Category: surveys

Deconstruction of a Twitter fracas involving Ibram Kendi

October 31, 2021 • 9:15 am
The tweet by Ibram X. Kendi quoted by Hashmi and Greenwald below was later removed by Kendi.  You can see the article on The Hill to which Kendi refers here, and the original survey is here.

 

Now clearly Kendi removed it (though he denies it; see below) because it appears to show that lying about your race if you’re white improves your chances of being accepted in college. I can see no other reason for the removal, especially given the pushback he got from people like Greenwald.

However, the original survey of 1250 college students doesn’t have a control group: the percentage of students who didn’t fake minority status and still got admitted to college. You could sort of have a control by looking at the percentage of students who got into college in each of the four categories: Black, Native American, Latino, and Asian/Pacific islander. If there were huge disparities in the acceptance figures among these liars—say that those who pretended to be Black or Hispanic got in much more easily than those who pretended to be Asian/Pacific islander. Even that has problems, but would show that claiming you’re of oppressed minority status makes a difference in admission, facilitating it in the expected direction (Blacks and Hispanics are favored over Asians). But I can’t find that data.

Absent the control percentage, of students who didn’t lie and got in, you can’t make an airtight case for the advantage of lying.  I suspect, however, that fewer than 3/4 of applicants get into college. (But even that’s problematic, because the survey didn’t specify that the students got into the college of their choice, just got into a college.) All you can say is that there is a reason students lie about their ethnicity (the most common reason given was to get financial aid), and that nearly every college in the U.S. is looking for good minority students, exercising affirmative action to take them. (I am, by the way, in favor of a form of such affirmative action.)

The other thing to ponder when “deconstructing” these tweets is why Greenwald says there are “numerous obvious falsehoods” in what Kendi said.  I couldn’t see any immediately, but the second tweet below clarifies things a bit:

What Griswold means is that if one-sixth of white applicants lie about being Native Americans, and 77% of those get in, then one expects (if these results are general) about 12.8% of the truly white students in a college would be classified as Native Americans. (The true figure, of course, will be lower than this because not all students in a college are white). Still, I don’t know of any college, except perhaps ones in the Southwest, where even 5% of students are classified as Native Americans.

What this comes down to is that the data in the surveys cited by Kendi are surely bogus. That doesn’t mean that Kendi screwed up big time, because this line of reasoning takes time, and he may simply have tweeted out what he read as the headlines or bullet points in the survey or The Hill article.

Where he messed up was simply posting the tweet, probably because the data seem to go against his thesis that there is structural racism everywhere, which would predict that members of minorities don’t get preferential admission to college or financial aid. If they didn’t (and of course they do), there would be no motivation for white students to lie. In other words, the data (though they may be faulty) appear at first glance to falsify Kendi’s main thesis: there is inequity everywhere, and if you see it it reflects “structural racism” acting at the present. Everyone involved in colleges knows that this is not true for the admissions process, at least for black and Hispanic students.

In the end, though, Kendi probably did the only thing he could do: delete the tweet, for the man is loath to admit he’s wrong.  But he screwed up again when he started defending his original tweet, saying stuff like this:

Again, what we need here is a control group: a group of similar white students who didn’t lie about their ethnicity, and whether their admission rates were substantially lower than 81% (the admission figure quoted in the survey for students who lied). If there is such a difference, then Kendi is wrong.  But I suspect that lying does help one get into college or get financial aid, and students realize it (remember, over a third of  the sample lied about their ethnicity).  And if that is true, then the “tortured line of thinking” is not tortured at all.  If there is an advantage of lying, then it’s not just that you “think” you have an advantage. (That’s why Kendi deleted his original tweet.)

This is not to deny that there is structural racism in various institutions or organizations. But if a more sophisticated analysis and explanation for the data show preferential college admission of minorities, then there is no structural racism in the college admissions process.  Indeed, there, at least, it’s an advantage to be a minority.  And we know this because colleges practice affirmative action.

I suspect that Kendi’s answers reflect his being flummoxed by all this. If I were Kendi, I would have simply removed the tweet. He’d still be excoriated by people who captured the screenshot, but he’s going to get into more trouble if he tries to debate. I’ve given him some material to defend himself in this post, but there’s simply no doubt that there is no “inequity” in college admissions for blacks or Hispanics.

