Colder States report sleeping more—65 degrees F evidently the ideal
Check out the Center for Disease Control (CDC) figures on sleep in the USA map in this Stat report. Note how dark it gets up in our neck of the woods of Minnesota, i.e., us being able to sleep better.
Dark is good but even better is the cold according to this new YouTube video posted by the Wall Street Journal.
There being a sweet spot on temperature makes perfect sense to me, this being based on many sleepless nights camping in the cold or hot. However the worst night I can remember was an overnight ice-fishing outing with a bunch of boy scouts. They could not stop fiddling with the space heater, which cycled us from freezing to boiling for some hours before finally stabilizing at a reasonable temperature. That is when the farting began and the giggling commenced. The whole troop deserved a merit badge for flatulence. But I digress…
A Data Sherlock’s best friend: IBM’s Watson
According to this report last week by eWeek, more than 1 million users have registered for IBM’s Watson Analytics service since it launched a little over 1 year ago. Evidently this artificially intelligent (AI) statistician-in-a-box will enable “citizen data scientists” to decipher patterns in the massive pile of information that now flow in from all quarters. Current clients featuring by eWeek range from multinational law firm using it to identify new areas of practice to a UK a care provider looking for factors that improve worker safety. IBM itself now operates an enterprise called Watson Health that deciphers medical imagery, and they bought the digital assets of the Weather Company to help businesses defend themselves against Mother Nature.*
Unfortunately for one of the early adopters of Watson—the MD Anderson Cancer Center at University of Texas (UT)—AI’s current IQ still falls far short of initial hopes.
“On Jeopardy! [Where Watson made its name 5 years ago by defeating the human champions] there’s a right answer to the question [actually the right question for the answer], but, in the medical world, there are often just informed decisions.”
— Lynda Chin, chief innovation officer for health affairs, UT
So it seems that, for the moment, at least, human statistical Sherlocks will not be replaced by AI’s overseen by amateurs at sleuthing out the culprits for cancer or other highly prized information. However, Watson might be as capable an assistant as ‘his’ literary namesake.
*1/6/16 Financial Times “Big Read” on “Artificial Intelligence”, p 5 sidebar.
Too many dogs at farmers markets?
Today’s “Gray Matter” column in the New York Times provides an exceptionally well-balanced report that casts doubt on the healthiness of food from farmers markets—read it here. What caught my eye is how the author, a professor at University of Minnesota (my alma mater), lays out a number of positive correlations (being careful not to conclude causation) between farmers markets and various food-borne illnesses, including one attributed to the ‘droppings’ from dogs and the like. But the thing I most admire is him admitting to “a number of dogs that did not bark”, i.e., a number of outbreaks that did not show a statistically significant connection to farmers markets.
This suggestion of possible health issues with farmers markets is heavily hedged—very possibly it will not be borne out by subsequent research. Nevertheless it would only be prudent to thoroughly wash locally grown and sold produce.
Sine illusion makes peaks and valleys on graphs look overly variable
An article in the latest Journal of Computational and Graphical Statistics (JCGS, Vol 24, Num 4, Dec 2015, p1170)) alerted me to a fascinating misperception called the “sine illusion” that causes misinterpretation of trends in variability. See it nicely illustrated here by vision researcher Micheal Bach. The JGCS, Susan VanderPlas and Heike Hofmann, detail “Signs of Sine Illusion—Why We Need to Care” and provide methods to counteract its misleading effects.
If you see a scatter plot that goes up and down with seemingly large scatter at the bends, get out a ruler to get the true perspective. That is my take home message for those like me who like to be accurate in their assessments of data.
“The illusion is explained in terms of a perceptual compromise between the vertical extent and the greater overall dimensions of the section at the turn of the sine-wave figure.”
– RH Day and EJ Stecher, “Sine of an illusion,” Perception, 20; 1991, 49–55.
A knotty problem—how to keep track of stuff without computers or even pen and paper
Posted by mark in history, Uncategorized on January 3, 2016
The New York Times reports today on the recent discovery of several knotted string records, called khipus, that ancient Incas used to record things such as the colorful potatoes I photographed at a Peruvian market. From what I saw on my travels there—see this blog on Incan agriculture experiments, a great deal of food must have been produced and stored.
Based on this Times picture I suspect these “mops that have seen better days”, as George Gheverghese Joseph, a mathematics historian at the University of Manchester, U.K., put it, must be a bit easier to untangle than Christmas lights. But then there remains the problem of deciphering them.
Thus far researchers have picked up on mathematical aspects of the khipus. However, the latest trove of colored strings provides a chance at figuring out the Incan scheme for identifying what was being counted. Here is where database capability and statistical methodology comes in handy.
I amazes me how all of the technology we now have at our disposable is challenged by methods developed 600 or so years ago. Hats off to Incan ‘thinken!
“Many now think that although khipu probably began as accounting tools, they had evolved into a writing system—a kind of three-dimensional binary code, unlike any other on Earth—by the time the Spanish arrived.”
— Cracking the Khipu Code Science magazine.
