Cause and effect on myopia

Photo PNC / https://flic.kr/p/f2uKgJ

It’s funny for a non-eye specialist to read articles about myopia, which in my line of work rarely means myopia, literally, which is nearsightedness. Takes some getting used to.

Anyway, in my book I use as an example of misleading correlations the link between night lights and myopia in children. Checking it to make sure it is still a good example to keep for the second edition, I was glad to see that it holds up well.

Here’s the story. In a 1999 paper (paywalled | sci-hub), Quinn and colleagues reported a “strong association between myopia and night-time ambient light exposure during sleep in children before they reach two years of age.” That is, kids who slept with night lights were more likely to be nearsighted. This was potentially big news, because we actually don’t fully understand why people become nearsighted, except we know it has to do with reading a lot and spending a lot of time indoors as a kid. They had some idea that light penetrating the eyelids at night might do something, but no real mechanism, just an association over a few hundred kids.

The paper didn’t have some important variables controlled, notably parents’ nearsightedness. Since the condition is also genetic, this was acknowledged as a problem. Still, they wrote:

Although it does not establish a causal link, the statistical strength of the association of night-time light exposure and childhood myopia does suggest that the absence of a daily period of darkness during early childhood is a potential precipitating factor in the development of myopia.

As I stress ad nausem in this post, the “strength” of an association is not an argument for its causal power. And neither is the number of studies in which the association is found. Real spurious findings can produce very strong, easily-reproducible results. And when researchers have a story to fit the rationale can seem strong. Also, the prospect of publishing in a top journal like Nature has to figure in there somewhere. (This problem is endemic in studies of, for example, family structure and child outcomes, among many other subjects.)

In this case there is a very nice explanation, which was reported less than a year later by Zadnik and colleagues (paywalled | sci-hub), who found no association between night lights and myopia – but they did report a very strong relationship between night lights and parents’ myopia. The same pattern was reported in another response to the Quinn paper, in the same issue, by Gwiazda and colleagues. It appears that nearsighted parents like to leave night lights on. Alternately, some other factor causes parental nearsightedness, child nearsightedness, and night light preference, such as education level (e.g., more-educated people read more and use night lights more).

Several other studies have also failed failed to confirm the night-light theory, and now the thing seems to have blown over. It’s not a perfect example, because the bivariate correlation isn’t always found, but I like it as a family-related case. So I think I’ll keep it in.

Black men raping White women: BJS’s Table 42 problem

I’ve been putting off writing this post because I wanted to do more justice both to the history of the Black-men-raping-White-women charge and the survey methods questions. Instead I’m just going to lay this here and hope it helps someone who is more engaged than I am at the moment. I’m sorry this post isn’t higher quality.

Obviously, this post includes extremely racist and misogynist content, which I am showing you to explain why it’s bad.

This is about this very racist meme, which is extremely popular among extreme racists.

The modern racist uses statistics, data, and even math. They use citations. And I think it takes actually engaging with this stuff to stop it (this is untested, though, as I have no real evidence that facts help). That means anti-racists need to learn some demography and survey methods, and practice them in public. I was prompted to finally write on this by a David Duke video streamed on Facebook, in which he used exaggerated versions of these numbers, and the good Samaritans arguing with him did not really know how to respond.

For completely inadequate context: For a very long time, Black men raping White women has been White supremacists’ single favorite thing. This was the most common justification for lynching, and for many of the legal executions of Black men throughout the 20th century. From 1930 to 1994 there were 455 people executed for rape in the U.S., and 89% of them were Black (from the 1996 Statistical Abstract):

For some people, this is all they need to know about how bad the problem of Blacks raping Whites is. For better informed people, it’s the basis for a great lesson in how the actions of the justice system are not good measures of the crimes it’s supposed to address.

Good data gone wrong

Which is one reason the government collects the National Crime Victimization Survey (NCVS), a large sample survey of about 90,000 households with 160,000 people. In it they ask about crimes against the people surveyed, and the answers the survey yields are usually pretty different from what’s in the crime report statistics – and even further from the statistics on things like convictions and incarceration. It’s supposed to be a survey of crime as experienced, not as reported or punished.

