Search Results for 'GSS'


Robin Hanson responded to a Washington Post article (possibly inspired by Spotted Toad) on the increasing percentage of males 18-29 who report not having sex in the past year with some speculation on whether that was attributed to women of that age group (who reported a smaller increase in celibacy) shifting toward older men or to that subset of 18-29 year old men with more partners. It struck me that since the source of this data was the General Social Survey, which asks respondents their age as well as the number of partners, it should be answerable directly rather than guesses from respondents to a twitter poll. My initial attempt to do so was stymied by a newer GSS interface which generated errors when I tried to construct variables, but an anonymous commenter elsewhere pointed me toward the old interface which was still working. The parameters I used were as follows:
Row: PARTNERS
Column: AGE(r:18-29; 30-39; 40-49; 50-59)
Control(s): YEAR
Selection filter(s): SEX(1), YEAR(2008, 2010, 2012, 2014, 2016, 2018), NUMMEN(0), PARTNERS(0-8) (more…)

Agnostic has a post up in which he uses the GSS to look into some stats on gun-ownership, which inspired me to do the same to investigate some questions he might be interested in. The variables are OWNGUN and MARRIED, with SEX as the control variable. (more…)

I relayed that claim from Greg Cochran a while back, but inspired by this Half Sigma post I decided to present the actual numbers from the GSS.

Row: CHILDS

Column: WORDSUM

Control: SEX

Summary Statistics for SEX = 1(MALE)
Eta* = .06 Gamma = -.02 Rao-Scott-P: F(80,38080) = 2.10 (p= 0.00)
R = -.04 Tau-b = -.01 Rao-Scott-LR: F(80,38080) = 2.02 (p= 0.00)
Somers’ d* = -.01 Tau-c = -.01 Chisq-P(80) = 213.72
Chisq-LR(80) = 206.45
*Row variable treated as the dependent variable.
Summary Statistics for SEX = 2(FEMALE)
Eta* = .12 Gamma = -.08 Rao-Scott-P: F(80,38080) = 4.34 (p= 0.00)
R = -.11 Tau-b = -.06 Rao-Scott-LR: F(80,38080) = 3.85 (p= 0.00)
Somers’ d* = -.06 Tau-c = -.06 Chisq-P(80) = 485.56
Chisq-LR(80) = 430.81
*Row variable treated as the dependent variable.
Summary Statistics for all valid cases
Eta* = .08 Gamma = -.05 Rao-Scott-P: F(80,38080) = 4.70 (p= 0.00)
R = -.08 Tau-b = -.04 Rao-Scott-LR: F(80,38080) = 4.33 (p= 0.00)
Somers’ d* = -.04 Tau-c = -.04 Chisq-P(80) = 522.38
Chisq-LR(80) = 481.45
*Row variable treated as the dependent variable.

I’ve come across that “Firepower” character in the comments section at Overcoming Bias, there as well he seemed quite sure of himself in the absence of supporting evidence. (more…)

I’ll copy my comment from bhtv below the fold. Like Glenn Loury (usually one of my favorites), I appreciate Mark’s forebearance. I was very glad he stuck for science and evidence over the construction of political narratives. Glenn’s heart cries over the harm our criminal justice system does, how can he then turn away from a serious examination of how to improve it and the lives of the people it affects? Yes, people have political agendas, but it is not mere pretense of science to look at politicized issues through a reductionist lens. Not even Loury seems to dispute that Kleiman is a liberal, why look his gift-horse in the mouth and turn the question to politics? Adopting a Peter Singer type utilitarian stance and Glenn’s avowed (and I think sincere) priorities, couldn’t we argue that Loury is in fact obligated to spend a lot more time shouting from the roof tops that we need to find out whether a “magic bullet” has in fact been discovered, or disprove it so we can start investigating other avenues? Loury says in effect “Yeah, yeah, yeah, viewers of bhtv have heard about your book already, you can stop talking about it”. To that I say, hell no! The problems discussed have not been fixed, H.O.P.E type programs still receive very limited funding, and the political push to change that is still weak. It’s similar logic (via an economist) that says you should keep donating to the charity you think does the most good because you haven’t done enough to solve the problem yet. (more…)

