Russ Lyons sends along this news article by Ian Johnston, who writes:

The prestigious medical journal The Lancet has been attacked for publishing an academic paper that claimed eating too little salt could increase the chance of dying from a heart attack or stroke.

Johnston summarizes the study:

Researchers from the Population Health Research Institute in Canada, studied more than 130,000 people from 49 different countries on six continents and concluded people should consume salt “in moderation”, rather than trying to reduce it in accordance with government guidelines across the world. . . .

The paper compared the health of people who tests showed had consumed low levels of sodium (up to three grams a day), medium amounts (four or five grams) and high levels (seven grams or more).

“Those participants with four to five grams of sodium excretion had the lowest risk” of death or suffering a “major cardiovascular disease event”, the researchers reported.

Among people who had high blood pressure, eating high and low levels of salt “were both associated with increased risk”. And for people without high blood pressure, consuming less than three grams a day was “associated with a significantly increased risk” – 11 per cent – of death or a serious cardiovascular event.

But there are critics:

Professor Francesco Cappuccio, head of the World Health Organization’s Collaborating Centre for Nutrition, attacked both the methods used in the study and the journal for agreeing to publish it.

“It is with disbelief that we should read such bad science published in The Lancet,” he said.

Professor Cappuccio said the article contained “a re-publication of data” used in another paper.

“The flaws that were extensively noted in their previous accounts are maintained and criticisms ignored,” he said.

The measurement of salt intake used by the researchers, he said, was “flawed” because it was done by testing urine samples given in the morning and then “extrapolated to 24-hour excretion” using an “inadequate” equation.

Professor Cappuccio also said the participants were “almost exclusively from clinical trials of sick people that have a very high risk of dying and are taking several medications”.

Now I don’t know what to think. I really don’t. I haven’t looked at the paper or the criticisms. If the study really is so flawed, though, I can’t say I’m surprised that it was published in the Lancet, a journal that’s famous for producing headline-grabbing papers that are later refuted, such as that Iraq survey, or that article claiming that gun laws could reduce firearm deaths by 200% or whatever, or, most notoriously, that paper by Andrew Wakefield [no link needed]. The Lancet may well publish some high-quality work but they do seem to have a weakness for bold claims and publicity.

I suppose the Lancet will publish a letter by Cappuccio or someone else with the criticisms? Perhaps a reader can keep us up to date here.

P.S. I’m sure I eat too much salt for my own good. I have a big jar of pretzels just sitting here in my office!

The post Should you abandon that low-salt diet? (uh oh, it’s the Lancet!) 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 Should you abandon that low-salt diet? (uh oh, it’s the Lancet!) appeared first on All About Statistics.

How far away is the nearest hospital? How far is the nearest restaurant? The nearest gas station? These are commonly asked questions whose answers depend on the location of the person asking the question.

Recently I showed an algorithm that enables you to find the distance between a set of locations and a fixed set of reference sites. The set of locations (the people) can be arbitrary, so a good way to summarize the problem is to draw a contour plot that shows the distance from every point in a region to the nearest reference site (a hospital or store). In SAS, you can create this graph by using the SPP procedure, which analyzes spatial point processes. The graph is called an empty-space plot.

Example: The location of large US cities

Some people love big cities and the diverse opportunities that they provide. They want to know how far it is to the nearest big city. Other people hate the noise and congestion; they want to be as far away from a big city as possible. To demonstrate how to create an empty-space plot, let's look at the spatial distribution of US cities that have population more than 200,000.

data BigCities;
set maps.uscity;                          /* part of SAS/GRAPH */
where statecode not in ("AK" "HI" "PR") & pop > 200000;
if pop>500000 then Name=City;             /* label really big cities */
else Name = "        ";
run;
 
title  "Large US Cities";
title2 "Population > 200,000";
proc sgplot data=BigCities;
  scatter x=x y=y / markerattrs=(symbol=CircleFilled) datalabel=Name;
run;

The scatter plot supports a well-known fact: cities are not settled in random locations. People tend to settle near beneficial geographic features, such as proximity to rivers, harbors, and other strategic resources. Therefore the cities are not spread uniformly throughout the country. There are clusters of cities on both coasts and there is a large empty space in the rugged northwest. The large empty space contain locations that are much farther away from a big city than would be expected for a random uniform assortment of 86 cities.

