Please comment on the article here: Gianluca Baio's blog

The post Slightly off (by 26911) appeared first on All About Statistics.

Hey, did you hear about that study last year, where some researchers claimed to find that a 20-minute doorstep conversation with skeptical voters could change views on same-sex marriage? It was published in the tabloids and featured on This American Life? And it turned out it was all a fraud, that one of the authors of the paper made up the data and the other author had to retract it from the journal? Remember that?

OK, good. Well, here’s some news. David Broockman and Joshua Kalla, two of the people who figured out about that earlier study being faked, then up and ran their own study, and look what they found:

A single approximately 10-minute conversation encouraging actively taking the perspective of others can markedly reduce prejudice for at least 3 months. We illustrate this potential with a door-to-door canvassing intervention in South Florida targeting antitransgender prejudice. . . . 56 canvassers went door to door encouraging active perspective-taking with 501 voters at voters’ doorsteps. A randomized trial found that these conversations substantially reduced transphobia. . . . These effects persisted for 3 months, and both transgender and nontransgender canvassers were effective. The intervention also increased support for a nondiscrimination law, even after exposing voters to counterarguments.

Wow. As I wrote at the sister blog, this has to cause us to rethink the idea that persuasion is so difficult—at least for fluid issues such as gay rights where public attitudes are changing fast.

Betsy Levy Paluck wrote an accompanying article summarizing the research on persuasion:

What do social scientists know about reducing prejudice in the world? In short, very little. Of the hundreds of studies on prejudice reduction conducted in recent decades, only ~11% test the causal effect of interventions conducted in the real world. Far fewer address prejudice among adults or measure the long-term effects of those interventions.

Also, I’d expect persuasion to be less effective in the real world than in Broockman and Kalla’s experiment, because in a live political setting you’ll see efforts at persuasion from both sides.

Finally, let me remind you that when the suggestion came up to replicate the LaCour and Green study, I was dismissive:

Ulp. There are lots and lots of studies people are interested in doing, and I’m sure this activist group in Los Angeles has a long to-do list. Do you really think they should spend their precious time, money, and human resources to study an idea that is contradicted by an entire 900-paper literature and whose only claim to plausibility was a made-up experiment?? . . . You gotta be kidding.

I was dismissive, and I was wrong. I’m glad nobody was listening to me on this one!

P.S. One weird thing about publishing on the Monkey Cage is . . . newspaper comment sections are the worst. We got some scary-ass bigots out there. I mean, sure, no surprise, lots of people have hateful attitudes and they have every right to express their views. Still it upsets me a bit to see the actual words.

The post Gay persuasion update appeared first on Statistical Modeling, Causal Inference, and Social Science.

Please comment on the article here: Statistical Modeling, Causal Inference, and Social Science

The post Gay persuasion update appeared first on All About Statistics.

I got two of these in the same day!

1. A colleague emails me that a colleague emailed him regarding a study on women in the workplace. The headline conclusion is: “Corporate America is not on a path to gender equality.” My colleague’s colleague writes:

This coincides with my prior beliefs, but for exactly that reason I thought it important to dig into the evidence and see what the latest numbers show. . . .

He then went into all sorts of problems he had with the report, and then summarized:

I am just curious why the report chooses to focus on “we aren’t there yet, and therefore are not on the path” when in fact it looks from the data like we are *exactly* on the path.

I *do* have a lot of sympathy for the report. I think the recommendations at the end are sensible, and that many of the statements about forms of cultural bias and inequality in the home, etc., are correct. But these are based on other data. This study claims to present novel conclusions that are not supported by the data presented.

My colleague sent me the above with the note that it might be of interest.

I replied: The last thing I want to do, as a man, is to say that progress in gender equality is just fine—so I think I’ll stay out of this one!

2. Someone else points me to a news article entitled, “Height May Be Linked to Increased Cancer Risk, Study Contends,” and asks, “What do we do with something so absurd?”

I replied: Too difficult for me to understand . . . I don’t want to touch this one!

P.S. The real selection bias, though, comes when I don’t even tell you I’m not writing about a topic!

The post Selection bias, or, some things are better off left unsaid appeared first on Statistical Modeling, Causal Inference, and Social Science.

Please comment on the article here: Statistical Modeling, Causal Inference, and Social Science

The post Selection bias, or, some things are better off left unsaid appeared first on All About Statistics.

Please comment on the article here: Gianluca Baio's blog

The post Advanced Bayesian Modelling with BUGS course appeared first on All About Statistics.

Please comment on the article here: Econometrics Beat: Dave Giles' Blog

The post The Econometric Game – Update appeared first on All About Statistics.

The history of data journalism goes back much farther than most people assume. Long before computers or punch cards, and before even the first newspapers the way we know them today, data was being published. ProPublica’s Scott Klein has been digging up a lot of interesting history.

He wrote an article for this year’s Malofiej book, which he also posted on the ProPublica blog. You should also subscribe to his email newsletter Above Chart, where he goes into more fascinating examples. And finally, watch his wonderful presentation at Tapestry 2016 below.

Please comment on the article here: eagereyes

The post Links: Scott Klein on the History of Data Journalism appeared first on All About Statistics.

In this episode, Hilary and I talk about the difficulties of separating data analysis from its context, and Feather, a new file format for storing tabular data. Also, we respond to some listener questions and Hilary announces her new job.