I was talking to some Indians about their regional differences in cuisine, and expressed some interest in south indian food since I hadn’t eaten it. They replied that I probably wouldn’t like it since “south indian” food by its nature is vegetarian, even though many south indians themselves are not. I tried looking up an explanation for why that would be and found a pakistani forum where it was claimed that 84 percent of south indian households are non-vegetarian, whereas as the lowest rates of non-vegetarianism are in the north at 53 per cent (east was 94 and west was 58, with an overall percentage of nearly 71). So if south indians are unusually likely to NOT be vegetarians, it makes it very curious that their archetypical cuisine is vegetarian. Indian restaurants I’ve seen here also seem to reverse the reported representation of fish vs mutton vs chicken, although it could be that I’ve just ignored the seafood since I’m into turf than surf.

UPDATE: Agnostic sometimes talks about vegetarians really being unhealthy sugar-stuffers, and so I figured I’d check out the GSS to see the correlation between NOMEAT and INTRWGHT. Unfortunately, those are for disjoint sets of years. I then decided to try googling, and the first few results were complaints on some forum about an inability to find data! Fortunately I found this.

UPDATE: Audacious Epigone tackled this better a few years back.
origen01 engaged in some thread necormancy on an old post, writing that blacks have admiration rather than vitriol toward whites. As you might have guessed, this sent me to the GSS. There is a question FEELWHTS which reads as “In general, how warm or cool do you feel towards white or Caucasian Americans?” Here are the results by race.
(more…)

An anonymous commenter at Sailer’s wrote “[H]ow many straight men with Ivy League educations enjoy the opera? […] In my estimation, the type of man who enjoys the opera is typically gay and is therefore definitely not the type who would want to marry a woman with an Ivy League education.”
Again, this seems amenable to the gss.
Row: OPERA
Column: SEXSEX
Control: SEX
(more…)

Severn at Sailer’s wrote “Educated women tend not to have high IQ’s. The subjects they take in college do not require it. I’d say the typical “communications major” is not any more intelligent than a plumber, and is probably a good deal less so.” That sounded like something that could be investigated in the GSS.

Row: WORDSUM
Column: EDUC(r: “LT HS” 1-11; “HS” 12; “SC” 13-15; “C” 16; “GT C” 17-22)
Control: SEX
(more…)

I haven’t done a GSS analysis in a while, and since Sister Y assumed I had, that provides the angle. The initial post was about Gordon Gallup’s theory for the evolution of homophobia, but Sister Y thought I had shown surprisingly higher numbers of children among men who report male sexual partners. I linked to the Inductivist over there, but I’ll see if I can replicate his results with the GSS.

Mean number of children by gender and sexual partners EXCLUSIVELY MALE BOTH MALE AND FEMALE EXCLUSIVELY FEMALE
MALE .83 1.11 1.73
FEMALE 1.90 .96 1.20

Gallup theorizes that parents who kept their children away from gays were more likely to avoid having their children become gay and thus have more grandchildren. Wilkinson complains that no evidence was provided for that being the case (only that people behave as if it were), which leaves open the possibility that the behavior, while evolved (heritability should be examined), is adaptation-executing rather than fitness-maximizing. It’s not a pure physiological reflex, but something that requires some higher brain functioning (at least enough to respond to hypotheticals given by the experimenter). An idea as simple as desire that one’s children associate with role-models desirable of emulation could be at work even if in some particular case the child’s behavior is unaffected. On the other hand, if Cochran’s theory is right then orientation may be literally contagious at certain stages of development.

On an unrelated note: Frances Wooley finds an anti-feminist subtext in “dumb men” ads.
UPDATE:
George Weinberg was surprised that some of the “exlusives” managed to have children at all. So I should clarify that the above was based on partners in the previous year. There is another question for the past five years.

