Chance News 105: Difference between revisions

From ChanceWiki
Jump to navigation Jump to search
Line 47: Line 47:
The article gives the success rate for the 2-point play over this 20-year period as 47.5%, compared with 99% for the kick. An easy expected value calculation gives 0.95 points for the 2-point play compared with 0.99 points for the kick.  It is not surprising, then, that 2-point attempts have been elected only 5% of the time (typically for strategic reasons involving the difference in scores between the two teams at the time).  
The article gives the success rate for the 2-point play over this 20-year period as 47.5%, compared with 99% for the kick. An easy expected value calculation gives 0.95 points for the 2-point play compared with 0.99 points for the kick.  It is not surprising, then, that 2-point attempts have been elected only 5% of the time (typically for strategic reasons involving the difference in scores between the two teams at the time).  


The article describes several proposed changes that were under consideration to encourage 2-point attempts. As announced [http://www.nfl.com/news/story/0ap3000000493347/article/nfl-moves-extra-point-to-15yard-line-for-2015-season here], the final decision was to to move the kick to the 15-yard line, where the estimated success rate is only 90%.  (n addition, the defense can score 2 points by returning failed attempts by either method; e.g., blocking the kick or intercepting a pass).
The article describes several proposed rule changes that were under consideration to encourage 2-point attempts. As announced [http://www.nfl.com/news/story/0ap3000000493347/article/nfl-moves-extra-point-to-15yard-line-for-2015-season here], the final decision was to to move the kick to the 15-yard line, where the estimated success rate is only 90%.  (In addition, the defense can score 2 points by returning failed attempts by either method; e.g., blocking the kick or intercepting a pass.)


==A fake study on chocolate==
==A fake study on chocolate==

Revision as of 20:17, 16 July 2015

Quotations

Small news item:
“When ants go exploring in search of food they end up choosing collective routes that fit statistical distributions of probability [mixture of Gaussian and Pareto], according to a team of mathematicians who analyzed the trails of a species of Argentine ant. ....”

“The Mathematical Pattern of Ants Marching”, Science 2.0, May 12, 2015

Submitted by Margaret Cibes


“The problem is simple but daunting. The foundation of opinion research has historically been the ability to draw a random sample of the population. That’s become much harder to do.”

--Nate Silver, quoted in: What's the matter with polling, New York Times, 20 June 2015

Submitted by Bill Peterson

Forsooth

Statistical Significance.
Definition: A mathematical technique to measure whether the results of a study are likely to be true. Statistical significance is calculated as the probability that an effect observed in a research study is occurring because of chance.

in: Glossary of Terms, Agency for Health Care Research and Quality, Dept of Heath and Human Services

Submitted by Bill Peterson

Italian scientists cleared in Aquila case

“Scientists jailed for manslaughter ... cleared”
Daily Mail.com, November 11, 2014

In 2012 seven Italian scientists were sentenced to 6 years each, and fined a total of about $7 million in damages, because of their “misleadingly reassuring statement” to the people of Aquila just prior to a 6.3 earthquake that resulted in the deaths of 309 people.

Some of the scientists defended their actions in a 2013 Scientific American article [1].

In 2014 six of them were cleared of all charges, and the seventh, who had claimed that there was “no danger,” had his sentence reduced to two years.

For more about this case, see Chance News 91 and Chance News 89.

Submitted by Margaret Cibes

Expected value in the NFL

Mike Olinick sent a link to the following:

N.F.L.explores making the 2-point conversion more tempting
by Victor Mather, New York Times, 19 May 2015

As football fans know, following a touchdown, the ball is placed at the two-yard line, where a team has the option of either attempting a 1-point kick or a 2-point play from scrimmage. The 2-point option was introduced 20 years ago to make things more interesting, since the kick had become an almost automatic point.

The article gives the success rate for the 2-point play over this 20-year period as 47.5%, compared with 99% for the kick. An easy expected value calculation gives 0.95 points for the 2-point play compared with 0.99 points for the kick. It is not surprising, then, that 2-point attempts have been elected only 5% of the time (typically for strategic reasons involving the difference in scores between the two teams at the time).

The article describes several proposed rule changes that were under consideration to encourage 2-point attempts. As announced here, the final decision was to to move the kick to the 15-yard line, where the estimated success rate is only 90%. (In addition, the defense can score 2 points by returning failed attempts by either method; e.g., blocking the kick or intercepting a pass.)

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).

(See also Chance News 96 for a story of how Bohannon instructively gamed the peer review system.)

Tanzania’s new Statistics Act under fire

The June 2015 issue of Significance magazine contained a brief piece about Tanzania’s new Statistics Act, noting that an up-to-date version of the act was not publicly available at press time. A National Geographic blogger[2] stated that, as of May 13, 2015, there was no official word about whether the president had signed the act.

Nevertheless, a Tanzanian blogger has prepared a 6-page critique of the act -- “The Statistics Act, 2013 – A rapid analysis”, August 4, 2015 [sic] -- in which he discusses his 5 key areas of concern:
1. Uncertainty about who is allowed to generate statistics and what authorization is required
2. Restrictive rules about disseminating survey micro-data
3. Obstacles to whistleblowing without any public interest protections
4. Severe restrictions on the publication or communication of any “contentious” statistical information, and making illegal the publishing/communicating of “false” statistical information or information that “may result in the distortion of facts,” with no protections for acting in good faith
5. Severe penalties for those found guilty of offenses under the bill

For more discussion of potentially problematic aspects of the act, by an English blogger with the adopted Tanzanian name of Mtega[3], see:
“Three (government) statistics that could be illegal under Tanzania’s new Statistics Act”, and
“Four bills later”.

For the act itself, without any 2014 amendments, see: “The Statistics Act, 2013”, United Republic of Tanzania, June 2013.

Submitted by Margaret Cibes

A Dilbert forsooth?

Hedging your bets on the God Hypothesis
by Scott Adams, Scott Adams Blog, 3 July 2015

Adams juxtaposes the following two apparently contradictory statistics, presented in a recent article from Business Insider (3 July 2015), which begins, "Although the US might seem more polarized than ever, Americans can still find common ground on plenty or topics."

  • 54% of Americans are "very confident" in the existence of a supreme being. (Source: 2014 AP-GfK poll)
  • 70% of Americans identify as Christian, although the number has been declining in recent years. (Source: Pew Research Center 2014 Religious Landscape study)

As Adams notes, Christians are presumably a subset of all people who believe in a supreme being.

Discussion
Looking more closely at the survey questions, the AP-GfK asked for degree of confidence in the statement, "The universe is so complex, there must be a supreme being guiding its creation;" whereas Pew asked respondents to self-identity according to religious affiliation. What effect might this have on the results?

Submitted by Bill Peterson

Surrogate markers

Paul Alper sent a link to the following cartoon, from the Statistically funny blog.

Biomarkers-Unlimited.jpg

There is actually a serious discussion here, and the accompanying text cites viral load in HIV studies as an important example. However,

There is a lot of controversy about surrogate outcomes - and debates about what's needed to show that an outcome or measure is a valid surrogate we can rely on. They can lead us to think that a treatment is more effective than it really is.

Yet a recent investigative report found that cancer drugs are being increasingly approved based only on surrogate outcomes, like "progression-free survival." That measures biomarker activity rather than overall survival (when people died).

Or, quoting Paul on his favorite surrogate marker: "If your car's dashboard light goes on, covering it up with duct tape solves the problem immediately but the car dies at the same time as if you had done nothing."