College textbooks up over 1000 percent since ’77
Bernie Sanders say it’s time to make college tuition free. That would really be radical. A more attainable goal is to make textbooks more affordable. According to an NBC story posted just prior to the current school year, college textbook prices have risen 1,041 percent since 1977, now amounting to over $1200 per student per year. My high school classmate Mark Perry, Professor of Economics at University of Michigan, warns that:
“College textbook prices are increasing way more than parents’ ability to pay for them.”
This tide of expenses for books has been slowed somewhat by the advent of rentals, e.g., $34 for one semester versus $157 to buy Montgomery’s 8th Edition of Design and Analysis of Experiments.
Another way to save that no one dreamed of in ’77 is by buying an e-textbook—Montgomery’s DAOE book costing only $67 in this format.
However, the big breakthrough in reducing the cost of college comes from the University of Minnesota’s Open Textbook Network (OTN), which began offering textbooks for free three years ago. Students over this period saved an estimated 1.5 million dollars, primarily over the past year.*
Professor Gary Oehlert of U Mn School of Statistics—a long-time advisor to Stat-Ease and author of A First Course in Design and Analysis of Experiments**—provides this endorsement of this worthy initiative by my alma mater: “There are several other similar open textbook depositories (OpenStax, etc), but OTN was one of the first to have reviews for the books as well as perhaps being the only one to have a support model for obtaining serious reviews of the books. It also has a broader range of texts than one might anticipate, with math books ranging from high school level through some fairly advanced topics.”
Powerful forces from for-profit publishers and authors who prefer being paid for their hard labor will naturally restrict the spread of free books. Even so, the OTN will certainly put a damper on the rampant inflation of the cost of texts. That will be a big relief for hard-pressed students and their parents.
*Source: “One for the Books”, Minnesota Alumni magazine, Winter 2016, p.12.
**Freely available here under Creative Commons license
Merry DXmas!
Design-Expert software produced this graphic with some wizardry by our Consultant Wayne Adams. Amazing!
Happy holidays!
Technology advanced beyond any hope for healthy curiosity
I am watching the Syfy’s series “Childhood’s End” this week. It is based on a science fiction novel by British author Arthur C. Clarke, one of my favorites growing up. One of the main characters is a very bright boy who at the end of the premier episode last night becomes an astrophysicist, despite this scientific profession being made entirely superfluous by the advanced technology of the alien Overlords.
This morning Robert Scherrer, the chairman of the department of physics and astronomy at Vanderbilt University, lamented in an editorial* for Wall Street Journal that children no longer have any reason to be interested in science, mainly because most of our household gadgets fall into the category of magic—alluding to Clarke’s observation that “any sufficiently advanced technology is indistinguishable from magic.”
“The world’s now placid, featureless, and culturally dead: nothing really new has been created since the Overlords came. The reason’s obvious. There’s nothing left to struggle for, and there are too many distractions and entertainments.”
― Arthur C. Clarke, Childhood’s End
Sleep well or get fat
Posted by mark in Uncategorized, Wellness on November 10, 2015
It’s not often I see a study that focuses on variability but that is what drives this recent New York Times story detailing how “Poor Sleep May Spur College Weight Gain.” Sleep for Science Lab researchers kept track of how 132 first-year students at Brown University slept over a nine week period, during which more than half of them gained nearly six pounds. The increase in weight comes as no surprise (one subject gained 18 pounds!) but this correlating to deviations in sleeping times is provocative. Whether this is causal, or just an offshoot of other upsets in lifestyle that come when students break loose from their parents, remains to be seen in controlled experiment.
Meanwhile, the National Sleep Foundation (NSF) widened ranges earlier this year (2/2/15) in these new recommendations in hours per day broken down by age:
- Newborns (0-3 months): 14-17 (previously 12-18)
- Infants (4-11 months): 12-15 hours (14-15)
- Toddlers (1-2 years): 11-14 hours (12-14)
- Preschoolers (3-5): 10-13 hours (11-13)
- School age children (6-13): 9-11 hours (10-11)
- Teenagers (14-17): 8-10 hours (8.5-9.5)
- Younger adults (18-25): 7-9 hours (new age category)
- Adults (26-64): 7-9 hours (no change)
- Older adults (65+): 7-8 hours (new age category)
My interpretation of all this is to get your teens to follow a fairly regular schedule for sleep (good for those of all ages, I feel sure), but don’t worry too much about the exact amount, provided it falls within the recommended guidelines of NSF.
P.S. In this report published November 4 in Sleep Review, Russell Foster, professor of circadian neuroscience at Oxford, recommends you brush your teeth in the dark to ensure a good sleep. That gives me a bright idea: glow in the dark toothpaste! 🙂
Breaking free of standard practices that no longer make sense
Posted by mark in Consumer behavior, Uncategorized on October 20, 2015
Building off my previous blog on “Why no one wants to monkey around with how things have always been done” I am passing along an insightful anecdote by David Morganstein, President of the American Statistical Association, about his wife’s standard practice to slice a quarter inch off of every ham and toss it in the trash. Read this amazing story and others like it in “The Slice of Ham, How Do You Know?”.
Most of our assumptions have outlived their uselessness.
– Marshall McLuhan, famed for predicting the World Wide Web almost thirty years before it was invented.