Chance News 48: Difference between revisions

From ChanceWiki
Jump to navigation Jump to search
 
(17 intermediate revisions by 3 users not shown)
Line 31: Line 31:


Submitted by Charles Grinstead
Submitted by Charles Grinstead
----
Helene Hadsell, an 83 year old with a "4-Step Secret That Has Helped Her Win Limitless Prizes & Contests, ... And Even Cheat Death" and "The Woman Who Wins Every Contest She Enters," has written a book,
<center>"Against All Odds: The Art of Mind Enhanced Winning."[http://www.thewinningsage.com/lp/win-lottery-in-life-1?sr=1&gclid=CIWxouff8JoCFQIWFQodIXdzJg]</center>
Submitted by Margaret Cibes


== Winning system for dice game?==
== Winning system for dice game?==
Line 84: Line 89:
However, you won’t be able to read the Kuklo paper to find out what is meant by “a significantly higher union rate in the group treated with rhBMP-2 (92% vs 76%)” because the paper has been withdrawn by Scott and the JBJS.  The reasons may be found in [http://www.nytimes.com/2009/05/13/business/13surgeon.html Wilson and Meier’s reporting in The New York Times].  The headline is “Doctor Falsified Study on Injured G.I.’s, Army says.”  Not only did he “not obtain the Army’s required permission to conduct the study” but also “Army investigators found that Dr. Kuklo forged the signatures of four Walter Reed doctors on the article before submitting it last year to a British medical journal, falsely claiming them as co-authors.”  In addition, “the total number of patients Dr. Kuklo reported as having been treated for extensive lower leg wounds at Walter Reed during the study period—138 soldiers—was greater than the number for which the hospital could find records.”
However, you won’t be able to read the Kuklo paper to find out what is meant by “a significantly higher union rate in the group treated with rhBMP-2 (92% vs 76%)” because the paper has been withdrawn by Scott and the JBJS.  The reasons may be found in [http://www.nytimes.com/2009/05/13/business/13surgeon.html Wilson and Meier’s reporting in The New York Times].  The headline is “Doctor Falsified Study on Injured G.I.’s, Army says.”  Not only did he “not obtain the Army’s required permission to conduct the study” but also “Army investigators found that Dr. Kuklo forged the signatures of four Walter Reed doctors on the article before submitting it last year to a British medical journal, falsely claiming them as co-authors.”  In addition, “the total number of patients Dr. Kuklo reported as having been treated for extensive lower leg wounds at Walter Reed during the study period—138 soldiers—was greater than the number for which the hospital could find records.”


Further, according to Scott, the forgery came to light because “Shortly after the paper was published we received correspondence from one of the persons identified as a co-author indicating than the alleged co-authors had not seen the manuscript prior to publication and they had not signed the letter of transmittal.  It was further disclosed that much of the paper was essentially false.”
Further, according to Scott, the forgery came to light because “Shortly after the paper was published we received correspondence from one of the persons identified as a co-author indicating that the alleged co-authors had not seen the manuscript prior to publication and they had not signed the letter of transmittal.  It was further disclosed that much of the paper was essentially false.”


Discussion
Discussion
Line 204: Line 209:


A study in Experimental Aging Research shows that old people may be subject to the effects of "stereotype threat," which refers to a situation in which "some members of stigmatized groups, when faced with stressful situations, expect themselves to do worse – a prophecy that fulfills itself."<br>
A study in Experimental Aging Research shows that old people may be subject to the effects of "stereotype threat," which refers to a situation in which "some members of stigmatized groups, when faced with stressful situations, expect themselves to do worse – a prophecy that fulfills itself."<br>
The study is said to describe an experiment in which psychologists at North Carolina State University in Raleigh recruited 103 volunteers, ages 60 to 82, to perform simple arithmetic and recall tests. The researchers told about half of the participants that the purpose of the tests was "to examine aging effects on memory" and were asked to write down their ages before beginning the tests.  A control group was told that the tests had been constructed to correct for age-related bias, but these participants were not asked to write down their ages.  According to the article's author, members of the "treated" group performed significantly worse on the memory tests than the control group, and they even performed worse than they had on a pre-experiment screening test.<br>
The study is said to describe an experiment in which psychologists at North Carolina State University in Raleigh recruited 103 volunteers, ages 60 to 82, to perform simple arithmetic and recall tests. The researchers told about half of the participants that the purpose of the tests was "to examine aging effects on memory" and were asked to write down their ages before beginning the tests.  A control group was told that the tests had been constructed to correct for age-related bias, but these participants were not asked to write down their ages.  According to the article's author, members of the "treated" group performed significantly worse on the memory tests than the control group, and they even performed worse than they had on a pre-experiment screening test.<br>
In the print version of this article (June 1, 2009, issue), the author cites Baruch College psychologist Catherine Good, who advises preparers of standardized tests to move questions about personal demographics to the ends of tests, in order to mitigate the effect of "stereotype threat" on student test takers.  Professor Good, along with Steve Stroessner and Lauren Webster, runs the website Reducing Stereotype Threat [http://reducingstereotypethreat.org].  The author also credits Stanford University social psychologist Claude Steele [http://steele.socialpsychology.org/] with the original use of the term "stereotype threat."<br>
In the print version of this article (June 1, 2009, issue), the author cites Baruch College psychologist Catherine Good, who advises preparers of standardized tests to move questions about personal demographics to the ends of tests, in order to mitigate the effect of "stereotype threat" on student test takers.  Professor Good, along with Steve Stroessner and Lauren Webster, runs the website Reducing Stereotype Threat [http://reducingstereotypethreat.org].  The author also credits Stanford University social psychologist Claude Steele [http://steele.socialpsychology.org/] with the original use of the term "stereotype threat."<br>
Submitted by Margaret Cibes<br>
Submitted by Margaret Cibes<br>


