between 2012 and 2013 as a y-axis (some prénoms are gaining popularity, some are losing popularity.

  5. JRLRC July 10, 2013 at 3:45 pm #

    “Following” Voeten, I would say, in general, the cause -or the heaviest causal factor- of the performance is not (directly) the name itself but the cause of the name -or one of the relative causal factors. What would be “causing” the scores is what “caused” the name of the students recieving them.

  6. Total July 10, 2013 at 5:03 pm #

    “Total : The graders do not know the name of the students. Nor do they know the highschool in which the Bac was prepared.”


  7. Widmerpool July 11, 2013 at 11:12 am #

    At the risk of revealing my parochialism, I will express my surprise that no one seems to name their little boy Jacques anymore.

  8. Andras July 11, 2013 at 11:34 am #

    I’d love to see a similar chart for United States SAT scores. I wonder if the raw data is available.

    • Baptiste July 11, 2013 at 11:49 am #

      > Widmerpool : Jacques is indeed too oldfashioned to be used
      > Andras : that would be very interesting indeed

      • Thomas B July 11, 2013 at 2:02 pm #

        Wow, names in France have maybe a 7 year shelf life.

        Seems like US names linger around 25 years, glancing at Robert and James, though Michael has hung around for about 50:

        I wonder what drives name churn… I would say most of the common US names are biblical, that religiosity might be an indicator, but Robert is just a random name of German origin, and it hung on just as easily. (Maybe Robert Frost or Robert Mitchum or Robert F. Kennedy drove that, though?)

  9. Baptiste July 11, 2013 at 3:02 pm #

    This will please (partially) Andrew Gelman : boys and girls are now colored in blue (girls) and yellow (boys), and the names are shuffled (a bit) in order not to overlap.

  10. Rem July 11, 2013 at 6:55 pm #

    Wow, thanks for the male/female version. Am I right that close to the top twenty names are all female? Is there a sex difference and does it also show up on math and science if you break this down by subject?