Mean number of children by gender and sexual partners EXCLUSIVELY MALE BOTH MALE AND FEMALE EXCLUSIVELY FEMALE
MALE .76 1.08 1.72
FEMALE 1.90 1.12 1.22

At Overcoming Bias, Doug argued that the rich tend to be rural despite the image of urban “SWPLs”. That didn’t sound quite right to me, so I decided to check the GSS.

Frequency Distribution
Cells contain:
-Column percent
-Weighted N
SRCBELT
1
12 LRGST SMSA’S
2
SMSA’S 13-100
3
SUBURB, 12 LRGST
4
SUBURB, 13-100
5
OTHER URBAN
6
OTHER RURAL
ROW
TOTAL
W
E
A
L
T
H
1: Less than $5,000 12.8
12.3
17.5
28.8
8.3
17.0
10.5
26.6
12.1
73.1
13.0
21.5
12.0
179.3
2: $5,000 to $20,000 12.5
12.0
14.5
23.9
11.1
22.6
10.2
25.8
12.6
76.4
15.1
24.9
12.5
185.6
3: $20,000 to $40,000 9.8
9.4
14.1
23.2
2.6
5.3
6.6
16.8
9.7
59.0
8.3
13.6
8.6
127.3
4: $40,000 to $75,000 11.7
11.3
11.1
18.4
10.3
21.0
7.2
18.1
8.1
49.1
8.3
13.7
8.8
131.7
5: $75,000 to $100,000 6.5
6.2
7.2
11.9
10.3
21.1
5.6
14.2
6.1
37.2
7.3
12.0
6.9
102.7
6: $100,000 to $150,000 12.6
12.1
4.7
7.7
4.7
9.6
6.9
17.4
8.5
51.3
9.2
15.2
7.6
113.4
7: $150,000 to $250,000 13.4
12.9
9.6
15.9
8.0
16.2
15.1
38.2
13.4
81.4
11.0
18.1
12.3
182.7
8: $250,000 to $500,000 9.7
9.3
15.7
25.9
21.9
44.7
18.9
47.7
16.5
100.2
14.2
23.5
16.9
251.2
9: $500,000 to $1 million 8.4
8.1
4.9
8.0
14.8
30.1
14.7
37.2
9.7
58.6
9.3
15.3
10.6
157.3
10: $1 million to $2 million 1.9
1.9
.5
.9
4.2
8.5
3.5
8.9
2.6
15.7
4.0
6.6
2.9
42.5
11: $2 million to $3 million .0
.0
.3
.4
.9
1.9
.2
.4
.4
2.2
.0
.0
.3
5.0
12: $3 million to $4 million .0
.0
.0
.0
2.3
4.8
.0
.0
.0
.0
.3
.4
.4
5.2
14: $5 million to $10 million .0
.0
.0
.0
.6
1.3
.0
.0
.1
.9
.0
.0
.1
2.1
15: Above $10 million .9
.9
.0
.0
.0
.0
.5
1.3
.1
.9
.0
.0
.2
3.0
COL TOTAL 100.0
96.4
100.0
165.1
100.0
204.2
100.0
252.7
100.0
605.9
100.0
164.8
100.0
1,489.1
Means 5.01 4.42 6.12 5.83 5.28 5.11 5.36
Std Devs 2.83 2.71 3.01 2.91 2.86 2.89 2.91
Unweighted N 95 171 180 229 601 188 1,464
Color coding: <-2.0 <-1.0 <0.0 >0.0 >1.0 >2.0 Z
N in each cell: Smaller than expected Larger than expected
Summary Statistics
Eta* = .17 Gamma = .00 Rao-Scott-P: F(65,2600) = 1.61 (p= 0.00)
R = .02 Tau-b = .00 Rao-Scott-LR: F(65,2600) = 1.54 (p= 0.00)
Somers’ d* = .00 Tau-c = .00 Chisq-P(65) = 114.53
Chisq-LR(65) = 109.10
*Row variable treated as the dependent variable.
Text for ‘WEALTH’

1092. Please estimate your total wealth. IF ASKED: Wealth means 
the value of your house plus the value of your vehicles, stocks 
and mutual funds, cash, checking accounts, retirement accounts 
including 401(k) and pension assets, and any other assets minus 
what you owe for your mortgage and your debts.