Line 243: Line 251:
"Facebook and Procrastination" [http://www.cjr.org/the_observatory/facebook_and_procrastination.php],  
"Facebook and Procrastination" [http://www.cjr.org/the_observatory/facebook_and_procrastination.php],  
by Earle Holland, Columbia Journalism Review, May 8, 2009<br>
by Earle Holland, Columbia Journalism Review, May 8, 2009<br>
This article describes the unfortunate media frenzy over an Ohio State University graduate student's preliminary research results about an association between students' uses of Facebook and their GPAs.  It describes what was "a good lesson in the inherent risks of reporters’ cavalierly covering the social sciences, as well as the risks that young researchers can face in dealing with the news media."<br>
This article describes the unfortunate media frenzy over an Ohio State University graduate student's preliminary research results about an association between students' uses of Facebook and their GPAs.  It describes what was "a good lesson in the inherent risks of reporters’ cavalierly covering the social sciences, as well as the risks that young researchers can face in dealing with the news media."<br>
According to the article's author, Aryn Karpinski, the study's co-author, stipulated that the research was a pilot study and that, while the authors did  find a correlation between Facebook use and GPA, they did not find a causal relationship.  She also acknowledged that more research was needed.  In a March press release reporting the study, the University asked that reports of the study not be published until the study had been presented at an April 16 educational research conference.<br>  
According to the article's author, Aryn Karpinski, the study's co-author, stipulated that the research was a pilot study and that, while the authors did  find a correlation between Facebook use and GPA, they did not find a causal relationship.  She also acknowledged that more research was needed.  In a March press release reporting the study, the University asked that reports of the study not be published until the study had been presented at an April 16 educational research conference.<br>  
Nevertheless the report appeared before the conference, in the Sunday Times of London, which stated:
Nevertheless the report appeared before the conference, in the Sunday Times of London, which stated:
<blockquote>Research finds the website [Facebook] is damaging students’ academic performance. … Facebook users … are more likely to perform poorly in exams, according to new research. … The majority of students who use Facebook every day are underachieving by as much as an entire grade compared with those who shun the site.</blockquote>
<blockquote>Research finds the website [Facebook] is damaging students’ academic performance. … Facebook users … are more likely to perform poorly in exams, according to new research. … The majority of students who use Facebook every day are underachieving by as much as an entire grade compared with those who shun the site.</blockquote>
Line 281: Line 291:
<blockquote>Four members of the sample ran for the U.S. Senate [and one was the renowned journalist Ben Bradlee]. One served in a presidential Cabinet, and one was president [JFK]. There was a best-selling novelist [not Norman Mailer]. But hidden amid the shimmering successes were darker hues. As early as 1948, 20 members of the group displayed severe psychiatric difficulties. By age 50, almost a third of the men had at one time or another met Vaillant’s criteria for mental illness.  ....  “They were normal when I picked them,” [Bock] told Vaillant in the 1960s. “It must have been the psychiatrists who screwed them up.”</blockquote>
<blockquote>Four members of the sample ran for the U.S. Senate [and one was the renowned journalist Ben Bradlee]. One served in a presidential Cabinet, and one was president [JFK]. There was a best-selling novelist [not Norman Mailer]. But hidden amid the shimmering successes were darker hues. As early as 1948, 20 members of the group displayed severe psychiatric difficulties. By age 50, almost a third of the men had at one time or another met Vaillant’s criteria for mental illness.  ....  “They were normal when I picked them,” [Bock] told Vaillant in the 1960s. “It must have been the psychiatrists who screwed them up.”</blockquote>
According to the article's author, "[I]t turned out that the lives were too big, too weird, too full of subtleties and contradictions to fit any easy conception of “successful living.”<br>   
According to the article's author, "[I]t turned out that the lives were too big, too weird, too full of subtleties and contradictions to fit any easy conception of “successful living.”<br>   
Vaillant has his own goal with respect to the Grant Study:  "His central question is not how much or how little trouble these men met, but rather precisely how—and to what effect—they responded to that trouble."<br>
Vaillant has his own goal with respect to the Grant Study:  "His central question is not how much or how little trouble these men met, but rather precisely how—and to what effect—they responded to that trouble."<br>
He "identified seven major factors that predict healthy aging, both physically and psychologically [– e]mploying mature adaptations ..., education, stable marriage, not smoking, not abusing alcohol, some exercise, and healthy weight."
He "identified seven major factors that predict healthy aging, both physically and psychologically [– e]mploying mature adaptations ..., education, stable marriage, not smoking, not abusing alcohol, some exercise, and healthy weight."
<blockquote>Of the 106 Harvard men who had five or six of these factors in their favor at age 50, half ended up at 80 as what Vaillant called “happy-well” and only 7.5 percent as “sad-sick.” Meanwhile, of the men who had three or fewer of the health factors at age 50, none ended up “happy-well” at 80. Even if they had been in adequate physical shape at 50, the men who had three or fewer protective factors were three times as likely to be dead at 80 as those with four or more factors.</blockquote>
<blockquote>Of the 106 Harvard men who had five or six of these factors in their favor at age 50, half ended up at 80 as what Vaillant called “happy-well” and only 7.5 percent as “sad-sick.” Meanwhile, of the men who had three or fewer of the health factors at age 50, none ended up “happy-well” at 80. Even if they had been in adequate physical shape at 50, the men who had three or fewer protective factors were three times as likely to be dead at 80 as those with four or more factors.</blockquote>
Line 312: Line 324:
The Wall Street Journal, June 1, 2009<br>
The Wall Street Journal, June 1, 2009<br>


