Count on placebo effect to counteract colds this coming season
Posted by mark in Consumer behavior on November 23, 2018
In anticipation of seasonal sniffles and hacking coughs I restocked my supply of throat lozenges today. The selection ranged from generics at less than one dollar for a 30-count bag to brands going for several times that price. I see on Amazon a popular choice called “Thieves” that go for $40 for the same quantity that I bought for 97 cents!
Evidently (per National Institute of Health) cough medicines’ effect is mainly placebo. If you do not believe this, go ahead and buy the most expensive lozenges—it will surely make you feel better: Belief enhances the power of a placebo, which means “I shall please” in Latin. You also should consider stocking up in a costly painkiller called Placeberol—laced with Inertaminophen that’s featured in the New York Times Magazine earlier this month* (just kidding).
By the way, the opposite of placebo is the nocebo effect—a belief that an inert ingredient will cause harm to your health. It can be a killer according to this report by the BBC that you can think yourself to death.
My prescription is to maintain a positive attitude to the efficacy of cough drops but pay as little for them as possible. While awaiting the cure for the common cold, I don’t see any alternative.
Talking turkey on extrapolation—do not stick your neck out!
“Extrapolating patterns beyond their natural range can lead to false conclusions” PhD statistician Christine Anderson-Cook in the October issue of Quality Progress. * That should be obvious to everyone. Unfortunately, though, one naturally “eye fits” the final leg of every graph straight out into the beyond of the X axis, thus overlooking the possibility of an imminent bend such as the one shown here by Gapminder. A classic case of this occurred a decade ago when a consensus of forecasters predicted that all of the world economies would experience continued expansion in 2009—none foreseeing the Great Recession. **
“Avoid Linear extrapolation … The turkey’s first 1000 days are a seemingly unending succession of gradually improving circumstances confirmed by daily experience. What happens on Day 1001? Thanksgiving.”
-John E. Sener (Source: Consortium for the Advancement of Undergraduate Statistics Education)
Be careful out there!
* “Straight Line or Not?” p45.
** Financial Times, “An astonishing record – of complete failure”, Tim Harford, May 30, 2014
“Data are profoundly dumb”
Posted by mark in Basic stats & math, Uncategorized on October 22, 2018
This is the controversial view of Judea Pearl and Dana Mackenzie expressed in “Mind over Data”—the lead article in the August issue of Significance. In this excerpt from The Book of Why these co-authors explain “how the founders of modern statistics ‘squandered’ the chance to establish the science of causal inference”. They warn against “falsely believing the answers to all scientific questions reside in the data, to be unveiled through clever data-mining tricks.”
“Lucky is he who has been able to understand the cause of things.”
– Virgil (29 BC)
Pearl and Mackenzie are optimistic that the current “Causal Revolution” will lead to far greater understanding of underlying mechanisms. However, by my reckoning, randomized controlled trials remain the gold standard for establishing cause and effect relationships. Only then can the data speak loud and clear.
Science class a real blast
Making your own erupting volcano is a classic experiment to do at home or in science class. Some, such as the ammonium dichromate one in the video, should not be done without strict supervision and safety precautions.
However, the standard experimental volcano with vinegar and baking soda generally does not cause much harm; that is, until an overzealous mom in India took it to a new level with a blast that injured 59. Fortunately, as reported here, the two students who went to hospital came out intact. Perhaps a non-working volcano might be advised in future.
Hurricane forecasters getting in the zone with their cone
Posted by mark in Nature, Uncategorized on September 13, 2018
Owning a home in Florida, I keep a close eye on the National Hurricane Center (NHC) forecasts throughout the June-through-November season. I see that Florence is now predicted to make landfall around mid-day tomorrow just south of Wilmington, North Carolina—the gateway to the aptly named Cape Fear. Even though that’s less than 24 hours away, it’s still subject to a broad zone of uncertainty–on the order of 100’s of miles. However, it’s very impressive that the NHC forecast made 48 hours ago remains right on. According to a Wall Street Journal report yesterday their 5-day conical projections have improved by more than 100 nautical miles (115 land miles) in radius over the past decade. That’s even more impressive.
However, WSJ advises we’d best remain very leery of the NHC’s spaghetti plots (an alternative to the cone), because they include very simple forecasts along with those that are state-of-the-art. This can be very disconcerting as I discovered when Irma came along a year ago to graze my place along the west coast of Florida. You can see in this Business Insider report on Irma that, only 3 days beforehand, this hurricane’s predicted landfalls ranged from Louisiana to Massachusetts. The article says that NHC suggests that people not focus on the specific tracks. That seems obvious to me based on the ridiculously high variance. I’d like to see these tracks vary by thickness according to the sophistication of the models—the thinner the weaker.
In any case, let’s hope that Florence fizzles out, after all.
Models now running the world
Models will run the world according to this recent op-ed by the Wall Street Journal. This sounds a lot less alarming than a prior essay on “Why Software is Eating the World”. However, one (software) enables the other (modeling).
