Should we practice glacial profiling?

If a fat job applicant waddles into your office, should you assume he is lazy?

The MIDUS Study measured the body mass index (BMI) as well as personality characteristics of a large sample of adults. Here are the BMI/personality correlations for men and women:

Correlations

Men (white, ages 40-59, n = 830)
Conscientiousness -.07
Achievement -.07
Control -.10

Women (white, ages 40-59, n = 978)
Conscientiousness -.21
Achievement -.12
Control -.06

The correlations are weak; they are a bit stronger for women. So there is a slight tendency for heavier people to be less achieving, less conscientious, and lacking in self-control, but no more than a slight tendency.

Religion and finding fault

One of the tenets of my religion that I like is self-criticism. It checks the human tendency to focus on the faults of others by saying, "Okay, but what about you, pal?" In the last post, I railed against extreme liberals for their dirtiness, but then looked around and had to admit my place wasn't looking so hot. My impulse is to vanquish enemies, but the message is always in my head that the world that needs fixing is inside. Self-esteem is thought to be so important nowadays, but the attitude of "I'm a sinner--wise up" does me a lot of good.

The General Social Survey asked respondents if they tend to find fault in others. Answers ranged from "strongly disagree" (1) to "strongly agree" (5).  Here are the means by religion (sample size = 1,505):

Mean "finding fault" score

Protestant 2.50*
Catholic 2.58*
Jewish 2.88
No religion 2.76

*significantly lower than those with no religion


Christians are significantly less likely than people with no religion to see themselves as tending to find fault in others. The Jewish sample size is very small which works against reaching statistical significance, but the phrase that popped into my head at seeing the high mean was "culture of critique."  I should mention the usual caveats that the numbers do not identify cause and effect, nor do we know the validity of the measure.


 

Who's the laziest?

GSS respondents were asked to rate the laziness of various groups, giving them a score somewhere between hardworking (1) and lazy (7). The mean ratings given by blacks and whites are shown above (sample size = 1,226).

Blacks and whites rank other groups differently. For blacks, the laziest to the least lazy goes like this: Hispanics, blacks, whites, southern whites, Jews, and Asians. For whites, it's blacks, Hispanics, southern whites, whites, Asians, and Jews. Based on my own rankings, it would appear that I am a white person.  

It's interesting, given that immigration enthusiasts love to advertise the work ethic of Hispanics, that ordinary blacks and whites perceive them to be comparatively lazy. And you can't respond that blacks and whites are completely self-serving in their rankings: both groups feel that Jews and Asians are harder working than they are.  

Social class and social potency

A social potency score was calculated for MIDUS Study participants based on the following items: is forceful and decisive; is persuasive and likes to influence others; enjoys or would enjoy leadership roles; enjoys being noticed, being the center of attention. Here are the means by sex and social class:

Mean social potency score

Males (n = 443)
Low 9.68*
Middle 10.19*
High 10.63*

Females (n = 745)
Low 8.86
Middle 9.69*
High 10.09*

*significantly higher than low-status females

All groups are significantly more socially potent than lower-class women. The mean for high-status men is more than four-tenths of an standard deviation (sd) higher than the mean for low-status men--a moderate difference. And the gap between low- and high-status women is also of moderate size--one-half of an sd.

According to this study, the heritability of social potency is 55 percent.

So folks from higher classes are more dominant than others. Just like the last post about high-status people being more driven, this is no surprise. But to extreme nurturists (i.e., just about every professor I ever had) where you end up in the stratification system has little or nothing to do with innate abilities or personality traits. IQ is only a proxy for parents' social class. People are chameleons who adopt the personality of their social position.   

Sexual orientation and harm avoidance

I calculated mean harm avoidance scores as described in the last post by sex and sexual orientation. (N = 3,546, whites only)

Mean harm avoidance score

Males
Straight 11.18
Gay 10.97
Bisexual 10.82

Females
Straight 12.74
Lesbian 12.24
Bisexual 11.92

Compared to straight guys, gay and bisexual men have lower harm-avoidance scores (the difference is trivial, however). Lesbians and bisexual women also have lower means than their straight counterparts. The female straight-bisexual gap is about one-third of a standard deviation--a moderate difference. None of the female groups approach the low danger-averse scores of the male groups.

Race, ethnicity, gender and harm avoidance

A harm avoidance score was calculated for MIDUS Study participants based on the following items: does not or would not enjoy participating in dangerous adventures or activities (e.g., skydiving); being in some natural disaster (e.g., forest fire); being caught in a sudden and dangerous emergency (e.g., a hold up); deliberately risking serious bodily injury; (e.g., riding a runaway horse). Instead prefers safer activities and experiences even if they are tedious or aggravating.  Here are the mean scores by race/ethnicity and sex (N = 3,739):

Mean harm avoidance score

For females, the gap between Norwegian Americans (the highest) and Mexican Americans (the lowest) is eight-tenths of a standard deviation--a large difference. For males, the black-Mexican gap is four-tenths of an sd--a moderate difference.  Mexican Americans are less danger-averse than other groups. It's interesting too that of all males, blacks score the highest on harm avoidance.