See charts of (a)"Daily closing share price since 1926, adjusted for splits and spinoffs," (b) "Monthly closes, inflation-adjusted," and (c) "Monthly closes shown on a logarithmic scale." [http://s.wsj.net/public/resources/documents/WSJ_GMStockA_090531.pdf]<br>
See charts for General Motors stock of (a)"Daily closing share price since 1926, adjusted for splits and spinoffs," (b) "Monthly closes, inflation-adjusted," and (c) "Monthly closes shown on a logarithmic scale." [http://s.wsj.net/public/resources/documents/WSJ_GMStockA_090531.pdf]<br>
Submitted by Margaret Cibes<br>
Submitted by Margaret Cibes<br>
==Governor proposes state Keno game==
"Rell Gives Keno Address: www.youlosebig.com"  [http://www.courant.com/news/local/columnists/hc-colin-mcenroe-keno-0531.artmay31,0,5262574.column]<br>
by Colin McEnroe, The Hartford Courant, May 31, 2009<br>
The governor of Connecticut recently proposed a state-run Keno operation as a way of raising an additional 60 million dollars.<br> 
According to the columnist, "if you play Keno a lot, you will be more likely to get back about 60 dollars of every 100 dollars you put in."
<blockquote>The basic idea is that 20 numbers of a possible 80 are picked at random. You have the opportunity, beforehand, to guess as many as 10 of them.  ....  States with Keno machines typically run the games every five minutes. In Michigan, for example, Club Keno drawings take place every five minutes seven days a week from 6:05 a.m. until 1:45 a.m.</blockquote>
The columnist also states that "Many of the machines are in bars ..., where patrons can consume nourishing and judgment-improving beverages while they decide how much money to lose.  A weekly Powerball ticket is to Keno games as a large cup of coffee is to crack cocaine."<br>
Submitted by Margaret Cibes<br>
==The U.K. stats police==
"This U.K. Sheriff Cites Officials for Serious Statistical Violations" [http://online.wsj.com/article/SB124398720006979435.html]<br>
by Carl Bialik (The Numbers Guy), The Wall Street Journal, June 3, 2009<br>
The U.K. instituted a Statistics Authority in 2007 with the "budget, authority and independence to question other government agencies on the numbers they release to the public."<br>
Poll results have suggested the public's distrust of government statistics:
<blockquote>Among 1,708 adults who answered a 2004 survey by the Office for National Statistics ... , just one in six agreed with the statement, "Official figures are produced without political interference."  ....  A poll conducted three years later found that just 33% of U.K. respondents said they tend to trust official statistics, the lowest rate among the 27 member states of the European Union.</blockquote>
Discussion<br>
1.  Did the U.K. public appear to have become more trusting of government statistics, or less trusting, over the period 2004-2007?<br>
2.  What unstated assumption(s) do you need to make in order to be able to interpret the margin of error given with the table?<br>
Submitted by Margaret Cibes<br>
==Alcohol's Good for You?==
From istostat<br>
There is an article in today's New York Times Health section about moderate drinking and health benefits.  It discusses the difficulty in determining the causation that underlies this and other correlations of health variables, and could be interesting for discussion in a conceptually based intro stats course.
http://www.nytimes.com/2009/06/16/health/16alco.html?em
"Alcohol's Good for You?  Some Scientists Doubt It" by Roni Caryn Rabin, New York Times, 6/16/09
Annette Gourgey
CUNY

Latest revision as of 11:54, 11 July 2011

Quotations

Wall Street takes your money and their experience and
turns it into their money and your experience

For source see Item 6

In other words, the population of a city is [according to Zipf's law], to a good approximation, inversely proportional to its rank [within its country]. Why this should be true, no one knows.

Steven Strogatz.
Guest Column
The New York Times
May 19 2009

Submitted by Paul Alper

Note: Strogatz is a good writer for the public and your students might enjoy reading this.

Laurie Snell


Forsooths

This costs about a billion dollars. I'm not going to say how many zeros after it, but it's a lot.

CT State Senator, Floor debate, CT-N, May 30, 2009

Submitted by Margaret Cibes


A quote from today's NY Times. In the sports section, there's an article on lightning strikes and golf. John Jensenius, a National Weather Service lightning safety expert, is quoted in the article as saying:

I study the case histories of all lightning fatalities.
Often, if people had gotten inside 5 or 10 minutes earlier, they would be alive.

Submitted by Charles Grinstead


Helene Hadsell, an 83 year old with a "4-Step Secret That Has Helped Her Win Limitless Prizes & Contests, ... And Even Cheat Death" and "The Woman Who Wins Every Contest She Enters," has written a book,

"Against All Odds: The Art of Mind Enhanced Winning."[1]

Submitted by Margaret Cibes

Winning system for dice game?

"Accused Cheater's Trial Includes Lesson In Craps"[2]
by Karen Florin, The Day (New London, CT), May 20, 2009

A Tennessee man has been charged with cheating at the craps table of the Foxwoods Casino in Ledyard, CT. During the trial he claimed that he was a "professional gambler who had a winning system that did not involve bribing dealers to pay him for late or illegal bets." (Closing arguments were scheduled for May 21, 2009.)

Jurors watched him provide an in-court demonstration of why his $3,000 or $5,000 bets are successful:

The makeshift craps table was “hot” for a while and [the defendant], who has spent the last eight months in prison, appeared happy to be reunited with the dice. He cradled them in his hand and shook them as [his] defense attorney ... walked him through an explanation of his strategy, which involves combinations of numbers that “go with” other numbers. If the shooter rolls a four, for example, [the defendant] recommends betting on 4, 6, 9, 10, 2, 3 or 11 to come up next.

The defendant described his success:

"The strategy works if you manage your money right and cash out when you're winning,” he said. He said his best run occurred in 2002, when he won $73,000 at Foxwoods. Other times, he won big and then lost big, he said.

Discussion

According to Wikipedia [3], in the game of craps, a shooter rolls a pair of dice until a "point" of 4, 5, 6, 8, 9, or 10 is established. He/she continues rolling the dice until either the "point" is rolled again (a win) or a 7 is rolled (a loss).

1. If the shooter rolls a 4, what is the probability of a 2, 3, 4, 6, 9, 10, or 11 coming up next on a pair of fair dice?

2. If the shooter rolls any other "point," what is the probability of a 2, 3, 4, 6, 9, 10, or 11 coming up next on a pair of fair dice?

3. Can you think of a reason why the defendant might have chosen this set of outcomes?

4. Can you identify any other set(s) of outcomes that have the same probability of coming up on any roll of a pair of fair dice?

5. It appears that bets are placed on an ultimate win or loss, not on the outcome of a particular roll. However, even if the defendant were correct in his reasoning and one wanted to bet on a particular outcome immediately following the roll of a 4, do you think that the defendant's strategy would be helpful in choosing an outcome to bet on?

Submitted by Margaret Cibes

Financial decision-making advice?

"Many Bought Shares High, Sold Low" [4], by Mary Pilon, The Wall Street Journal, May 18, 2009

The article describes several people who took large losses by selling their stocks around the time of the "March 9 12-year low," and before the more recent 26% "rebound."

One financial advisor told an investor about his "two-Ambien" test for decision-making. (An Ambien is a sleeping pill.)

If two Ambien can allow you to sleep, ... then it still might make sense to stay invested.

Many advisors tell clients not to divest during "downturns," because "buying high and selling low is a formula for awful returns." One advisor says that:

They know that in hindsight, it wasn't the best thing to do.... But it was what they had to do emotionally. Math and the mind don't always add up.

