Tuesday, November 07, 2006

Who has the most good friends? I was interested to know which American ethnic groups say they have the most close friends. According to the General Social Survey, this is the mean number reported:

Mean number of close friends (sample sizes in parentheses):

Norwegians 12.7 (32)
Scandanavians 9.7 (60)
Germans 7.8 (198)
English/Welsh 7.7 (198)
Mexicans 7.4 (38)

USA 7.2

American Indians 7.2 (43)
Irish 6.9 (167)
Poles 6.2 (54)
Blacks 6.1 (91)
Italians 5.0 (75)

This is not the pattern I expected. Comments?


  1. I know some Norwegians and have spent time with them in Norway. The men have great ethnic/national solidarity and drink and ski extensively together.

    I have read about the Vikings. Norway has never been invaded (except for WWII) and is proud of all that past history.

  2. Anonymous10:02 PM

    Wasn't Norway ruled by Sweden for an extensive period of time before gaining independence fairly recently?

  3. I guess Garrison Keillor has been lying to us all these years about Norwegians!

  4. In my post on personality at GNXP, there's a list of countries ranked by Eysenck's Psychoticism score -- that's the interaction between low Agreeableness and low Conscientiousness on the Big Five measures. It looks like people in the more Agreeable countries (i.e., those with low Psychoticism) have more friends. Hard to tell how much of the variance is accounted for with so few countries on the friends list, plus Poland isn't listed in the personality survey -- nor are Blacks, but I think that in the US you could safely assign them a score near that of Italians.

    So, you could rank the countries on the friends list for Agreeableness (or low P), and see what Kendall's tau is.

  5. OK, had some free time and checked Kendall's tau. For the rankings of friends & reversed Psychoticism, tau = +0.38 -- not shabby for social science data.

    The friends ranking is as you posted. I used the global personality survey wherever I could, but made the following assumptions: I assigned English/Welsh P = 3 and the Irish P = 4 (UK as a whole listed as 3.5 in the survey -- assumption is that Irish are slightly more disagreeable than the English / Welsh). This didn't affect the rankings, though. For Poles, I used the nearest ethnic group (Western Slav), namely Czechs -- but Czechs are the most disagreeable of all the world, so I dialed the score down to P = 7 just so Poles wouldn't be so extreme. Blacks I assigned P = 8 since they're plainly pretty disagreeable (crime rates, getting into fights or curse-outs in public, etc.). For American Indians, I figured Mexicans are a mix of Amerindians and Spaniards -- Mexicans score P = 4.5 and Spaniards P = 3, so I assigned American Indians P = 5.

    So, here's the ranking by reversed Psychoticism scores, making the above assumptions:

    1, 8, 3, 6, 2, 7, 4, 9, 10, 5 (matched to countries in descending order of # of close friends).

    Kendall's tau then gives correlation of +0.38.

  6. Anonymous ... you are very likely correct. I got carried away. The first paragraph is OK, but their history of formal government is complex, as you suggest.

    I'd like to add to my first paragraph, however. The Norwegians I knew would gather in large restaurants and drink and sing songs, both drinking songs and patriotic songs. And I think this does go continuously back to the Vikings.

    In their working life they are pretty much as stolid and reliable as the stereotype. But it is clear that there is something going on underneath.

    This may pass away if immigration becomes substantial; recent research shows that diversity undermines trust not only between ethnicities but within ethnicities.

  7. Anonymous10:00 PM

    Agnostic, don't lump Poles in with Czechs, they are quite different. The Poles, as evidenced by polka and my trips to the country itself, are very friendly and helpful. Although people bump into each other in the streets without an excuse me, its really because an unintentional bump is nothing to take seriously. They are otherwise very considerate and open. The Czechs are far more uptight and glum. Lumping the two peoples together based on mere geographic proximity is intellectually lazy.

  8. Norway was invaded by Sweden in 1814.

    I would not put too much stock in these data. How do you define "close friend"?

    I also have a hard time beliving Norway is that different from say the US or Germany.

    Clearly scandinavians are more "friend" oriented than "family" oriented than many other cultures. On the other hand they also seem less social in general, so I don't know which effect dominates.

  9. Dear Inductivist:

    Perhaps you should post sample size for each group along with scores. Maybe there are only, say, 20 or so Norwegians in the sample, so it's not worth coming up with ad hoc theories about the amiability of Norwegians?

  10. Steve: I added sample sizes to the list. My rule of thumb is a minimum of 30 cases, and means are more reliable, of course, than percent distributions. I'll combine Scandanavians to increase sample size and see what we get.

  11. A lot of US Scandinavians are in the Dakotas where there's less crowding. It could be that people in less crowded conditions form more friendships and so we are seeing the effects of geography. See my post Less Densely Packed People Friendlier Toward Neighbors.


Meta-analysis of clinical trials: Eat walnuts

I am always looking for easy eating choices that are good for you. This new meta-analysis of 26 clinical trials looked to see if walnuts ma...