Sandbox: Difference between revisions

From ChanceWiki
Jump to navigation Jump to search
(6 intermediate revisions by the same user not shown)
Line 5: Line 5:
:by John Bohannon, io9.com, 27 May 2015
:by John Bohannon, io9.com, 27 May 2015


Publishing under the pseudonym Johannes Bohannon, at his own respectably named [http://instituteofdiet.com Institute of Diet and Health],  Bohannon announced the results of a deliberately faulty study designed to show that eating chocolate promotes weight loss.  These findings should have sounded too good to be true, but that didn't stop a host of media outlets from uncritically reporting the story.
Publishing under the pseudonym Johannes Bohannon, at his own respectably named website, the Institute of Diet and Health],  Bohannon [http://instituteofdiet.com/2015/03/29/international-press-release-slim-by-chocolate/ announced the results] of a deliberately faulty study designed to show that eating chocolate promotes weight loss.  These findings should have sounded too good to be true, but that didn't stop the news and social media outlets from uncritically reporting the findings.  The i09.com story linked above includes screen captures from a number of these reports.


==Some math doodles==
The whole story is worth reading for details of how the hoax was conceived and conducted.  You can also listen to [http://www.npr.org/sections/thesalt/2015/05/29/410609184/trickster-journalist-explains-why-he-duped-the-media-on-chocolate-study an NPR interview] ("All Things Considered," 29 May 2015).  The study used Facebook to recruit a mere 15 volunteers, and randomly assigned them to one of three groups:  low carb diet, low carb diet plus a daily chocolate bar, and a control group.  Subjects weighed themselves daily, reported on sleep quality and other measures, and had variety of measures taken from blood work, for a total of 18 variables.  As Bohannon explains, as long as you don't specify up front what you are looking for, a study with so few subjects and so many variables is bound to turn up a statistically significant result.  On ion.com, he calls the design "a recipe for false positives."
<math>P \left({A_1 \cup A_2}\right) = P\left({A_1}\right) + P\left({A_2}\right) -P \left({A_1 \cap A_2}\right)</math>


Of course, such results will not be reproducible, but the story went viral before anyone asked hard questions.  NRP host Robert Siegel wondered if we can take some solace in the fact that reputable sources like the Associated Press or the ''New York Times'' did not pick up the story.  Bohannon responds that the tabloid press, with its much larger readership, had already done the damage.


==Parenting time==
For additional perspective on all of this, see [http://www.npr.org/sections/13.7/2015/06/08/412825282/what-junk-food-can-teach-us-about-junk-science What junk food can teach us about junk science] (NPR, 8 June 2015).
[http://www.nytimes.com/2015/04/02/upshot/yes-your-time-as-a-parent-does-make-a-difference.html?abt=0002&abg=1 Yes, your time as a parent does make a difference]<br>
by Justin Wolfers, "Upshot" blog, ''New York Times'', 1 April 2015


[http://www.nytimes.com/2015/04/03/upshot/why-a-claim-about-the-irrelevance-of-parenting-time-doesnt-add-up.html?rref=upshot&module=Ribbon&version=context&region=Header&action=click&contentCollection=The%20Upshot&pgtype=article&abt=0002&abg=1 Why a claim about the irrelevance of parenting time doesn’t add up]<br>
==Some math doodles==
by Justin Wolfers, "Upshot" blog, ''New York Times'', 2 April 2015
<math>P \left({A_1 \cup A_2}\right) = P\left({A_1}\right) + P\left({A_2}\right) -P \left({A_1 \cap A_2}\right)</math>
 
In this pair of articles, Wolfers seeks to debunk a study on parenting time that was widely reported in the media (he cites articles from [http://www.washingtonpost.com/local/making-time-for-kids-study-says-quality-trumps-quantity/2015/03/28/10813192-d378-11e4-8fce-3941fc548f1c_story.html The Washington Post], [http://www.theguardian.com/commentisfree/2015/apr/01/dont-stress-out-our-kids-are-just-fine-when-their-mothers-work-late The Guardian], and [http://www.today.com/parents/quality-over-quantity-new-study-brings-time-squeezed-parents-relief-t11746 NBC News], among others).  The study in question failed to find a significant correlation between parental time spent with children and outcomes later in life, such as scores on standardized tests.  The common theme of all the media reports was that "quality beats quantity" in parenting time.
 
Wolfers's first article pointed out that the study in question was based on a survey that asked parents about two specific days, one during the the week and one on the weekend.  He compares this with trying to predict your income by looking a particular day:  the result would vary wildly based on whether the day in question happened to be a payday.  Similary he quotes developmental psychologists saying, “What you did yesterday should not be taken as representative of what you did last year.”
 