Submitted by Margaret Cibes

Infuse and Kuklo

J. Scott, editor of the Journal of Bone and Joint Surgery writes about a paper, “Recombinant human morphogenetic protein-2 for type grade III open segmental tibial fractures from combat injuries in Iraq” by Timothy Kuklo, et al, which appeared in the JBJS in August, 2008:

“The paper described the management of 138 Gustillo Type IIIB and C tibial fractures in soldiers injured in Iraq. It was a retrospective study with some randomization of these patients into two groups, one of which received bone morphogenetic protein-2 (rhBMP-2) as part of the management and the other did not. The authors reported a significantly higher union rate in the group treated with rhBMP-2 (92% vs 76%). There was also a higher rate of further surgery required in the patients who did not receive BMP.”

For the non-specialist, rhBMP-2 is sold commercially as “Infuse” and is marketed by Medtronic. And, the paper “clearly seemed to represent a major contribution to the treatment of these severe complicated fractures which are difficult to manage and usually require several surgical procedures, careful wound management and extensive rehabilitation.”

However, you won’t be able to read the Kuklo paper to find out what is meant by “a significantly higher union rate in the group treated with rhBMP-2 (92% vs 76%)” because the paper has been withdrawn by Scott and the JBJS. The reasons may be found in Wilson and Meier’s reporting in The New York Times. The headline is “Doctor Falsified Study on Injured G.I.’s, Army says.” Not only did he “not obtain the Army’s required permission to conduct the study” but also “Army investigators found that Dr. Kuklo forged the signatures of four Walter Reed doctors on the article before submitting it last year to a British medical journal, falsely claiming them as co-authors.” In addition, “the total number of patients Dr. Kuklo reported as having been treated for extensive lower leg wounds at Walter Reed during the study period—138 soldiers—was greater than the number for which the hospital could find records.”

Further, according to Scott, the forgery came to light because “Shortly after the paper was published we received correspondence from one of the persons identified as a co-author indicating that the alleged co-authors had not seen the manuscript prior to publication and they had not signed the letter of transmittal. It was further disclosed that much of the paper was essentially false.”

Discussion

1. The NYT article states “During the six-month period ending last October, sales of Medtronic’s bioengineered products, principally Infuse, reached $419 million, according to a company filing.” A Medtronic spokesperson “confirmed that Dr. Kuklo was a paid consultant to the company and that the company financially supported some of his research at Walter Reed.”

A subsequent NYT article reveals that “Army doctors can accept money to consult for medical product companies if they are given approval. Military officials said Wednesday that they had not found records that Dr. Kuklo sought or received such permission.” A “consultants list shows that Medtronic paid about $943,000 from 2003 to 2008 to 22 doctors for consulting specifically about Infuse.” Kuklo’s name is not on that list “because he had a general consulting contract with Medtronic, rather than one specific to Infuse.”

2. Kuklo is now an associate professor at Washington University. Here we have this commentary from someone at St. Xavier University whose father “served on the [Washington University] medical school faculty for over 30 years”: “Dr Dan Riew, the Mildred B. Simon Distinguished Professor and Chief of Cervical Spine Surgery in the Department of Orthopedic Surgery, alleged that his colleague’s forging signatures of four phantom co-authors may have been subsequent to oral authorisation. Astonishingly, Dr Riew claimed that when a researcher is without a fax machine or is abroad, the forging of signatures may be the only alternative.” From the St. Louis Post-Dispatch is the headline, “WU colleagues say surgeon accused of fraud is honest, hardworking.” The university’s chief of orthopedic spine surgery writes, “The claims are largely false. Dr. Kuklo is a very honest investigator.”

3. Forging co-authors, although bizarre, is not unique to this instance. Another example may be found a short time ago in Chance News. An interesting book to read is Why Smart People do Dumb Things by Feinberg and Tarrant. The section on Cyril Burt focuses on his putative co-authors, sometimes known as “the ladies,” Ms. Howard and Ms. Conway. Then use Google to find assertions that Howard and Conway did exist and were not inventions of Burt.

4. Speculate on what motivates highly respected medical and scientific researchers to commit fraud.

Submitted by Paul Alper

Financial engineering as pseudo-science

"The Death of Kings," [5], by Nick Paumgarten, The New Yorker, May 18, 2009
Note: Readers may only be able to access the abstract online, without subscription.

In this 16-page article, author Paumgarten discusses the current economic crisis and presents some possible causes that were identified by several financial analysts he interviewed.

According to bond salesman Colin Negrych, "What Wall Street offers is the continual rationalization that ever-increasing indebtedness is sustainable .... It concocts believable, defensible arguments for the prices that they think things ought to be. Financial engineering fills the gap between people's desires and their wherewithal. So what you have is optimism buttressed by pseudoscience and statistical legerdemain."

Financial writer Paumgarten reports:

By that time, modern finance theory – the notion, borne of some elegant mid-century mathematics, that one could use models to value contingencies – had taken root in the world of financial practice. It gradually obscured "the sheer brute fact that the results of human activity cannot be anticipated," as the economist Frank Knight wrote in 1921. Yet anticipate it people did, or tried to, on trading desks and conference calls, amid what [David] Beim called "a rise in complexity." Mathematicians and physicists, cut loose by the decline of the space program, gravitated to Wall Street and began devising ways to measure, price, and package risk. It was a kind of decentralized Manhattan Project."....

"Financial engineering tapped into a strain in the investor's mind by replacing uncertainty with the appearance of certainty," [Simon] Mikhailovich said. Certainty came in a guise of inscrutability; the products designed to reassure also happened to befuddle. Many of the people responsible for evaluating the engineering considered their mystification to be further proof of its brilliance. They were, like Bernie Madoff's investors, comforted by their own ignorance. ....

Negrych quoted a line from a friend: "Wall Street takes your money and their experience and turns it into their money and your experience."

Submitted by Margaret Cibes

SAT Coaching

"SAT Coaching Found to Boost Scores – Barely" [6], by John Hechinger, The Wall Street Journal, May 20, 2009

In a May 2009 report, authored by Derek Briggs, chairman of the Research and Methodology Department at the University of Colorado in Boulder, the National Association for College Admission Counseling [7]

criticizes common test-prep-industry marketing practices, including promises of big score gains with no hard data to back up such claims. The report also finds fault with the frequent use of mock SAT tests because they can be devised to inflate score gains when students take the actual SAT."