I was actually surprised at the rarity of the wealthy in the biggest metropoli.

Sister Y suggested that smart people may use drugs & have sex because they are bored more often. The GSS has two variables, BOREDOM and BORED. Only the first one actually asks about boredom per se, but I report results for both.

Frequency Distribution
Cells contain:
-Column percent
-Weighted N
WORDSUM
1 2 3 4 5 ROW
TOTAL
BOREDOM 1: ALWAYS 35.3
3.1
.0
.0
.0
.0
2.9
1.5
1.2
1.0
3.0
5.6
2: OFTEN .0
.0
.0
.0
.0
.0
5.9
3.1
9.0
7.7
5.6
10.7
3: SOMETIMES .0
.0
50.0
9.2
34.0
8.7
41.2
21.5
34.7
29.6
36.3
68.9
4: HARDLY EVER 11.8
1.0
22.2
4.1
18.0
4.6
24.5
12.8
26.3
22.5
23.7
44.9
5: NEVER 52.9
4.6
27.8
5.1
48.0
12.3
25.5
13.3
28.7
24.5
31.5
59.8
COL TOTAL 100.0
8.7
100.0
18.4
100.0
25.5
100.0
52.1
100.0
85.3
100.0
190.0
Means 3.47 3.78 4.14 3.64 3.72 3.75
Std Devs 1.97 .88 .91 1.03 1.02 1.05
Unweighted N 7 15 27 51 81 181
Frequency Distribution
Cells contain:
-Column percent
-Weighted N
WORDSUM
6 7 8 9 10 ROW
TOTAL
BOREDOM 1: ALWAYS 1.5
2.0
.0
.0
1.3
1.0
2.4
1.0
3.2
1.0
1.4
5.1
2: OFTEN 6.5
8.7
2.9
2.6
3.9
3.1
11.0
4.6
6.5
2.0
5.6
20.9
3: SOMETIMES 34.4
46.0
32.6
28.6
32.9
25.5
32.9
13.8
32.3
10.2
33.3
124.1
4: HARDLY EVER 33.2
44.4
40.7
35.8
32.2
25.0
31.7
13.3
35.5
11.2
34.8
129.7
5: NEVER 24.4
32.7
23.8
20.9
29.6
23.0
22.0
9.2
22.6
7.2
24.9
93.0
COL TOTAL 100.0
133.8
100.0
87.8
100.0
77.6
100.0
41.9
100.0
31.7
100.0
372.8
Means 3.73 3.85 3.85 3.60 3.68 3.76
Std Devs .96 .82 .94 1.03 1.01 .94
Unweighted N 128 84 78 45 32 367
Summary Statistics
Eta* = .16 Gamma = -.03 Rao-Scott-P: F(40,1680) = 2.66 (p= 0.00)
R = -.04 Tau-b = -.02 Rao-Scott-LR: F(40,1680) = 1.74 (p= 0.00)
Somers’ d* = -.02 Tau-c = -.02 Chisq-P(40) = 89.93
Chisq-LR(40) = 58.96
*Row variable treated as the dependent variable.
Text for ‘BOREDOM’ 

1336. Now some more questions about your working conditions.
Please circle one code for each item below to show how often 
it applies to your work.
 How often... d. Are you bored at work?

Now for free-time.