The tweets may reflect reasons why Kendi doesn’t engage in live debates.

Pew survey of American Democrats and their views on Biden as an “old white male”: some good news and some puzzling news

April 22, 2020 • 8:45 am

This new article by Pew Research reports a survey of American Democrats, asking them if they were bothered by the fact that the Democratic Presidential nominee (Biden of course) is “a white man in his 70s.” The results are pretty much as you expect—most people don’t care that much (especially given his opponent!), but there’s one surprising result. Click on the screenshot to read the short report:

First, as the figure below shows (it also displays the question posed to people), the people least concerned with Biden being an “old white male” are, of course, those who originally supported Biden (79% aren’t bothered). But those who backed other candidates show more concern, except those who backed Sanders—the other old white male (79% not bothered). Those who originally favored either Warren (an old white woman) or Buttigieg (a young white gay male) were, as expected, more bothered  (26% and 43% not bothered, respectively.  Overall, 59% of total Dems aren’t that bothered about Biden’s age and race.

But the good news is that regardless of whether Dems are bothered by the OWM (old white male) syndrome, they are overwhelmingly in favor of Biden when he’s put up against Trump. 89% of all Dems disapprove of the way Trump is handling his job (8% approve!), with most of the disapproval being “strong”. The disapproval is a bit stronger among those who are worried about Biden being an OWM. But there’s no difference when it comes to voting: among all Dems surveyed, 85% intend to vote for Biden or lean towards him (4% are Trump/lean Trump!), with 10% being “neither or others”. When you divide up the Dems by whether they’re concerned about Biden’s being an OWM, there’s hardly any difference between the degree to which the two groups favor Biden:

This is heartening in that it shows that Democrats have largely come together behind one candidate, even though, as I believe—many interviewees must feel the same way—that Biden is not an overwhelmingly fantastic candidate. Still, only 85% of Dems say at this point they’ll probably vote for Biden. If too many Dems stay home because they wanted Sanders or Warren to be the nominee, we’ll be screwed.

But there is one surprising result. If you break down the Democrats by race, age, gender, ideology, and education, and ask them if they’re bothered by Biden being an OWM, here’s what you get (the red box is one I’ve added):

Men and women are about the same, and, as you might expect, the younger and more educated people are more bothered by the OWM syndrome (they are more likely, I think, to be more conscious of racial equity given the climate among liberals). The difference between postgraduates on one hand and those who had some high school education but no college on the other, is huge: 58% vs 24% respectively are bothered by the fact that Biden’s an OWM.

What surprised me was the division among races. I would have expected that minorities—African-Americans and Hispanics—would have been more bothered by OWM, for no candidate represents them, though the withdrawn candidates Cory Booker and Kamala Harris identify as black. But while 49% of white Democrats are bothered by Biden being an OWM, only 28% of blacks and 30% of Hispanics are.

It’s clear, then, that white Democrats are more bothered by OWM candidates than are blacks or Hispanics.  That’s not what I expected.

Now I can make up post facto theories (which are mine) about why this is so. For instance, you could say that blacks and Hispanics, who are overwhelmingly Democratic, are so relieved that Biden got the nomination—he polls more strongly in those communities than Buttigieg, Warren, or Sanders—that they don’t much care if he’s an OWM. Or one could say that this reflects white guilt. I’ve heard that the most woke antiracists are in fact white, but until now I had no data supporting that.

Or perhaps there’s another explanation. If you have a theory which is yours, put it in the comments below.

h/t: cesar

New Pew survey shows that Americans are rapidly becoming less Christian and more secular

October 18, 2019 • 11:10 am

The Pew Organization conducted surveys for their Religious Landscape Studies in 2007 and 2014, assessing the religiosity, non-religiosity, beliefs, and churchgoing habits of Americans. They continued these surveys last year and this year, though on a more restricted scale. Nevertheless, the new Pew Survey, whose summary you can see by clicking on the screenshot below (full pdf here), heartens me, substantiating my theory (which is not mine alone) that America is becoming less religious all the time. What surprises me, as you can see in the headline below, is that the decline in the last 12 years is so fast. 