(The Association's analysis was based on prep courses for the pre-Writing-Section version of the SAT.)

Several students allege that their prep company's practice tests were more difficult than the actual test, which could account for score gains from practice test to actual test. A company's response to the students is that those students were "outliers," and that "surveys of students at [their high school] generally show high satisfaction with the test-coaching company's results."

The report concludes that "on average, prep courses yield only a modest benefit, 'contrary to claims made by many test-preparation providers.'" It claims that "SAT coaching resulted in about 30 points in score improvement on the SAT, out of a possible 1600, and less than one point out of a possible 36 on the ACT ...."

According to the article's author, the College Board is "critical of colleges that select applicants based on small score differences that aren't statistically significant."

The article includes a table of claims/guarantees provided by seven test-preparation companies.[8]
In a blog [9], Matthew Fraser, of Education Unlimited, responds to the author of the report.

Discussion

1. Did you realize that some test-prep companies may be claiming student gains in SAT scores on the basis of comparing actual SAT scores to practice test scores, not to previous actual SAT scores? Do you think that companies should divulge (or be required to divulge) more information about their claims of improved scores?

2. While a 30-point difference (out of 1600 points) in two students' SAT scores might be highly significant to a college admissions officer, under what condition(s) might a 30-point difference be statistically significant?

Submitted by Margaret Cibes

SAT after 41 years

"High-School Senior: I Took the SAT Again After 41 Years" [10]
by Sue Shellenbarger, The Wall Street Journal, May 27, 2009

This columnist describes her experience re-taking the SAT at age 57, after a challenge from her teenage son.
While she had not studied for the test in 1967, this year she studied from the College Board's official guide. Her scores -- in the order 1967 adjusted score, 2009 practice test score, 2009 score -- were

Verbal/Critical Reading 800, 690-790, 800

Math 640, 430-490, 600

Essay N/A, N/A, 10 out of 12.

In a blog, the columnist describes how to update scores on any SAT taken before April 1995 [11]:

[A]dd about 70 to 80 points to your old verbal score, up to a maximum of 800 .... My 1967 verbal score was 740, so it became 800 after adjustment. If you scored between 200 and 500 [in math] originally, add 30 points. If you scored above 500, add 20 points. If you scored above 550, add 10 points. If you topped 600 in math on the old SAT, keep your original score, rounded to the nearest 10-point interval. For example, my 1967 math score was 636, so the College Board recommended I round it to 640. Scores today are reported only in 10-point intervals.

Submitted by Margaret Cibes

Multiple-choice test screens for potential gang members

"A New Approach to Gang Violence Includes a Multiple-Choice Test" [12]
by Nicholas Casey, The Wall Street Journal, May 20, 2009
Having experienced nine burglaries at his home and studied gang activity for four decades, retired social psychologist Malcolm Klein has joined with USC colleagues to design a multiple-choice test [13] that "they hope will empirically identify which children are headed toward a life on the street." The City of Los Angeles plans to use the 70-question test to screen 10-15 year olds for signs of potential gang membership. The children will not be told the purpose of the test.
According to the author of the article,

In Los Angeles, Dr. Klein's theories are appealing to policy makers eager to stretch limited resources. This year, the test is being given to children for the first time, and officials say they will use the results to determine whether some of the city's $24 million annual budget for gang prevention is being spent on children who aren't at high risk.

The emphasis on data is part of what policy makers have been calling an "epidemiological" strategy, drawing analogies between the spread of crime and disease. The focus is shifted from treating "symptoms" of gang activity -- violent crime, for example -- to prevention efforts that will stem proliferation.

So far, 958 children who live in active gang areas have taken the test; of that group, about one-third have been identified as potential future gang members and will be enrolled in prevention programs. But city officials won't know for several years whether the test failed to pick out children who went on to join a gang.

An LA-detective blogger [14] expresses his concern about the potential for test results to stigmatize young people.

Discussion

The test [15] is actually a "Youth Services Eligibility Interview," with more than 70 questions. It is in a multiple-choice format in the sense that an interviewee chooses an answer from a rating scale of "1" to "5" or from the pair "Yes" or "No." Some questions are open-ended.
1. Students are told at the beginning of the interview:

The reason for this survey is to find youth who might want to participate in a new city program. The program was designed to help young people develop successfully and keep them out of gangs. This survey will let us know if our free program will be helpful for you. This is not a test, and there are no right or wrong answers. All you have to do is answer honestly. The answers you give will stay private.

Do you agree that the children will not be told the purpose of the test?
2. What would a "false positive test" mean for a student? How about a "false negative test"? Can you suggest some repercussions of a false test, for better or worse?

Submitted by Margaret Cibes

Predicting winners and losers in Supreme Court cases

When the Justices Ask Questions, Be Prepared to Lose the Case, by Adam Liptak, The New York Times, May 25, 2009.

A second year law student at Georgetown unlocked a secret to predicting who would prevail in court cases argued before the U.S. Supreme Court. Just look at what happens during oral arguments for the case.

"'The bottom line, as simple as it sounds,' said the student, Sarah Levien Shullman, who is now a litigation associate at a law firm in Florida, 'is that the party that gets the most questions is likely to lose.'"

This was a very small study (10 small), but it inspired a replication by a future chief justice.

"Chief Justice Roberts heard about Ms. Shullman’s study while he was a federal appeals court judge, and he decided to test its conclusion for himself. So he picked 14 cases each from the terms that started in October 1980 and October 2003, and he started counting. 'The most-asked-question "rule" predicted the winner — or more accurately, the loser — in 24 of those 28 cases, an 86 percent prediction rate,' he told the Supreme Court Historical Society in 2004."

These small studies were replicated in a comprehensive study that looked at 2,000 oral arguments.

"If the two sides receive the same number of questions, the likelihood of reversal is 64 percent, which is in line with the usual probabilities; the court reverses more often than it affirms. But if the side seeking reversal gets 50 more questions than its adversary, the likelihood of a victory drops to 39 percent. And if that side manages to get the maximum number of extra questions in the study, which was 94, the likelihood of winning drops to 18 percent."