It’s an important survey that yields a lot of good information. But in this case the Bureau of Justice Statistics is doing a serious disservice in the way they are reporting the results, and they should do something about it. I hope they will consider it.

Like many surveys, the NCVS is weighted to produce estimates that are supposed to reflect the general population. In a nutshell, that means, for example, that they treat each of the 158,000 people (over age 12) covered in 2014 as about 1,700 people. So if one person said, “I was raped,” they would say, “1700 people in the US say they were raped.” This is how sampling works. In fact, they tweak it much more than that, to make the numbers add up according to population distributions of variables like age, sex, race, and region – and non-response, so that if a certain group (say Black women) has a low response rate, their responses get goosed even more. This is reasonable and good, but it requires care in reporting to the general public.

So, how is the Bureau of Justice Statistics’ (BJS) reporting method contributing to the racist meme above? The racists love to cite Table 42 of this report, which last came out for the 2008 survey. This is the source for David Duke’s rant, and the many, many memes about this. The results of Google image search gives you a sense of how many websites are distributing this:

Here is Table 42, with my explanation below:

What this shows is that, based on their sample, BJS extrapolates an estimate of 117,640 White women who say they were sexually assaulted, or threatened with sexual assault, in 2008 (in the red box). Of those, 16.4% described their assailant as Black (the blue highlight). That works out to 19,293 White women sexually assaulted or threatened by Black men in one year – White supremacists do math. In the 2005 version of the table these numbers were 111,490 and 33.6%, for 37,460 White women sexually assaulted or threatened by Black men, or:

Now, go back to the structure of the survey. If each respondent in the survey counts for about 1,700 people, then the survey in 2008 would have found 69 White women who were sexually assaulted or threatened, 11 of whom said their assailant was Black (117,640/1,700). Actually, though, we know it was less than 11, because the asterisk on the table takes you to the footnote below which says it was based on 10 or fewer sample cases. In comparison, the survey may have found 27 Black women who said they were sexually assaulted or threatened (46,580/1,700), none of whom said their attacker was White, which is why the second blue box shows 0.0. However, it actually looks like the weights are bigger for Black women, because the figure for the percentage assaulted or threatened by Black attackers, 74.8%, has the asterisk that indicates 10 or fewer cases. If there were 27 Black women in this category, then 74.8% of them would be 20. So this whole Black women victim sample might be as little as 13, with bigger weights applied (because, say, Black women had a lower response rate). If in fact Black women are just as likely to be attacked or assaulted by White men as the reverse, 16%, you might only expect 2 of those 13 to be White, and so finding a sample 0 is not very surprising. The actual weighting scheme is clearly much more complicated, and I don’t know the unweighted counts, as they are not reported here (and I didn’t analyze the individual-level data).

I can’t believe we’re talking about this. The most important bottom line is that the BJS should not report extrapolations to the whole population from samples this small. These population numbers should not be on this table. At best these numbers are estimated with very large standard errors. (Using a standard confident interval calculator, that 16% of White women, based on a sample of 69, yields a confidence interval of +/- 9%.) It’s irresponsible, and it’s inadvertently (I assume) feeding White supremacist propaganda. The Census Bureau generally wouldn’t report numbers like this (I think their cutoff is a sample of 50).

Rape and sexual assault are very disturbingly common, although not as common as they were a few decades ago, by conventional measures. But it’s a big country, and I don’t doubt lots of Black men sexual assault or threaten White women, and that White men sexually assault or threaten Black women a lot, too – certainly more than never. If we knew the true numbers, they would be bad. But we don’t.

A couple more issues to consider. Most sexual assault happens within relationships, and Black women have interracial relationships at very low rates. In round numbers (based on marriages), 2% of White women are with Black men, and 5% of Black women are with White men, which – because of population sizes – means there are more than twice as many couples with Black-man/White-woman than the reverse. At very small sample sizes, this matters a lot. But we would expect there to be more Black-White rape than the reverse based on this pattern alone. Consider further that the NCVS is a household sample, which means that if any Black women are sexually assaulted by White men in prison, it wouldn’t be included. Based on a 2011-2012 survey of prison and jail inmates, 3,500 women per year are the victim of staff sexual misconduct, and Black women inmates were about 50% more likely to report this than White women. So I’m guessing the true number of Black women sexually assaulted by White men is somewhat greater than zero, and that’s just in prisons and jails.