The following call to PROC SPP create an empty-space plot, which reveals regions that are far from a big city. The F option will be explained in a moment:

proc spp data=BigCities plots(equate)=(emptyspace(obs=on) F);
   process BigCities = (x, y / area=(-0.37,-0.19,0.32, 0.24)) / F;
run;

The dark blue color indicates geographic coordinates that are far from any large city. The darkest blue appears to be in the northwestern states, such as Montana, Idaho, and Wyoming. Western Texas also stands out. Those are great places to live for people who want to be far from large cities. There is also dark blue in regions that are uninhabitable (oceans and the Great Lake) or are not part of the US (Mexico and Canada).

I think the image is beautiful, but it is useful as well. It gives a visual summary of locations that are far from the reference sites (big cities).

Empty-space distances

If you want a more statistical analysis, the F option in the PROCESS statement and in the PLOTS= option tells PROC SPP to lay down a regular grid of points and compute the distance from each grid position to the nearest big city. (By default the grid is 50 x 50.) The graph shows the cumulative distribution of these distances in blue. You can compare the distribution of these distances to the expected distribution for randomly selected locations, which is shown in red and labeled "complete spatial randomness."

You can see that the empirical distribution rises more slowly than the theoretical curve. This indicates that the observed locations are more clustered than you would expect if the cities were randomly located. For details about the F option, see the documentation for PROC SPP.

The main purpose of this example is to demonstrate that PROC SPP makes it easy to compute the empty-space plot, which summarizes the distance between an arbitrary location and a pattern of points. In this article the points were large US cities, but obviously you could analyze other point patterns. The SPP procedure has many options that indicate whether the pattern appear to be randomly distributed or clustered. The procedure indicates that the locations of large US cities are not random.

The post The empty-space distance plot appeared first on The DO Loop.

Please comment on the article here: The DO Loop

The post The empty-space distance plot appeared first on All About Statistics.

The University of Melbourne is advertising for a “Professor in Statistics (Data Science)”. Melbourne (the city) is fast becoming a vibrant centre for data science and applied statistics, with more than 4700 people signed up for the Data Science Meetup Group, a thriving start-up scene, the group at Monash Business School (including Di Cook and me), and the Monash Centre for Data Science (including Geoff Webb and Wray Buntine). Not to mention that Melbourne is a wonderful place to live, having won the “World’s most liveable city” award from the Economist for the last 6 years in a row.

Actually, the Uni of Melbourne currently has two professorships on offer — the other being the Peter Hall Chair in Mathematical Statistics. (Not sure that anyone would actually feel qualified to have a job with that title!)

So any professors of statistics out there looking for a new challenge, please consider coming to Melbourne. We’ll even invite you to visit us from time to time at Monash.

Please comment on the article here: Hyndsight

The post Come to Melbourne, even if not to Monash appeared first on All About Statistics.

Swupnil made this graph for a research meeting we had today:

It looks so cool. I think it’s the gray colors.

So here’s my advice to you: If you want to make your graphs look cool, use lots of gray.

The post Gray graphs look pretty 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 Gray graphs look pretty appeared first on All About Statistics.

Please comment on the article here: Big Data, Plainly Spoken (aka Numbers Rule Your World)

The post Why are GMAT scores going up? appeared first on All About Statistics.

• Introduction to the tidyverse tools - Jiddu Alexander
  
• Performance profiling and improvement in R - Nils Glück
  
• Batch processing of R-Scripts with Excel - Klaus Jacobi
  

Please comment on the article here: mages' blog

The post Next Kölner R User Meeting: Friday 14 October appeared first on All About Statistics.