The article continues with a discussion of the predictive power of particular words. Pleasant words like approve, confidence, and guidance do not seem to influence the prediction but unpleasant words like "abusing," "failed," and "hostile" that are directed at a particular party will decrease the chances they will prevail.

Submitted by Steve Simon

Stereotype threat may affect senior citizens

"How Stereotypes Defeat the Stereotyped" [16], by John Cloud, TIME, online version of May 9, 2009

A study in Experimental Aging Research shows that old people may be subject to the effects of "stereotype threat," which refers to a situation in which "some members of stigmatized groups, when faced with stressful situations, expect themselves to do worse – a prophecy that fulfills itself."

The study is said to describe an experiment in which psychologists at North Carolina State University in Raleigh recruited 103 volunteers, ages 60 to 82, to perform simple arithmetic and recall tests. The researchers told about half of the participants that the purpose of the tests was "to examine aging effects on memory" and were asked to write down their ages before beginning the tests. A control group was told that the tests had been constructed to correct for age-related bias, but these participants were not asked to write down their ages. According to the article's author, members of the "treated" group performed significantly worse on the memory tests than the control group, and they even performed worse than they had on a pre-experiment screening test.

In the print version of this article (June 1, 2009, issue), the author cites Baruch College psychologist Catherine Good, who advises preparers of standardized tests to move questions about personal demographics to the ends of tests, in order to mitigate the effect of "stereotype threat" on student test takers. Professor Good, along with Steve Stroessner and Lauren Webster, runs the website Reducing Stereotype Threat [17]. The author also credits Stanford University social psychologist Claude Steele [18] with the original use of the term "stereotype threat."

Submitted by Margaret Cibes

Making corporate directors accountable to investors

"The Financial Page: Board Stiff" [19], by James Surowiecki, The New Yorker, June 1, 2009

The article discusses some changes in the methods of selecting corporate directors, changes which have been implemented to provide more accountability to investors.

Among other efforts, companies are trying to appoint more directors who are "independent" and from a wider pool, in response to new regulations and investor pressure. However, two research projects indicate that these steps may not be as effective as thought.

The academics Sanjai Bhagat [University of Colorado at Boulder] and Bernard Black [University of Texas at Austin] ... have found no evidence that simply appointing more independent directors improves corporate performance. .... James Westphal, a business professor at the University of Michigan, has found that professionally diverse boards are actually less likely to challenge the C.E.O.

The article's author identifies two issues that may be causing the problems: (1) C.E.O.'s, not shareholders, still nominate most directors, and (2) "independent" directors do not always have the appropriate experience to challenge management. Currently the S.E.C. is considering a proposal for large investors, instead of C.E.O.'s, to nominate candidates. Going a step further,

"[C]orporate-law scholars Ronald Gilson and Reinier Kraakman ... have proposed that big institutional investors create a cadre of full-time directors, people whose only job would be to sit on corporate boards and look after shareholder value."

Submitted by Margaret Cibes

The financial crisis: risk, prisoner's dilemma, and pschology

"Outsmarted" [20], by John Lanchester, The New Yorker, June 1, 2009

The article reviews three contemporary books about the origins of the financial crisis.

First is Gillian Tett's book, "Fool's Gold," which tells the story of how J.P. Morgan helped create the concept of the credit derivatives, or "credit-default-swap." Lanchester calls this book "gripping and indispensable."

The quality of the loans is critical, because those loans are the bank’s earning assets ... [and this is] the very core of what banking is. But the model of packaging plus securitization spurned the principle that a bank had to individually assess and monitor every loan. The mathematics of valuation models—horrendously complex equations to assess probabilities and correlations, cooked up in mad-scientist style by the firms’ “quants”—took on the burden of assessing statistical risk. The idea that a banker looks a borrower in the eye and takes a view on whether he can trust him came to seem laughably nineteenth-century. As for the risks? Well, as Lawrence Summers said when he was Deputy Secretary of the Treasury, “The parties to these kinds of contract are largely sophisticated financial institutions that would appear to be eminently capable of protecting themselves from fraud and counterparty insolvencies.”

Second is U.S. Court of Appeals Judge Richard A. Posner's book, "A Failure of Capitalism," a discussion of factors behind the crisis.

“A Failure of Capitalism” argues that the risks taken by the banks were rational, for two main reasons. First, it’s only with the benefit of hindsight that we can know that a bubble in prices was taking place. Bankers had to assign a probability to the prospect that there was a bubble, and, second, to the prospect that, if there was a bubble and it burst, house prices would fall by twenty per cent or more—this being the decline that precipitated the general crisis of bank insolvency. Now, suppose that the risk of both things happening was one per cent. Whether an event with that likelihood is worth worrying about depends on what its consequences will be. From the larger point of view, the consequences included systemic meltdown; but Posner invites us to focus our attention on what they looked like for individual bankers. They had strong incentives for taking the maximum amount of risks in their lending, since risks are correlated with rewards, and the bankers were so well paid that they didn’t really have to worry about being laid off. .... Besides, if a bank avoids these risks, and its competitors don’t and therefore make more money during the boom, the cautious bank risks going out of business anyway, because its clients will walk away. .... The conclusion: “Risky behavior of the sort I have been describing was individually rational during the bubble. But it was collectively irrational.”

Third is Robert J. Shiller and George A. Akerlof's book, "Animal Spirits: How Human Psychology Drives the Economy, and Why It Matters for Global Capitalism," a book with "barely a page ... without a fascinating fact or insight." The title is taken from a phrase in John Maynard Keynes:

We can't know the future, and therefore our inclination to act, to do things "can only be taken as a result of animal spirits – of a spontaneous urge to action rather than inaction."

The article's author cites Shiller and Akerlof, as he concludes:

“What allows capitalism to function is the regulations," they write. This should be an enduring lesson of the crisis—an understanding that the rules governing the operating of markets were not handed down on stone tablets but are made by men, and are in constant need of revision, supervision, and active, imaginative enforcement.

Submitted by Margaret Cibes

Careless media reporting of a Facebook study

"Facebook and Procrastination" [21], by Earle Holland, Columbia Journalism Review, May 8, 2009

This article describes the unfortunate media frenzy over an Ohio State University graduate student's preliminary research results about an association between students' uses of Facebook and their GPAs. It describes what was "a good lesson in the inherent risks of reporters’ cavalierly covering the social sciences, as well as the risks that young researchers can face in dealing with the news media."
According to the article's author, Aryn Karpinski, the study's co-author, stipulated that the research was a pilot study and that, while the authors did find a correlation between Facebook use and GPA, they did not find a causal relationship. She also acknowledged that more research was needed. In a March press release reporting the study, the University asked that reports of the study not be published until the study had been presented at an April 16 educational research conference.