The second article provides responds to some readers's objections, and gives a particularly careful discussion of statistical issues related to "errors in variables."  Wolfers acknowledges that randomness in the sample could reasonably be expected to guarantee that the average parenting time balances out correctly in the measure, in that some parents will respond about more time-intensive days, and others about lesser days.  The real problem, he explains, comes in correlating this with another measure.  To illustrate, he constructs a set of three scatterplots, the first showing a positive correlation between the time you spent with your children today and the  time you typically spend, the second showing a positive correlation between test scores and the time you typically spend, but the third showing near-zero correlation between test scores and the time you spend with your children today.


Submitted by Bill Peterson


==Accidental insights==
==Accidental insights==

Revision as of 14:34, 9 June 2015

A fake study on chocolate

Paul Alper sent a link to the following:

I fooled millions into thinking chocolate helps weight loss. Here's how.
by John Bohannon, io9.com, 27 May 2015

Publishing under the pseudonym Johannes Bohannon, at his own respectably named website, the Institute of Diet and Health], Bohannon announced the results of a deliberately faulty study designed to show that eating chocolate promotes weight loss. These findings should have sounded too good to be true, but that didn't stop the news and social media outlets from uncritically reporting the findings. The i09.com story linked above includes screen captures from a number of these reports.

The whole story is worth reading for details of how the hoax was conceived and conducted. You can also listen to an NPR interview ("All Things Considered," 29 May 2015). The study used Facebook to recruit a mere 15 volunteers, and randomly assigned them to one of three groups: low carb diet, low carb diet plus a daily chocolate bar, and a control group. Subjects weighed themselves daily, reported on sleep quality and other measures, and had variety of measures taken from blood work, for a total of 18 variables. As Bohannon explains, as long as you don't specify up front what you are looking for, a study with so few subjects and so many variables is bound to turn up a statistically significant result. On ion.com, he calls the design "a recipe for false positives."

Of course, such results will not be reproducible, but the story went viral before anyone asked hard questions. NRP host Robert Siegel wondered if we can take some solace in the fact that reputable sources like the Associated Press or the New York Times did not pick up the story. Bohannon responds that the tabloid press, with its much larger readership, had already done the damage.

For additional perspective on all of this, see What junk food can teach us about junk science (NPR, 8 June 2015).

Some math doodles

<math>P \left({A_1 \cup A_2}\right) = P\left({A_1}\right) + P\left({A_2}\right) -P \left({A_1 \cap A_2}\right)</math>


Accidental insights

My collective understanding of Power Laws would fit beneath the shallow end of the long tail. Curiosity, however, easily fills the fat end. I long have been intrigued by the concept and the surprisingly common appearance of power laws in varied natural, social and organizational dynamics. But, am I just seeing a statistical novelty or is there meaning and utility in Power Law relationships? Here’s a case in point.

While carrying a pair of 10 lb. hand weights one, by chance, slipped from my grasp and fell onto a piece of ceramic tile I had left on the carpeted floor. The fractured tile was inconsequential, meant for the trash.

BrokenTile.jpg

As I stared, slightly annoyed, at the mess, a favorite maxim of the Greek philosopher, Epictetus, came to mind: “On the occasion of every accident that befalls you, turn to yourself and ask what power you have to put it to use.” Could this array of large and small polygons form a Power Law? With curiosity piqued, I collected all the fragments and measured the area of each piece.

Piece Sq. Inches % of Total
1 43.25 31.9%
2 35.25 26.0%
3 23.25 17.2%
4 14.10 10.4%
5 7.10 5.2%
6 4.70 3.5%
7 3.60 2.7%
8 3.03 2.2%
9 0.66 0.5%
10 0.61 0.5%
Montante plot1.png

The data and plot look like a Power Law distribution. The first plot is an exponential fit of percent total area. The second plot is same data on a log normal format. Clue: Ok, data fits a straight line. I found myself again in the shallow end of the knowledge curve. Does the data reflect a Power Law or something else, and if it does what does it reflect? What insights can I gain from this accident? Favorite maxims of Epictetus and Pasteur echoed in my head: “On the occasion of every accident that befalls you, remember to turn to yourself and inquire what power you have to turn it to use” and “Chance favors only the prepared mind.”

Montante plot2.png

My “prepared” mind searched for answers, leading me down varied learning paths. Tapping the power of networks, I dropped a note to Chance News editor Bill Peterson. His quick web search surfaced a story from Nature News on research by Hans Herrmann, et. al. Shattered eggs reveal secrets of explosions. As described there, researchers have found power-law relationships for the fragments produced by shattering a pane of glass or breaking a solid object, such as a stone. Seems there is a science underpinning how things break and explode; potentially useful in Forensic reconstructions. Bill also provided a link to a vignette from CRAN describing a maximum likelihood procedure for fitting a Power Law relationship. I am now learning my way through that.

Submitted by William Montante


The p-value ban

http://www.statslife.org.uk/opinion/2114-journal-s-ban-on-null-hypothesis-significance-testing-reactions-from-the-statistical-arena