The BJS seems to have stopped releasing this form of the report, with Table 42, maybe because of this kind of problem, which would be great. In that case they just need to put out a statement clarifying and correcting the old reports – which they should still do, because they are out there. (The more recent reports are skimpier, and don’t get into this much detail [e.g., 2014] – and their custom table tool doesn’t allow you to specify the perceived race of the offender).

So, next time you’re arguing with David Duke, the simplest response to this is that the numbers he’s talking about are based on very small samples, and the asterisk means he shouldn’t use the number. The racists won’t take your advice, but it’s good for everyone else to know.

Not-too-racist Whites of America: Do you want to be that person?

Everyone is at least a little racist. But hardly any Whites really actively want to hate Blacks and people from other racial-ethnic minority groups. And I bet that includes a lot of people who are thinking of voting for Trump. So this is for people who don’t want to be haters, or even seen as haters.

Think about it this way: Voting is pretty symbolic. Your individual vote is really not going to make the difference. But it says something about who you are, to yourself at least, and to anyone else who knows.

So look at these polling results for African Americans in five key states. Between 2 and 5 out of every 100 Black voters says they support Trump:

If you vote for Trump, because you’re angry about politicians who never get anything done, or you don’t trust Hillary, or you think it’s time for a change in Washington, think about this: do you want to spend the next four or eight years knowing that you voted against virtually every Black person who you will know or meet during that time?

Maybe they’re wrong. But I think, if you’re not the hating kind, it might gnaw at you, and you might feel better if you didn’t take that stand against them. I’m not trying to change your political views in a short blog post, but I do think we’ll get along better – and you will too – if you don’t vote for Trump.

---

Sources:

NYT/Siena: North Carolina, 9/16-9/19

Detroit Free Press: Michigan, 9/10-9/13

NYT/Siena: Florida, 9/10-/14

Monmouth: Georgia, 9/15-9/18

UMW: Virginia, 9/6-9/12

Poverty, marriage, and single mother update

With the annual Census report on poverty out, here are two quick updates.

First, updating this post, the share of all poverty (using official rates) found in single-mother families remains lower than it was from 1974 to 2000. Since 1995, as the poverty rate has gone up and down between 10 and 15 percent, the share of poor people in single-mother families has fallen. As of 2015, 34% of poor people are found in single-mother families.

Marriage has declined, and single motherhood has increased, but that has not produced a poverty population more dominated by single-mother families. Of course these families are more likely to be poor than married-couple families, but they’re not the main poverty story.

Second, updating this post a little, it’s important to keep two major trends in the back of your mind when thinking about social change. The first is that marriage has declined precipitously since 1960. It’s unremitting decline is one of the major social facts of our time. The other trend to keep in mind is that poverty rates fell a lot after the 1960s, but since then have bounced around at an atrocious 10-15%. Now try to keep them both in mind at once: marriage falls, poverty goes up and down. This year’s update puts those together (sorry people who hate this kind of figure), as change in the percentage of women married, and change in the percentage of the population poor.

For a recent op-ed on poverty and marriage, here’s the unpaywalled version of my essay in the Washington Post‘s Post Everything.

Polarization written on the body Google

These are polarizing times in America. And what better to understand that then a highly polarized  measure?

I took the forecast margin of victory for each state for Clinton and Trump, as of today, on Five Thirty Eight. The scores range roughly from -28 to +28, and I reverse them to get the positive score for each candidate (I excluded DC). Then I asked Google Correlate what searches were most correlated with each list of state scores. All the searches here are correlated with the candidate margins at .83 or higher.

Here’s the map as of today:

The Clinton list is dominated by vegetarianism and yoga, Top Chef, and the kind of annoying movies that liberals just love (Before Sunset).

The Trump list is racist anti-Obama stuff, patriotism, and, mostly, the kind of guns you don’t use for hunting. Google gives 100 for each list; I deleted those that weren’t easily categorized. (You can see the full lists here and here.) Here are the highlights:

Really, you people are so predictable.

