In the Dallas Morning news there are many other articles using and abusing
statistics to lesser and greater degrees. In all cases sources are not
quoted and footnoting is not a feature. To get more accurate reporting
about health or science in newspapers the basic aim of newspapers would
need to change. Sensation sells and advertisers give business to newspapers
that sell. Change this.

There are so many articles that have counter-intuitive statistics, that
don't quote sources or substantiate reporting with interviews. Take this
example from the same newspaper.

Girth of a nation:Doctors warn of health crisis as obesity gains on Americans
http://www.dallasnews.com/specials/obesity/

>Almost 10 years ago, when the number of overweight Americans hovered
>around 25 percent of the population, public-health experts set out to
lower that
>statistic by the century's end. Instead, the population tightened its grip on
>cheeseburgers and remote controls. Obesity rates soared.

>Now, about half the population is at least somewhat overweight. And nearly
>one-quarter of adult Americans are considered obese, carrying an extra 30
>pounds or more.

...many of the same mistakes, over exaggeration, lack of sources, quotes
from experts... Yet why take the cigarette article as an example of bad
science when there are other articles with similar exaggerations,
misleading evidence. If you click on the Health button in the Dallas
Morning Times every article is an example of this type of reporting and
most newspapers do the same. Is "objective scientific truth" the aim of the
articles?

Statistical sources are always questionable. What is being measured? How
was it measured? What kind of sample of the entire population was tested?
What were the assumptions underlying the research? The fact that various
results are quoted in this article giving different statistical outcomes is
of concern showing use of a variety of sources, not quoted in a newspaper
article.

Ron Roizen, your statement about alcohol paradigms in your essay, "How Does
the Nation's 'Alcohol Problem' Change From Era to Era?  Stalking the Social
Logic of Problem-Definition Transformations Since Repeal" asks the
> ...Sixty-Four-Dollar Historical Question ... Should we
>regard the current drift away from the de-politicizing alcoholism paradigm
and
>toward the new, re-politicizing public health approach as historical
progress or
> retrogression?
http://www.roizen.com/ron/postrepeal.htm

The same question arises around smoking, freedom to smoke, how smoking has
been viewed, and how it is viewed today in various parts of the world
compared to the past. And for this reason the link in the article to the
tobacco industry is not a diversion but part of the whole.

At the moment there is a political fight going on between cigarette
manufacturers, doctors and the anti-smoking lobby. It is part of our
present paradigm and in the west the anti-smoking lobby has improved it's
stake in the argument.

This article chosen as an example of poor reporting is a glimpse of part of
the battle being aired in the public domain. It's sources and statistics
are probably as good and as bad as most newspaper reporting. I echo Steve
and Roy's call for better statistics and reporting. The demand is important.

As a social and political animal I do not like seeing tobacco companies
peddle their wares in third world countries where there is less overt
criticism of the tobacco industry. I hope that better research will be done
and that statistics will be used appropriately. My biases are open. I
support the basis of the newspaper report, to limit (by uncoersive means)
the use of cigarettes by people who are stressed and ill. I hope the
Harvard researchers will add weight to this outcome

Melanie Lazarow
History and Philosophy of Science Liaison Librarian
University of Melbourne