The gender gap is largest for Norwegians and smallest for Mex-Ams. Mex-Am women are real outliers here:  they are less danger-averse than most male groups.

Estimates of heritability range from 0.2 to 0.6. 

The overall gender gap is over one-half of a standard deviation. I'll repeat my point from the last post: The person who steps forward to slay the Jabberwocky is probably going to be a dude. Ok--maybe a Mexican chick.

Race, ethnicity, sex, and social potency

A social potency score was calculated for MIDUS Study participants (all Americans) based on the following items: is forceful and decisive; is persuasive and likes to influence others; enjoys leadership roles; enjoys being noticed, being the center of attention. Here are the means by race/ethnicity and sex (N = 3,735):

 
 
 
 
 
 
 
 
 
 
 
 
 
 
*significantly higher than men of English ancestry
**significantly higher than women of English ancestry
 
The mean for Jewish men is sixth-tenths of a standard deviation (sd) higher than the mean for English men--a fairly large difference.  The Jewish-English gap for women is smaller--a third of an sd.
 
The gender gap is largest for Mexican Americans, and smallest for Norwegian Americans.
 
The total male-female gap is about one-third of standard deviation.  According to this study, the heritability for social potency is 55%.   
 
In the new movie, little Alice slays the Jabberwocky (instead of a boy as in the original poem) and then steps into her father's shoes as an international business visionary. The story line is precisely where it belongs: in a fantasy.     

Race and sophistication

MIDUS participants were asked: "How well does the following describe you: sophisticated."  Answers ranged from "not at all" (1) to "a lot" (4).  Here are the means by race/ethnicity (sample size = 3,922):

Mean sophistication score
  
White 1.30
Black 1.74*
Amerindian 1.26
Asian 1.56
Mexican 1.38

*significantly higher than whites

Blacks are half of a standard deviation more sophisticated than whites.

Sexual orientation and social closeness

A social closeness score was calculated for participants in the MIDUS study based on the following items: is sociable; likes to be with people; takes pleasure in and values close interpersonal ties; is warm and affectionate; turns to others for help and comfort. Here are the means by sexual orientation (sample size = 3,850).

Mean social closeness score

Males
Straight 11.45
Gay 11.67
Bisexual 10.79

Females
Straight 12.09*
Lesbian 11.09
Bisexual 11.52

* significantly higher than straight males

The mean for straight females is a bit more than one-quarter of a standard deviation (sd) higher than the mean for straight males--a small difference. If we ignore significance (the lesbian sample is only 23 people) the lesbian mean is four-tenths of an sd lower than the mean for straight women--a moderate difference. Lesbians are less likely to be people people.

This is consistent with my recent finding that, compared to straight women, lesbians are less likely to have emotional intimacy in their sexual relationships.

Race and aggression among children

In a study of 404 eight-year-old boys and girls from Finland (one Finnish- and one Swedish-speaking group) Poland, and Chicago (one white and one black group) blacks were found to be more physically, verbally, and indirectly aggressive (passive-aggressive stuff) than the other groups. This was true for boys and girls, and regardless of whether self- or peer-estimations were used. (The black kids were inner-city).

The racial gap was larger for girls. This is consistent with the analysis of adults in the MIDUS study I described earlier.

According to this study, broad heritability for aggression among adults is around 70%.


(Karin Osterman, Kaj Bjorkqvist, Kirsti M.J. Lagerspetz, Ari Kaukiainen, L. Roweli Huesmann, and Adam Fraczek. 1994. Peer and Self-Estimated Aggression and Victimization in 8-Year-Old Children From Five Ethnic Groups. AGGRESSIVE BEHAVIOR, 20, 411-28.)

Race and well-being

A well-being score was calculated for MIDUS respondents based on the following items: has a cheerful happy disposition; feels good about self; sees a bright future ahead; is an optimist; lives an interesting, exciting life; and enjoys the things he or she is doing. Here are the means by race/ethnicity:

Mean well-being score

White 8.95
Black 9.47*
Amerindian 9.08
Asian 9.50
Mexican 9.06

*significantly higher than whites

The black mean is about three-tenths of a standard deviation higher than the white mean--a small-to-moderate difference. How are these folks so happy in such a terrible country?

According to Tellegen and Lykken, the stable part of happiness is 80% heritable. 