But what of the Before Sunset-lover working in the Obama Jokes town? The Biggest Gun husband and the Vegetarian Sushi wife with their Ayurvedic Massage therapist next door? Of course, this method will never show the nuances of social life, the moments when people reach out from their silos and grasp, however fleetingly, the hands of those whom the winds of fortune and arbitrary social divisions have attempted to sweep away from them forever. And it won’t show the big, messy middle, the people who do use guns for hunting, eat tofu but aren’t vegetarian, listen to Tom Tom Club and also learn country guitar. I’d be happy to see something about them out there today.

Teen birth rate low but Bible remains a concern

In 2012 I did a post about teen birth rates, abstinence, and Google searches for Antichrist stuff. The most important point was that abstinence education doesn’t work. In this post I use the percentage of teen (women) having a birth, and see what people are Googling in places with more teen births.

This is an inductive approach that generates ideas and surprises. Out of the billions of things people search for, which searches are most correlated with a demographic or social pattern across states? For example, the relationship between low marriage rates and searches about Kanye West is very strong (even controlling for a bunch of demographics), and state suicide rates are highly correlated with lots of searches about guns. If you think these are random flukes, you may be right — but then look at what searches correlate with racial/ethnic composition of states.

So for teen births, this is easy to get from the American Community Survey via IPUMS (I used the 2010-2014 combined file), which asks of each person if they had a baby in the previous year. Teen birth rate is the percentage of women ages 15-19 who did. Then you surf over to the Google Correlate tool and upload the teen birth rates file. The result is the 100 searches that are most highly correlated with the state file you uploaded. Someone with the keys to Google could get more, but this is what any member of the public can do.

We know that teen births are most common in the Southwest and South, and that broad pattern is really what’s most important: Republican-dominated states, the Bible belt, and places with a lot of poor young people.* Here’s the broad strokes:

The Google searches is just for thinking about subtler cultural relationships and generating ideas.

Among the top 100 searches most correlated with teen births, American muscle cars stand out: Mustangs (13), Camaros (5), Hummers, Chargers, along with related things like transmissions. Next, however, is Bible stuff. There are 12 searches that correlate with the teen birth rate at .80 or higher on the list:

nephilim
original bible
book of enoch
bible talk
the book of enoch
i believe in god
yeshua
book of enoch pdf
bible names
the truth shall set you free
yhwh
truth shall set you free

Here’s a map showing the ACS teen births rates on the left and searches for “original bible” on the right, correlation .83:

(A little disturbingly, “what is cinnamon” is also high on the list [correlation .81] — cinnamon is often promoted as a “natural” medicine to cause miscarriage.)

I exported the correlation file from Google and then averaged those 12 searches, producing a bible searches index that correlates with teen births at .87 (all the search correlations come out as z-scores, so the average has mean of 0 and s.d. of .93). Here are the results:**

I’m no Bible expert, and this could all be a total coincidence, but I think some real research on it might be pretty interesting. Maybe the people who say the Bible is awesome for families and teen births are bad should look into it.***

Followup: Of course, if you only look at the highest correlations out of billions, you find high correlations. So I don’t expect a research award for discovering that. And that fact that these bible searches are from certain niches of Christianity is an interesting tidbit but just as food for thought. The more theory-driven version of this research might start with searches for just the word “bible”and test the hypothesis that it’s correlated with teen births.  That relationship is not as strong (correlation .74), but it’s still plenty to go on:

I take from this weaker finding that the stronger pattern above is not just a fluke or an artifact of the method.

• Follow the Google tag to see the many posts using this stuff.
• Follow the teen birth tag for more, including the argument that the teen birth rate is a myth, and the racial implications of promoting delayed births.

 

---

* This survey measure is correlated .89 with the 2008 list of state teen birth rates published by the National Center for Health Statistics. I would have a better sense of which is the right one to use if Google Correlate would say what time period is used for their analysis, but I can’t find that anywhere. When I used the NCHS list instead of my ACS list, it was more dominated by muscle cars and had less Bible stuff, as only “book of enoch” was in the top 100, correlated .87 with teen births.