At 13:24 29/11/00 -0500, you wrote:
>A story today on cigarette sales to the mentally ill today seems
>scientifically problematic though it emanated from a study by Dr. Karen
>Lasser, et.al. She is a primary care physician at Cambridge Hospital and
>research fellow at Harvard Medical School.
>
>The story carried in the press lacks accurate data and expert analysis of
>the significance of the report. We picked up from a link in the Physician's
>News Digest of Southeasten Pennsylvania which carried a summary of the
>story. The Latter gave the Dallas News link below which ran the LA Times
>Story.Citation:  Rosie Mestel, 11/27/2000, "Cigarette sales raise concerns -
>Mentally ill make up big part of consumers" Los Angeles Times.
>http://www.dallasnews.com/science/health/224391_smoking_27dis..html
>
>The veracity of the data in the article of numbers of mentally ill and
>smokers in the nation is questionable. Moreover, there is nothing new in
>reporting that the mentally ill and drug addicts smoke; it is common
>knowledge in the field. The proportion of all smokers in the nation who are
>mentally ill is subject to question.<<From the survey, the scientists
>estimated that 44.3 percent of the cigarettes smoked in the United States
>are smoked by the mentally ill.>> How in the world the Harvard researchers
>leaps to the conclusion is not explained by the LA Times article. While it
>is plausible that the majority of smokers do at one time drink, use drugs,
>or attend psychological counseling, they are not necessary mentally ill.
>Surely addicts and dysfunctional abusers of substances and those who have
>been diagnosed with serious psychological/psychiatric disorders can be
>classified as mentally ill but in the absence of definitive data, it is
>difficult to accept the premise on face value that approximately 50% smokers
>nationally are mentally ill or that about 50% of those who purchase
>cigarettes are mentally ill. The statement is counter-intuitive and contrary
>to cursory observations at cigarette stands and counters in retail stores;
>though in some communities the proportion may be high. Moreover, the
>newspaper article contradicts itself: First it uses the 44.3 percent figure.
>then it uses 20%: << Smoking rates are high in such groups. In addition, in
>any given year, one in five U.S. adults is estimated to suffer from some
>form of mental illness. One in 20 suffers from a severe mental illness>>
>However, the lead of the articles states 50%: << Nearly half of all
>cigarettes purchased in the United States are smoked by people who suffer
>from mental illnesses, according to Harvard Medical School research.>>
>
>The story lacked interviews and substantive comments from the authors of the
>original study as well as comments from epidemiologists or medical/health
>statisticians. However the interviews with experts about the consequences
>and what should or could be done about smoking by the mentally ill are
>important and should be expanded upon. The LA Times article states << The
>authors of the journal's paper also suggest that the tobacco industry may
>have been aware that mentally ill people were a psychologically vulnerable
>segment of their market. The scientists quote a 1981 document by R.J.
>Reynolds Tobacco Co., which noted that some smokers smoked for "mood
>enhancement" and "positive stimulation" or because smoking "helps perk you
>up," "gain self-control," "calm down," and "cope with stress."   R.J.
>Reynolds Co. did not return several telephone calls seeking comment.>>
>
>The role of cigarette companies is a spin off story and diverts attention
>from the problems of accurate reporting and analysis outlined above.
>Moreover the conclusion of the researcher and her quote about the cigarette
>companies adds bias to the article. The statement incorrectly infers and
>leads one to conclude that smokers who seek to get perked up, "gain
>self-control," "calm down," and "cope with stress" are mentally ill.
>
>The LA Times is an excellent institution and it has the reputation of
>excellent reporting. So are broadcast media such as NBC and ABC news.
>However in the past we had the occasion to criticize the latter two for bad
>medical research reporting and now we criticize the LA Times cigarette
>story. Our criticism applies to others who ran the story without adequate
>analysis. Publishing an important medical or scientific story without
>investigating/analyzing it or by hyping/overplaying it can mislead readers
>and policymakers. The first step at correcting the problem is for reporters
>and editors being aware and vigilant that medical and scientific research
>news releases are just an invitation to investigate the story and to delve
>into the significance. There is a responsibility to take the time to get and
>report accurate data and other expert interpretations of the original report
>that allow readers to decide the significance on the merits and not on the
>hype.
>
>The rush to publish is not excusable for bad reporting and poor analysis of
>stories. Moreover, medical/health and scientific reporters and editors
>should be technically competent in the fields they cover. Generally
>speaking, a journalism degree and experience in local news reporting are not
>sufficient for the critically important mission of public communications of
>science and health/medicine affairs to the nation. We invite scholarly
>articles and expert commentary about the role of the news media in science
>and health policy.
>
>Best wishes,
>Stephen Miles Sacks, MPA, Ph.D., Editor and Publisher
>SCIPOLICY(tm)-The Journal of Science and Health Policy
>Box 504, Haverford, PA 19041
>Website: http://www.Scipolicy.net  E-mail: [log in to unmask]
>
>Volume 1, Number 1 of Scipolicy focuses on The Future of Large-Scale Health
>Systems and  includes 10l articles on health systems and the problems,
>changes in institutional ethics, and a case study of the University of
>Pennsylvania Health System. Subscriptions and orders for individual copies
>can be placed on line at http://www.Scipolicy.net.  Proposals and
>contributed articles are  welcome.
>