If you have questions you’d like us to answer, you can send them to nssdeviations @ gmail.com or tweet us at @NSSDeviations.

Subscribe to the podcast on iTunes.

Please leave us a review on iTunes!

Show notes:

• NSSD Patreon page

• Feather git repository

• Apache Arrow

• FlatBuffers

• Roger’s blog post on feather

• NausicaaDistribution

• New York R Conference

• Every Frame a Painting

Download the audio for this episode.

Please comment on the article here: Simply Statistics

The post Not So Standard Deviations Episode 13 – It’s Good that Someone is Thinking About Us appeared first on All About Statistics.

David Allison writes:

UAB’s Office of Energetics (http://www.soph.uab.edu/energetics/home) seeks a post-doctoral fellow with statistical training for a fellowship in obesity-related research as part of a highly interactive collaborative interdisciplinary team.

· Collaborative Style: We are seeking a good-spirited team-player who gets along well with others from diverse intellectual, social, and demographic backgrounds and who enjoys being part of a group and helping others.
· Communication Ability: Excellent English-language written and oral communication abilities are essential. Must be able to communicate with scientists from many disciplines and who have statistical training ranging from zero to highly expert.
· Citizenship: Though not formally required, US citizenship or permanent residency is desirable.
· Statistical Abilities: Abilities should be commensurate with a PhD in biostatistics or statistics and include strength and experience in applied data analysis. No knowledge of or experience in obesity is required. We will provide that training. The ability to work rigorously and carefully to assure quality control is essential.
· Duties: Participate in multiple scientific projects related to obesity, energetics, and related topics. Conduct statistical analyses, write sections of manuscripts, write sections of grant proposals, and participate in design of new studies. Many opportunities to publish papers as co-author and some as lead author. Participate in training seminars and workshops.
· Salary: Greater than or equal to NIH stipend levels (see: http://grants.nih.gov/grants/guide/notice-files/NOT-OD-16-047.html)
· Benefits: See – https://www.uab.edu/postdocs/benefits
· Start Date: As early as possible.
· Duration of Appointment: Initial appointment is for 1 year and is potentially renewable for up to an additional 3 years by mutual agreement.
· Environment: See: http://www.norc.uab.edu/news/161
· Other: Women, persons from traditionally underrepresented groups, and persons with disabilities are encouraged to apply.

Looks interesting, and this is a research group that’s trying to do real science, not just spit out p-values.

The post Postdoc in Alabama on obesity-related research using statistics appeared first on Statistical Modeling, Causal Inference, and Social Science.

Please comment on the article here: Statistical Modeling, Causal Inference, and Social Science

The post Postdoc in Alabama on obesity-related research using statistics appeared first on All About Statistics.

Kaveh sends along this, from a recent talk at Berkeley by Katherine Casey:

It’s so gratifying to see this sort of thing in common use, only 15 years after Francis and I introduced the idea (and see also this more recent paper with Carlin).

The post These celebrity photos are incredible: Type S errors in use! appeared first on Statistical Modeling, Causal Inference, and Social Science.

Please comment on the article here: Statistical Modeling, Causal Inference, and Social Science

The post These celebrity photos are incredible: Type S errors in use! appeared first on All About Statistics.

Last week I showed how to generate random points uniformly inside a 2-d circular region. That article showed that the distance of a point to the circle's center cannot be distributed uniformly. Instead, you should use the square root of a uniform variate to generate 2-D distances to the origin. This result generalizes to higher dimensions. If you want to simulate points uniformly in the d-dimensional ball, generate the radius to be proportional to the dth-root of a uniform random variate.

Although it is possible to generalize the polar mapping and obtain a parameterization of the d-dimensional ball, there is an easier approach to simulate points in a ball. The basic idea is to simulate d independent standardized normal variables, project them radially onto the unit sphere, and then adjust their distance to the origin appropriately. You can find this algorithm in many textbooks (Devroye, 1986; Fishman, 1996), but Harman and Lacko (2010) summarize the process nicely. To paraphrase:

If Y is drawn from the uncorrelated multivariate normal distribution, then S = Y / ||Y|| has the uniform distribution on the unit d-sphere. Multiplying S by U^1/d, where U has the uniform distribution on the unit interval (0,1), creates the uniform distribution in the unit d-dimensional ball.

It is easy to carry out this operation in SAS. In Base SAS, you can write a DATA step that uses arrays. The following program provides a SAS/IML solution. I used SAS/IML Studio to create a 3-D rotating scatter plot of the simulated cloud of points. You can download the SAS program that contains the DATA step and the IMLPlus program that creates the animation.

proc iml;
call randseed(12345);
d = 3;                               /* dimension = number of variables */
N = 1000;                            /* sample size = number of obs     */
radius = 2;                          /* radius of circle */
Y = randfun(N // d, "Normal");       /* Y ~ MVN(0, I(d)) */
u = randfun(N, "Uniform");           /* U ~ U(0,1)       */
r = radius * u##(1/d);               /* r proportional to d_th root of U */
X = r # Y / sqrt(Y[,##]);            /* Y[,##] is sum of squares for each row */
/* X contains N random uniformly distributed points in the d-ball */

The post Generate points uniformly inside a d-dimensional ball appeared first on The DO Loop.

Please comment on the article here: The DO Loop

The post Generate points uniformly inside a d-dimensional ball appeared first on All About Statistics.