Models are good in my estimation—beautiful both in mathematical aspects and for their double meaning for those who wear clothes well. For example, the weather-casters often debate during hurricane season whether the American or the European models will prove most accurate. Taking “models” in the fashionable sense, it’s no contest—the winner is American Karly Kloss, who’s Kode with Klossy camps empower girls to code. She is the cross over—a model who can model.
“In the hunt for competitive advantage, model-driven companies will accelerate away from the pack.”
– Steven A. Cohen and Matthew W. Granade
Marketers trick math-challenged consumers with ploys on percentages
Posted by mark in Consumer behavior on August 5, 2018
WSJ’s “The Numbers” columnist, Jo Craven McGinty, advised readers “To Shop Smart, Mind the Percentages” in this weekend’s issue. It turns out that, as I blogged back in 2007, percentages are puzzling to many people. Put yourself to this test from McGinty: You can buy a regular container of ice cream at 33% off (option 1) or pay the usual price for a 33% more of it as a free bonus (option 2). If you picked the first option without any hesitation, you go to the head of the class. Those of you—likely the majority of the general population–who went for option 2 are the target for the marketers.
“People always go for the bonus.” – Quote in WSJ from Akshay Rao, marketing professor, University of Minnesota and co-author of When Two and Two is Not Equal to Four: Errors in Processing Multiple Percentage Changes
The remainder who withheld judgement until they do the calculation get full credit for knowing that percentages require thinking to work out their effect. Kudos to you for being math-savvy.
“To be statistically literate, one must be able to form arithmetic comparisons of any two numbers.”
– Milo Schield, Department of Business, Accounting and MIS, Augsburg College, Minneapolis, “Common Errors In Forming Arithmetic Comparisons”, Sept 1999, Association of Public Data Users, Volume 1.51 Journal Of Significance
Fireworks that do not go Fourth deserve a resounding fizgig
My word for today is “fizgig”, a type of firework that makes a loud hiss. I consider this an onomatopoeic word, given the “fiz” characterizes the sound.
Residents of Saint Paul, where I grew up, must be fizzing their mayor today after he canceled the city’s fireworks this year due to budget concerns. Boo, hiss!
More commonly, fireworks are frowned upon due to safety concerns. For example, a Florida television station broadcast a warning yesterday that Independence Day revelers should be careful not to “be a statistic” by shooting off fireworks. I don’t get this. Isn’t being a statistic a good thing?
An enterprising fireworks vendor turned the statistics around in a very creative way by touting a long-term trend toward fewer injuries per pound of pyrotechnics—citing a decrease of more than 50% since 1994. I find this fellows numerical and sentimental arguments in favor of fireworks very compelling: Check it out here.
“It ought to be solemnized with Pomp and Parade, with Shews, Games, Sports, Guns, Bells, Bonfires and Illuminations from one End of this Continent to the other from this Time forward forever more.”
– John Adams, July 3, 1776
Are you seeing red sorting out how much sunscreen to apply and at what strength?
As pointed out in this weekend’s Wall Street Journal, this is the burning question for those of us in the northern hemisphere as we enjoy our brightest days of the year. According to The Numbers columnist Jo Craven McGinty, the answers are:
- At least 2 milligrams per square centimeter—about one-fifth the depth of a piece of paper. Although this seems very thin, most people only put on about half that amount, so you’d best apply your sunscreen twice.
- Go for at least an SPF of 50. However, do not sweat it if all you have on hand is SPF 30, or the level 50 costs more than you care to pay—the difference is minor as you can see in the graph I made by fitting the FDA data WSJ
provided to a curve (using Design-Expert® software).
It seems to me it would be best to stay out of the sun as much as possible, and, when you do, cover up, but seek the advice of your physician or other qualified health provider, such as a dermatologist. Be careful out there!
PS. McGinty detailed a study that found a significant advantage to increasing SPF from 50 to 100. It being done via a randomized split on 199 snow skiers bodes well. However, the research was sponsored by a sunscreen maker. An independent, contrary view of 100 SPF pros and cons is laid out here.
Fascination for pendulums piqued by Foucault’s in France
Posted by mark in science, Uncategorized on June 26, 2018
Earlier this month I visited the Pantheon in Paris where I observed this attendant recalibrating Foucault’s pendulum.
This French scientist’s elegant scientific demonstration of earth’s rotation has delighted observers like me since 1851. For more on this story read this Ask Smithsonian blog. Unfortunately, one morning in 1998, the cable on the 52-foot long pendulum at Smithsonian’s National Museum of Natural History (originally History and Technology when opened in 1964) snapped, nearly ‘clocking’ a staffer with its wayward 240-pound brass bob. This Foucault device being unAmerican and dangerous, it was removed in favor of the Star-Spangled Banner Preservation Project, thus eliminated a favored place for folks to gather.
By the way, I am now reading The Discoverers by The Librarian of Congress, Daniel Boorstin—the first in his wonderful Knowledge Trilogy. There, coincidentally, I learned that Galileo—only 19-years old at the time (1583) and bored by a church service in Pisa—became distracted by the swinging of a chandelier. By timing his pulse, he observed the time of a pendulum being independent of its arc length—an important discovery of a property called isochronism. This simple discovery, as pointed out by Boorstin, began a new age where science developed from observation and measurement rather than pure speculation.