I’ll give some salient data and graphs below. First, the take-home message from the report (my emphasis):

The religious landscape of the United States continues to change at a rapid clip. In Pew Research Center telephone surveys conducted in 2018 and 2019, 65% of American adults describe themselves as Christians when asked about their religion, down 12 percentage points over the past decade. Meanwhile, the religiously unaffiliated share of the population, consisting of people who describe their religious identity as atheist, agnostic or “nothing in particular,” now stands at 26%, up from 17% in 2009.

Both Protestantism and Catholicism are experiencing losses of population share. Currently, 43% of U.S. adults identify with Protestantism, down from 51% in 2009. And one-in-five adults (20%) are Catholic, down from 23% in 2009. Meanwhile, all subsets of the religiously unaffiliated population – a group also known as religious “nones” – have seen their numbers swell. Self-described atheists now account for 4% of U.S. adults, up modestly but significantly from 2% in 2009; agnostics make up 5% of U.S. adults, up from 3% a decade ago; and 17% of Americans now describe their religion as “nothing in particular,” up from 12% in 2009. Members of non-Christian religions also have grown modestly as a share of the adult population.

I’m lumping together the atheists, agnostics, and “nothings in particulars” as “nonbelievers,” and that group has risen 9% in 9 years. While self-described atheists are at only 4%, it’s still a statistically significant rise from 2% ten years ago, and of course we know how reluctant Americans are in telephone polls like these to say they are atheists. I suspect the real number of people who don’t accept the existence of a god is much higher. Here are two plots from the survey:

I suspect, in the graph above, that “nothing in particular” pretty much means “nonreligious”, and that’s gone up 5% in just 10 years.

As in Europe, church attendance in the U.S. is dropping pretty rapidly:

The graph below shows that the secularization results not so much from people changing their minds as they age, but that each cohort becomes successively less religious than the last. Nonbelief in America spreads over the bodies of dead believers.

There are other results as well, none of them giving hope to people who think religion in America is here to stay. The “nones” have increased from 39 million in 2009 to 68 million in 2018/2019, while Christians have dropped from 178 million to 167 million over the same period. Non-Christians (Jews, Muslims, Buddhists, Hindus, etc.) have increased by only 2% over the same period—from 5% to 7%.

Women remain more religious than men, but both sexes are growing less religious. Republicans are still more Christian than are Democrats (79% versus 55% respectively), but both have dropped significantly in religiosity (Republicans were 86% Christian in 2009, Democrats 72%). 34% of Democrats are unaffiliated, while only 16% of Republicans are. As expected, the GOP could be termed “God’s Own Party.”

Finally, Black and Hispanic Democrats remain significantly more religious than White Democrats, while—the only “bad news” in the survey—there’s been no decline in the proportion of Protestants who describe themselves as “born again or evangelical” (about 60%, higher than I would have thought).

Overall, then the news is good—but only if you’re a nonbeliever.

 

h/t: cesar

What’s your meaning and purpose?

March 13, 2018 • 9:15 am

Here’s survey I’m taking to see whether a theory I have, which is mine, bears any resemblance to reality. Here are two questions I’d like readers to answer in the comments. Here we go:

If a friend asked you these questions, how would you answer them?

1.) What do you consider the purpose of your life?

2.) What do you see as the meaning of your life?

Now I know there are a lot of nonbelieving readers, so I don’t expect that many of the answers will involve “God.” I am not implying that either meaning or purpose must be conferred by some kind of deity—or even by forces of beings outside yourself. Further, you may consider the questions ambiguous or meaningless, in which case say so.

I got curious about this since yesterday Andrew Sullivan asserted that the last few centuries of human progress, showing big improvements in worldwide well being and material welfare, rob life of meaning, purpose, and spiritual sustenance. To claim that is to claim that people’s lives actually have those attributes. (You can also expatiate about what brings you “spiritual sustenance.”)

I’m trying to find out whether, in this audience, people really feel that there’s meaning and purpose in their lives, and, if they have some “spiritual sustenance,” where it comes from.

Sullivan also implied that atheists have no source of these attributes, so asking an audience comprising mainly the godless might be instructive.  Please humor me and answer the questions.

Thanks!