Frequency Distribution
Cells contain:
-Column percent
-Weighted N
WORDSUM
1 2 3 4 5 ROW
TOTAL
BORED 1: QUITE OFTEN 20.8
7.6
26.7
16.3
22.7
27.5
17.6
35.0
20.2
66.4
20.5
152.7
2: NOW AND THEN 37.3
13.6
39.5
24.1
46.8
56.8
43.4
86.2
40.5
133.0
42.0
313.7
3: ALMOST NEVER 41.9
15.3
33.8
20.6
30.5
37.0
39.0
77.6
39.3
129.3
37.5
279.8
COL TOTAL 100.0
36.5
100.0
61.0
100.0
121.3
100.0
198.7
100.0
328.8
100.0
746.2
Means 2.21 2.07 2.08 2.21 2.19 2.17
Std Devs .77 .78 .73 .72 .75 .74
Unweighted N 42 81 132 213 319 787
Frequency Distribution
Cells contain:
-Column percent
-Weighted N
WORDSUM
6 7 8 9 10 ROW
TOTAL
BORED 1: QUITE OFTEN 11.0
44.9
9.3
32.0
4.7
8.9
8.5
13.2
4.6
5.0
8.6
104.0
2: NOW AND THEN 49.1
200.9
37.3
128.4
43.5
82.5
38.4
59.7
26.7
29.0
41.5
500.5
3: ALMOST NEVER 39.9
163.2
53.4
183.8
51.8
98.4
53.1
82.7
68.7
74.6
49.9
602.7
COL TOTAL 100.0
409.0
100.0
344.1
100.0
189.8
100.0
155.6
100.0
108.6
100.0
1,207.1
Means 2.29 2.44 2.47 2.45 2.64 2.41
Std Devs .65 .66 .59 .65 .57 .64
Unweighted N 391 308 194 147 99 1,139
Summary Statistics
Eta* = .21 Gamma = .23 Rao-Scott-P: F(20,1960) = 3.89 (p= 0.00)
R = .20 Tau-b = .16 Rao-Scott-LR: F(20,1960) = 3.88 (p= 0.00)
Somers’ d* = .14 Tau-c = .18 Chisq-P(20) = 112.07
Chisq-LR(20) = 111.84
*Row variable treated as the dependent variable.
Text for ‘BORED’ 

285a. If sometimes or never: How often would you say you have time
 on your hands that you don't know what to do with?

Dain referenced such an association it in the comments. Here’s what the GSS reports:

Frequency Distribution
Cells contain:
-Column percent
-Weighted N
WORDSUM
1 2 3 4 5 ROW
TOTAL
HLTH5 1: YES .0
.0
.0
.0
.0
.0
2.1
1.1
4.1
3.2
2.5
4.2
2: NO 100.0
8.5
100.0
19.6
100.0
15.4
97.9
50.5
95.9
74.9
97.5
168.9
COL TOTAL 100.0
8.5
100.0
19.6
100.0
15.4
100.0
51.5
100.0
78.1
100.0
173.1
Means 2.00 2.00 2.00 1.98 1.96 1.98
Std Devs .14 .20 .16
Unweighted N 10 19 14 52 82 177
Frequency Distribution
Cells contain:
-Column percent
-Weighted N
WORDSUM
6 7 8 9 10 ROW
TOTAL
HLTH5 1: YES 4.3
4.2
4.3
3.7
2.2
1.1
4.9
2.1
.0
.0
3.6
11.2
2: NO 95.7
95.6
95.7
82.8
97.8
47.3
95.1
40.9
100.0
35.1
96.4
301.6
COL TOTAL 100.0
99.8
100.0
86.6
100.0
48.3
100.0
43.0
100.0
35.1
100.0
312.8
Means 1.96 1.96 1.98 1.95 2.00 1.96
Std Devs .20 .20 .15 .22 .19
Unweighted N 99 82 49 42 38 310
Summary Statistics
Eta* = .09 Gamma = -.07 Rao-Scott-P: F(9,378) = .88 (p= 0.54)
R = -.02 Tau-b = -.01 Rao-Scott-LR: F(9,378) = 1.37 (p= 0.20)
Somers’ d* = .00 Tau-c = -.01 Chisq-P(9) = 4.33
Chisq-LR(9) = 6.72
*Row variable treated as the dependent variable.
Text for ‘HLTH5’