Differences in traditionalism

As a follow-up to the last post, here are the mean traditionalism scores by race/ethnicity (sample size = 3,725):

Mean traditionalism score

Amerindian 9.11
Black 9.00
Norwegian 8.65
Asian Indian 8.64
German 8.52
French 8.46
Scottish 8.44
Mexican 8.43
Irish 8.34
English 8.14
Polish 8.06
Italian 7.88
Swedish 7.66
Jewish 6.88

The gap between Amerindians and Jews is one full standard deviation--a very large difference.

Heritability for traditionalism is about .6.

Race and aggression

An aggression score was calculated for MIDUS respondents based on the following items: is physically aggressive; enjoys upsetting and frightening others; enjoys scenes of violence; victimizes others for own advantage; will retaliate; is vindictive. Here are the means by race/ethnicity (Sample size = 3,392):

Mean aggression scores

Males (n = 1,783)
White 5.72
Black 6.13
Amerindian 5.75
Asian 8.43*
Mexican 5.50

Females (n = 2,209)
White 5.11
Black 5.84*
Amerindian 5.72
Asian 5.74
Mexican 5.61


* significantly higher than white counterpart

The high score for Asian males might be anomalous (there are only 7 respondents) but in an analysis of GSS data, I found that Chinese Americans were most likely to agree with a pro-violence statement. Black men are only two-tenths of a standard deviation above white men--a small difference. On the other hand, the mean for black women is more than four-tenths of a standard deviation above that of white women, a moderate difference. Notice how scores for Amerindian women are similar to their male counterparts, while Mexican American women are higher than the men.

More on recklessness

Here's more evidence that southern whites are more reckless than white northerners. The map (CDC data) shows motor vehicle mortality rates for people living in non-metropolitan areas. The North/South divide is even clearer here than with fatal firearm accidents. It's striking, given that northerners experience much worse weather conditions.

Recklessness among northern and southern whites

Figure 1. Fatal gun accidents

As a follow-up to the last post on racial differences in recklessness, I generated a map of fatal gun accident (FGA) rates by state. On many behaviors (e.g., violence, IQ, and education) southern whites seem to be between northern whites and non-Asian minorities (NAMs).

You can see from the map that the highest rates of FGAs among whites are concentrated in the non-coastal southern states. Idaho, Wyoming, and Montana--big hunting states--are the exceptions. According to the General Social Survey (GSS), the highest percentage of people who hunt are those who live in the mountain states.

High hunting rates might help explain the high FGA levels in the South, but people are just as likely to hunt in the north central region of the country (GSS), and their FGA rates are low.

Do people with certain personalities have more kids?

Using GSS data, I calculated the mean number of children of males and females who were at least 45 years old at the time of the survey. I used five questions that attempt to measure personality traits, and the table displays mean number of kids by possession of the traits.

The only categories that are significantly different are imaginative and unimaginative men. The guys without imaginations have fewer kids. Notice how few men will admit that they lack this trait. Those who do are probably particularly self-aware, or they might be especially unimaginative.

The other comparison that is almost statistically significant is finding fault among women. Females who admit to finding fault in others have fewer kids. As with imagination, few people are willing to admit that they do this.

Keep in mind that the heritabilities of personality traits are roughly 0.5. There is little evidence here of differential fertility across personality types. (Let's also remember that individual items are poor measures of traits).

Behavioral measures of personality

In this old research study (Perceptual And Motor Skills. 1976. Dec; Vol. 43, pp. 843-7), 1,154 pairs of high school students were observed in cafeteria queues. Blacks were found to touch others more frequently than whites. This is consistent with Francis Galton's claim that Africans are affectionate and gregarious. Over the years, my observation has been that blacks are more extraverted, but self-assessment data (GSS) don't back me up.

Behavioral studies, like this one on touching, seem to offer a more valid way to identify personality differences between races. One interesting approach is to study the items in people's offices or homes, as is described in Snoop: What Your Stuff Says about You.   

I can't call it a study, but I have gotten an impression of racial differences from behavior in the classroom. If I measure extraversion as the frequency of speaking up in class--or perhaps approaching me with questions--blacks are the most extraverted, whites are in the middle, and Mex-Ams are the most introverted.

If I measure conscientiousness in terms of the how perfectionistic students are when writing their papers, whites would score the highest, Mex-Ams in the middle, and blacks would come in last. Or if I measured the trait by punctuality and attendance, the racial order would be the same.

If I measure disagreeableness in terms of students arguing with me, blacks rank first, whites second, and Mex-Ams third.

It's not as easy to measure neuroticism, but based on expressing emotional difficulties to me, whites do it the most. Hispanics are less likely to talk to me, period. Blacks talk to me plenty, but are more likely to talk about problems in terms of circumstances, not feelings.