Nevertheless the report appeared before the conference, in the Sunday Times of London, which stated:

Research finds the website [Facebook] is damaging students’ academic performance. … Facebook users … are more likely to perform poorly in exams, according to new research. … The majority of students who use Facebook every day are underachieving by as much as an entire grade compared with those who shun the site.

This incorrect conclusion went on to become the basis of "hundreds of news stories from media around the world." Nor did any of the graduate student author's interviewers alter their reports after speaking with her. And then a second round of stories ran in USA TODAY and other media.

Some writers, like Carl Bialik, The Numbers Guy in The Wall Street Journal, discussed the faulty reporting.[22] And John Timmer, of Ars Technica , also discussed the issue.[23]

For the original data summary and blogs, including a response by Aryn Karpinski, see here [24].

Submitted by Margaret Cibes

Strategy for basketball et al.

"How David Beats Goliath: When underdogs break the rules" [25], by Malcolm Gladwell, The New Yorker, May 11, 2009
This article discusses how/why weaker individuals/groups can overcome stronger ones by applying non-conventional strategies. Examples include David vs. Goliath, T. E. Lawrence vs. Ottoman Army, Fordham vs. UMass (1971), and UKentucky vs. LSU. (1996).
Vivek Ranadivé, MIT grad, software-company founder, and Mumbai native, agreed to coach his daughter's basketball games and

was puzzled by the way Americans played basketball. .... A basketball court was ninety-four feet long. But most of the time a team defended only about twenty-four feet of that, conceding the other seventy feet. Occasionally, teams would play a full-court press ... for only a few minutes at a time.

Ranadivé also noticed that the actual basketball playing took place in about 24 minutes, not the official time of 48 minutes, due to "preliminaries and formalities."
His all-girl team played all 94 feet of the court with a full-court press, "which is legs, not arms. It supplants ability with effort."
His methods were not popular with critics, especially opposing coaches who accused his team of unfair play.

[T]he puzzle ... is that [the full-court press] has never become popular. People look at upsets ... and call them flukes. Ranadivé readily admitted that all an opposing team had to do to [win] was press back: the girls were not good enough to handle their own medicine. Playing insurgent basketball did not guarantee victory. It was simply the best chance an underdog had of beating Goliath.

The article's author goes on to examine a number of David-and-Goliath examples from history.

The political scientist Ivan Arreguín-Toft recently looked at every war fought in the past two hundred years between strong and weak combatants [202 cases]. The Goliaths, he found, won in 71.5 per cent of the cases. [He found that] and even in ... lopsided contests the underdog won almost a third of the time. .... What happened, .... when the underdogs ... chose an unconventional strategy [50 cases]? He went back and re-analyzed his data. In those cases, David’s winning percentage went from 28.5 to 63.6. When underdogs choose not to play by Goliath’s rules, they win, [he] concluded, “even when everything we think we know about power says they shouldn’t.”

He also discusses Randivé's related professional interest in applications of non-traditional methods with respect to real-time vs. batch processing.
Submitted by Margaret Cibes

Longitudinal, qualitative study of Harvard men

"What Makes Us Happy? [26], by Joshua Wolf Shenk, The Atlantic, June 2009

This article discusses the work of psychiatrist George Vaillant, who has run the 72 year old Harvard Study of Adult Development for the past 42 years. Participants are 268 Harvard men, primarily from the classes of '42, '43, and '44, who have been administered extensive medical and psychological tests and responded to surveys and interviews over their lifetimes since college. The study is called the Grant Study, after W. T. Grant, department-store magnate and original funder. The article's author provides interesting excerpts from individual reports, based on his access to the study's files.

The study's creator, Arlie Bock, said in 1938 that the study would “attempt to analyze the forces that have produced normal young men.” He defined normal as “that combination of sentiments and physiological factors which in toto is commonly interpreted as successful living.”

Who are these anonymous folks, some of whom have self-identified?

Four members of the sample ran for the U.S. Senate [and one was the renowned journalist Ben Bradlee]. One served in a presidential Cabinet, and one was president [JFK]. There was a best-selling novelist [not Norman Mailer]. But hidden amid the shimmering successes were darker hues. As early as 1948, 20 members of the group displayed severe psychiatric difficulties. By age 50, almost a third of the men had at one time or another met Vaillant’s criteria for mental illness. .... “They were normal when I picked them,” [Bock] told Vaillant in the 1960s. “It must have been the psychiatrists who screwed them up.”

According to the article's author, "[I]t turned out that the lives were too big, too weird, too full of subtleties and contradictions to fit any easy conception of “successful living.”

Vaillant has his own goal with respect to the Grant Study: "His central question is not how much or how little trouble these men met, but rather precisely how—and to what effect—they responded to that trouble."

He "identified seven major factors that predict healthy aging, both physically and psychologically [– e]mploying mature adaptations ..., education, stable marriage, not smoking, not abusing alcohol, some exercise, and healthy weight."

Of the 106 Harvard men who had five or six of these factors in their favor at age 50, half ended up at 80 as what Vaillant called “happy-well” and only 7.5 percent as “sad-sick.” Meanwhile, of the men who had three or fewer of the health factors at age 50, none ended up “happy-well” at 80. Even if they had been in adequate physical shape at 50, the men who had three or fewer protective factors were three times as likely to be dead at 80 as those with four or more factors.

He found that some factors did not matter to happiness in old age. These included "cholesterol levels at age 50," "social ease," "childhood temperament," and "physical health."

The men’s relationships at age 47 ... predicted late-life adjustment better than any other variable, except defenses. Good sibling relationships seem especially powerful: 93 percent of the men who were thriving at age 65 had been close to a brother or sister when younger.

The Harvard data also illustrate the phenomenon of distortion.

In 1946, for example, 34 percent of the Grant Study men who had served in World War II reported having come under enemy fire, and 25 percent said they had killed an enemy. In 1988, the first number climbed to 40 percent—and the second fell to about 14 percent. “As is well known," Vaillant concluded, "with the passage of years, old wars become more adventurous and less dangerous.”