** Here's the Stata command for making this figure (which I then prettied up a little):
gr twoway (scatter teenbirth biblesearch , mlabel(state) mlabposition(0) msymbol(i)) (lfit teenbirth biblesearch)

*** The 2010-2014 teen birth rates, from the IPUMS release of ACS data are these:

State	State	Teen birth rate (%)
Alabama	AL	2.44
Alaska	AK	2.727
Arizona	AZ	2.385
Arkansas	AR	2.886
California	CA	1.901
Colorado	CO	1.755
Connecticut	CT	0.902
Delaware	DE	1.644
District of Columbia	DC	2.088
Florida	FL	2
Georgia	GA	2.578
Hawaii	HI	1.991
Idaho	ID	2.202
Illinois	IL	2.009
Indiana	IN	2.69
Iowa	IA	1.477
Kansas	KS	2.432
Kentucky	KY	2.936
Louisiana	LA	2.36
Maine	ME	0.852
Maryland	MD	1.783
Massachusetts	MA	0.941
Michigan	MI	1.881
Minnesota	MN	1.428
Mississippi	MS	3.545
Missouri	MO	2.756
Montana	MT	2.065
Nebraska	NE	1.304
Nevada	NV	2.449
New Hampshire	NH	1.135
New Jersey	NJ	1.005
New Mexico	NM	3.5
New York	NY	1.494
North Carolina	NC	2.48
North Dakota	ND	2.328
Ohio	OH	1.901
Oklahoma	OK	3.214
Oregon	OR	1.568
Pennsylvania	PA	1.928
Rhode Island	RI	1.978
South Carolina	SC	2.829
South Dakota	SD	2.271
Tennessee	TN	2.974
Texas	TX	3.303
Utah	UT	1.666
Vermont	VT	1.073
Virginia	VA	1.636
Washington	WA	1.688
West Virginia	WV	2.146
Wisconsin	WI	1.305
Wyoming	WY	1.6

Do we get tenure for this?

For the occasion I titled this photo of Utah “Openness.”

Colleen Flaherty at Inside Higher Ed has written up the American Sociological Association’s committee report, “What Counts? Evaluating Public Communication in Tenure and Promotion.”

I was once a member of the ASA Subcommittee on the Evaluation Of Social Media and Public Communication In Sociology, which was chaired by Leslie McCall when they produced the report. (It is a subcommittee of the task force on engaging sociology, convened by then-President Annette Lareau.)

It’s worth reading the whole article, which also includes comments from Sara Ovink, McCall and me, in addition to the report. Having thought about this issue a little, I was happy to respond to Flaherty’s request for comment. These are the full comments I sent her, from which she quoted in the article:

1. We don’t need credit toward promotion for every thing we do. Scholars who take a public-facing stance in their work often find that it enhances the quality and quantity of their work in the traditional fields of assessment (research, teaching, service), so that separately rewarding the public work is not always necessary. I don’t need credit for having a popular blog – that work has led to new research ideas, better feedback on my research, better grad students, teaching ideas, invitations to contribute to policy, and book contracts.

2. We’d all love to be promoted for authoring a great tweet but no one wants to be fired for a bad one. Assessment of public engagement needs to be holistic and qualitative, taking into the account quality, quantity, and impact of the work. Simplistic quantitative metrics will not be useful.

3. It is also important to value and reward openness in our routine work, such as posting working papers, publishing in open access journals, sharing replication files, and disseminating open teaching materials. Public engagement does not need to mean separate activities and products, but can mean taking a public-facing stance in our existing work.

The SocArxiv project is one outcome of these conversations (links to latest info; submit a paper), especially relating to point #3 above. Academics who open up their work should be recognized for that contribution to the public good and for promoting the future of academia. In that spirit also I proposed a rule change for the ASA Dissertation Award, which now includes this:

To be eligible for the ASA Dissertation Award, candidates’ dissertations must be publicly available in Dissertation Abstracts International or a comparable outlet. Dissertations that are not available in this fashion will not be considered for the award.

It’s hard to change everything, but it’s not that hard to make some important changes in the right direction. Rewarding engagement and openness is an important step in the right direction.