1575. Now, I'm going to ask you about various events and 
conditions that happen to people. I'm interested in those that 
happened to you during the last 12 months, that is since 
(CURRENT MONTH), 2003. As I ask you about the specific events, 
please think carefully, so I can record things accurately. a. 
First, thinking about health related matters, did any of the 
following happen to you since February/March, 1990? 5. Used 
illegal drugs (e.g. marijuana, cocaine, pills)
Frequency Distribution
Cells contain:
-Column percent
-Weighted N
WORDSUM
1 2 3 4 5 6 7 8 9 10 ROW
TOTAL
HLTH5 1: YES .0
.0
.0
.0
.0
.0
2.1
1.1
4.1
3.2
4.3
4.2
4.3
3.7
2.2
1.1
4.9
2.1
.0
.0
3.2
15.4
2: NO 100.0
8.5
100.0
19.6
100.0
15.4
97.9
50.5
95.9
74.9
95.7
95.6
95.7
82.8
97.8
47.3
95.1
40.9
100.0
35.1
96.8
470.5
COL TOTAL 100.0
8.5
100.0
19.6
100.0
15.4
100.0
51.5
100.0
78.1
100.0
99.8
100.0
86.6
100.0
48.3
100.0
43.0
100.0
35.1
100.0
485.9
Means 2.00 2.00 2.00 1.98 1.96 1.96 1.96 1.98 1.95 2.00 1.97
Std Devs .14 .20 .20 .20 .15 .22 .18
Unweighted N 10 19 14 52 82 99 82 49 42 38 487
Color coding: <-2.0 <-1.0 <0.0 >0.0 >1.0 >2.0 Z
N in each cell: Smaller than expected Larger than expected
Summary Statistics
Eta* = .09 Gamma = -.07 Rao-Scott-P: F(9,378) = .88 (p= 0.54)
R = -.02 Tau-b = -.01 Rao-Scott-LR: F(9,378) = 1.37 (p= 0.20)
Somers’ d* = .00 Tau-c = -.01 Chisq-P(9) = 4.33
Chisq-LR(9) = 6.72
*Row variable treated as the dependent variable.
chart illustrating table
Text for ‘HLTH5’

1575. Now, I'm going to ask you about various events and 
conditions that happen to people. I'm interested in those that 
happened to you during the last 12 months, that is since 
(CURRENT MONTH), 2003. As I ask you about the specific events, 
please think carefully, so I can record things accurately. a. 
First, thinking about health related matters, did any of the 
following happen to you since February/March, 1990? 5. Used 
illegal drugs (e.g. marijuana, cocaine, pills)

I’m surprised by the low overall percentage of people that admit to using drugs, particular those with the lowest wordsum scores.

In a comment at Overcoming Bias, Michael Vassar responded to the claim that in our society on average the wealthy are least promiscuous: “If you have spent time with middle class and wealthy people, even a little time, that claim is obviously false. The wealthy are MUCH more promiscuous in my observation.”
I decided to check the GSS. (more…)

Agnostic claimed most whites are friendly towards it. I decided to check the GSS.

(more…)

Mark Kleiman was asked why European countries have abolished the death penalty while America and Asia continue to use it. His theory was that “a Westminster system gives well-educated civil servants more power, and the well-educated mostly dislike capital punishment”. The second part of that theory seemed testable, so I checked it out in the GSS, running CAPPUN (supporting the death penalty for murderers is coded as 1, opposing it is 2) by EDUC (years of education). First the aggregate statistics.

Summary Statistics
Eta* = .11 Gamma = .02 Rao-Scott-P: F(20,3820) = 17.89 (p= 0.00)
R = .00 Tau-b = .01 Rao-Scott-LR: F(20,3820) = 17.19 (p= 0.00)
Somers’ d* = .01 Tau-c = .01 Chisq-P(20) = 498.91
Chisq-LR(20) = 479.30
*Row variable treated as the dependent variable.

Then broken down by years of education. (more…)