A second study, the Glueck study, started in 1939 by Sheldon and Eleanor Glueck and taken over by Vaillant in the 1970s, focuses on "delinquent" and "nondelinquent" boys and is now part of the Harvard Study of Adult Development. This study has found that "industriousness in childhood ... predicted adult mental health better than any other factor, including family cohesion and warm maternal relationships."

But Vaillant has largely played down the distinctions among the samples. For example, while he allows that, in mortality rates, the inner-city men at age 68 to 70 resembled the ... Harvard cohorts at 78 to 80, he says that most of the difference can be explained by less education, more obesity, and greater abuse of alcohol and cigarettes. “When these four variables were controlled,” he writes, “their much lower parental social class, IQ, and current income were not important.” But of course those are awfully significant variables to “control.” Vaillant points out that at age 70, the inner-city men who graduated from college were just as healthy as the Harvard men. But only 29 Glueck men did finish college—about 6 percent of the sample.

Submitted by Margaret Cibes

CEOs: talented or just lucky?

Do CEOs Matter? [27], by Harris Collingwood, The Atlantic, June 2009
This article discusses whether a CEO has any special impact on the success of a business. The article was prompted by the current activity in the price of Apple stock in view of concerns about CEO Steven Jobs' health.

The debate over the centrality of the CEO began in earnest in the 1930s, with the work of Chester Barnard, a onetime president of New Jersey Bell. Barnard was one of the first scholars to recognize that the corporation was, first and foremost, a social organization, and that social phenomena such as groupthink and the development of rival factions influenced a corporation’s actions at least as much as coldly rational analysis and deliberation.

In a 1972 article, “Leadership and Organizational Performance: A Study of Large Corporations,” Stanley Lieberson and James O’Connor argued against overestimating the effect of a CEO on a company's performance:

[It] was seldom decisive in a company’s performance. They had the numbers to back up this view. Working with a database of 167 companies, they teased out the effects that various factors had on corporate profitability, from the competitive state of the industry to the size and structure of an individual company to the CEO’s managerial decisions. “Industry effects,” such as the amount of available capital and the stability of the market, accounted for almost 30 percent of the variance in corporate profits. “Company effects,” such as the firm’s historical place in the corporate pecking order, explained about 23 percent. “CEO effects” explained just 14.5 percent. And even this impact should be viewed skeptically: it unavoidably bundles CEO actions that were genuinely smart and skillful with those that were merely lucky.

Other scholars have concluded that

[E]xternal forces influence corporate performance far more than CEOs do. Indeed, more-recent studies have tended to find a smaller CEO effect than Lieberson and O’Connor did—ranging from 4.5 percent to 12.8 percent of profit variance.

The article's author cites Jeffrey Immelt, successor to Jack Welch as CEO of General Electric. Immelt recently stated that

[I]n the 1990s, “anyone could have run GE and done well.” [H]e added, “Not only could anyone have run GE in the 1990s, [a] dog could have run GE. A German shepherd could have run GE.” Welch, to his credit, more or less agreed with this assessment.

And three Harvard scholars argue in a recent paper that

[T]he right question is, When does leadership matter? Using advanced statistical techniques that go by a wonderfully CSI-style name, “variance decomposition analysis,” the authors examine 531 companies in 42 industries and isolate leadership effects from other determinants of corporate performance. They conclude that leadership matters sometimes. It doesn’t make much difference at electrical-utility companies, which are so constrained by government regulations and the cost of fuel that there’s very little room for the CEO to exercise any discretion.

Submitted by Margaret Cibes

Historical chart of GM stock prices

"From Blue Chip to Near Bankrupt" [28], The Wall Street Journal, June 1, 2009

See charts for General Motors stock of (a)"Daily closing share price since 1926, adjusted for splits and spinoffs," (b) "Monthly closes, inflation-adjusted," and (c) "Monthly closes shown on a logarithmic scale." [29]
Submitted by Margaret Cibes

Governor proposes state Keno game

"Rell Gives Keno Address: www.youlosebig.com" [30]
by Colin McEnroe, The Hartford Courant, May 31, 2009

The governor of Connecticut recently proposed a state-run Keno operation as a way of raising an additional 60 million dollars.

According to the columnist, "if you play Keno a lot, you will be more likely to get back about 60 dollars of every 100 dollars you put in."

The basic idea is that 20 numbers of a possible 80 are picked at random. You have the opportunity, beforehand, to guess as many as 10 of them. .... States with Keno machines typically run the games every five minutes. In Michigan, for example, Club Keno drawings take place every five minutes seven days a week from 6:05 a.m. until 1:45 a.m.

The columnist also states that "Many of the machines are in bars ..., where patrons can consume nourishing and judgment-improving beverages while they decide how much money to lose. A weekly Powerball ticket is to Keno games as a large cup of coffee is to crack cocaine."

Submitted by Margaret Cibes

The U.K. stats police

"This U.K. Sheriff Cites Officials for Serious Statistical Violations" [31]
by Carl Bialik (The Numbers Guy), The Wall Street Journal, June 3, 2009

The U.K. instituted a Statistics Authority in 2007 with the "budget, authority and independence to question other government agencies on the numbers they release to the public."
Poll results have suggested the public's distrust of government statistics:

Among 1,708 adults who answered a 2004 survey by the Office for National Statistics ... , just one in six agreed with the statement, "Official figures are produced without political interference." .... A poll conducted three years later found that just 33% of U.K. respondents said they tend to trust official statistics, the lowest rate among the 27 member states of the European Union.

Discussion

1. Did the U.K. public appear to have become more trusting of government statistics, or less trusting, over the period 2004-2007?

2. What unstated assumption(s) do you need to make in order to be able to interpret the margin of error given with the table?

Submitted by Margaret Cibes

Alcohol's Good for You?

From istostat

There is an article in today's New York Times Health section about moderate drinking and health benefits. It discusses the difficulty in determining the causation that underlies this and other correlations of health variables, and could be interesting for discussion in a conceptually based intro stats course.

http://www.nytimes.com/2009/06/16/health/16alco.html?em "Alcohol's Good for You? Some Scientists Doubt It" by Roni Caryn Rabin, New York Times, 6/16/09

Annette Gourgey CUNY