Justin Esarey writes:

This Friday, October 14th at noon Eastern time, the International Methods Colloquium will inaugurate its Fall 2016 series of talks with a presentation by Andrew Gelman of Columbia University. Professor Gelman’s presentation is titled “The Statistical Crisis in Science.” The presentation will draw on these two papers:

“Beyond Power Calculations: Assessing Type S (Sign) and Type M (Magnitude) Errors”
http://www.stat.columbia.edu/~gelman/research/published/retropower_final.pdf

“Disagreements about the Strength of Evidence”
http://www.stat.columbia.edu/~gelman/research/published/ChanceEthics12.pdf

To tune in to the presentation and participate in the discussion after the talk, visit http://www.methods-colloquium.com/ and click “Watch Now!” on the day of the talk. To register for the talk in advance, click here:

https://riceuniversity.zoom.us/webinar/register/6272124085322b4fdc2040ba88984b7b

The IMC uses Zoom, which is free to use for listeners and works on PCs, Macs, and iOS and Android tablets and phones. You can be a part of the talk from anywhere around the world with access to the Internet. The presentation and Q&A will last for a total of one hour.

A webinar isn’t as much fun as a live talk, but you can feel free to check this one out. According to Justin, it’s open to all and there will be lots of time for questions.

The post My online talk this Friday noon for the Political Methods Colloquium: The Statistical Crisis in Science 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 My online talk this Friday noon for the Political Methods Colloquium: The Statistical Crisis in Science appeared first on All About Statistics.

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

The post Masters of some appeared first on All About Statistics.

In a news article entitled, “Inequality might start before we’re even born,” Carolyn Johnson reports:

Another study, forthcoming in the Journal of Human Resources, analyzed birth outcomes in counties where the home team goes to the Super Bowl. . . . The researchers found that women in their first trimester whose home team played in the Super Bowl had measurably different birth outcomes than pregnant women whose teams did not go to the championship. There was a small, 4 percent increase in the probability of having a baby with low birth weight when the team won.

Garden. Of. Forking. Paths.

And this:

The magnitude of the change was tiny, but what was striking to Mansour [one of the authors of the study] was that it was detectable at all, in studying Super Bowl history from 1969 to 2004.

On the contrary, I’m not surprised at all. Given that researchers have detected ESP, and the effects of himmicanes, and power pose, and beauty and sex ratio, etc etc etc., I’m not surprised they can detect the effect of the Super Bowl. That’s the point of researcher degrees of freedom: you can detect anything.

As a special bonus, we get the difference between significant and non-significant:

The chances of having a low birth weight baby were a bit higher when the team won in an upset, suggesting that surprise may have helped fuel the effect. There was little effect when the team lost.

Really no end to the paths in this garden.

To her credit, Johnson does express some skepticism:

There’s a huge caveat to interpreting these studies. . . . That means researchers have to use natural experiments and existing data sets to explore their hypothesis. That leads to imaginative studies — like the Super Bowl one — but also means that they can’t be certain that it’s the prenatal experiences and not some other factor that explains the result.

But not nearly enough skepticism, as far as I’m concerned. To say “they can’t be certain that . . .” is to way overstate the evidence. If someone shows a blurry photo that purports to show the Loch Ness Monster, the appropriate response is not “they can’t be certain that it’s Nessie and not some other factor that explains the result.”

Sure, you can come up with a story in which the Super Bowl adds stress that increases the risk of low birth weight. Or a story in which the Super Bowl adds positive feelings that decrease that risk. Or a story about the relevance of any other sporting event, or any other publicized event, maybe a major TV special or an election or the report of shark attacks or prisoners on the loose or whatever else is happening to you this trimester. Talk is cheap, and so is “p less than .05.”

P.S. One more thing. I just noticed that the news headline was “Inequality might start before we’re even born.” What do they mean, “might”? Of course inequality starts before we’re even born. You don’t have to be George W. Bush or Edward M. Kennedy to know that! It’s fine to be concerned about inequality; no need to try to use shaky science to back up a claim that’s evident from lots of direct observations and a huge literature on social mobility.

The post No, I don’t think the Super Bowl is lowering birth weights 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 No, I don’t think the Super Bowl is lowering birth weights appeared first on All About Statistics.

Please comment on the article here: No Hesitations

The post Machine Learning vs. Econometrics, II appeared first on All About Statistics.