CHANCE News 4.11
(21 July 1995 to 18 August 1995)


Prepared by J. Laurie Snell, with help from William Peterson, Fuxing Hou, Jeanne Albert and Ma.Katrina Munoz Dy, as part of the CHANCE Course Project supported by the National Science Foundation. Please send comments and suggestions for articles
to jlsnell@dartmouth.edu

Back issues of Chance News and other materials for teaching a
CHANCE course are available from the Chance Web Data Base.


The published studies to date - and they're not great - suggest that the rupture rate and bleed rate (of silicon implants) is somewhere between 5 percent and 71 percent.
Food and Drug Commissioner Dr. DAVID KESSLER:

TED KOPPEL: I mean, that's a ridiculous range.

NOTE: If you would like to join a discussion group to share experiences using current chance events in class please send a note to jlsnell@dartmouth.edu.

NOTE: If you would like to join a discussion group to share experiences using current chance events in class please send a note to jlsnell@dartmouth.edu.
This issue of Chance News is not at complete as usual because of our attending meetings etc. We don't even have a Marilyn story! It's thanks to Jeanne Albert and Katrina Dy that we even have a Chance News this time.



Eunice Goldberg told us about the next article and provided her own enthusiastic account of the study being reported. This story make national news: For example Tom Brokaw talked about it on the evening news and the Washington Post used it as the basis of an editorial.

Double mystery.
The New Yorker, 7 August 1995, Pg. 45
Lawrence Wright

This article checks in on the current status of the nature vs. nurture debate by looking at twins, especially identical twins, who are reared apart. There is plenty of fascinating material here, including the case of Amy and Beth, a set of identical twins who have been studied since they were separated as infants in the nineteen-sixties. Their adoptive parents differed in many respects, from socio-economic status to acceptance of and later relationship with their adopted daughter. However, the two girls apparently remained very similar, if not practically identical, in many developmental and emotional ways. There are several other examples along these lines, but perhaps the most dramatic is the case of an identical pair of boys, separated at birth and subsequently both named Jim. The two Jims, when reunited 39 years later, found "that each been married twice, first to a woman named Linda and then to a woman named Betty. Jim Lewis had named his firstborn child James Alan, and Jim Springer had named his James Allen. In childhood, each twin had owned a dog named Toy. They had enjoyed family vacations on the same beach in Florida and had worked part time in law enforcement. They shared a taste for Miller Lite beer and Salem cigarettes." What is going on here?

Such evidence seems to weigh in heavily for the nature side of the debate. But of course, identical twins are not identical in all respects -- what accounts for the difference, genes or environment? The article now describes a pair of identical twins who differed in many significant areas, including height, strength, and intelligence. Perhaps most intriguing, the shorter and weaker of the pair suffered from a particularly savage form of muscular dystrophy, which would eventually kill her in her teens, while the stronger girl was completely spared. The mystery is solved by examining the genetic causes of the disease. According to the article, muscular dystrophy is carried on the maternal X chromosome. A female fertilized egg (or zygote) has two X chromosomes to choose from: one from the mother and one from the father. "If you took a normal girl and looked at a thousand cells, you'd expect five hundred of those cells to be using the mother's X and five hundred to be using the father's X -- just like tossing coins," geneticist John Burn says in the article. "Now and again, very rarely, a girl will toss heads in every cell...and switch off all her good copies."

Nurture does win out in some areas, though, including whether a twin develops a happy, more extroverted outlook on life. What about other social issues? Along these lines there are many other examples of research involving twins in the article, including several that are sure to be controversial: Is criminal behavior inherited? How about intelligence (The Bell Curve echoes)? In particular there is a section on some of the theories of David Lykken, author of the newly published book, "The Antisocial Personalities," who believes that, just as potential adopting parents must typically meet several pre-established criteria, "our [social] problems are not gong to be mitigated until we establish similar criteria for those who would produce children biologically."

1. How likely are the similarities between the two Jims? How might you measure this? Do you think that DNA controls what you name your child or the name of your spouse?

2. If choosing one parent's X chromosome over the other's is indeed "just like tossing coins", just how rare is it for a girl to "toss heads in every cell?" What in fact is meant here by "just like tossing coins?" Do you believe the statement made by Burns about the normal girl's X chromosomes?

3. If we are in fact victims of our genes, does that leave room for free will? The article closes by saying: "If it is true that our identical clone can sort through the world of opportunity and adversity and arrive at a similar place, then we may as well see that as a triumph of our genetic determination to become the person we ought to be." What does this mean? Do you agree?


Health benefits from soy protein.
The New York Times, 3 August 1995, A1 and A22
Natalie Angier

A new report in The New England Journal of Medicine states that soy protein significantly lowers cholesterol levels in people with moderately high to high cholesterol. Since elevated cholesterol levels sharply increase the risk of heart attacks and stroke, soy protein may prove a safe and painless way to battle America's no. 1 killer, cardiovascular disease. The New York Times reports that the higher a person's cholesterol level, the greater is the power of soy protein to reduce it. Dr. James W. Anderson from the University of Kentucky and his colleagues found that a diet of 47 grams of soy protein a day cuts cholesterol levels in a month by an average of 9.7 percent. People with extremely high cholesterol levels, those with over 300 mg per deciliter of blood, saw their cholesterol concentration plunge by 20 percent. Even better is the finding that soy protein fights the type of cholesterol that one wants to minimize, the so-called "bad cholesterol" or low-density lipoprotein levels, without affecting the amounts of "good cholesterol" or high-density lipoproteins. The most interesting part of this report is the way the study was carried out. The new study, according to the article, is not an original experiment but rather is a "meta-analysis." A "meta-analysis" is a review of all the studies done to date on the subject. After eliminating what they considered to be poorly designed or inappropriate studies, researchers then focus on the remaining ones to find their results. In this case, the researchers focused on 38 trials that included 730 subjects of both sexes, children and adults alike. "Meta- analyses" are often criticized for their subjectivity and for what they include and exclude. With this study's findings plus the fact that soy apparently has no detrimental side effects, one should expect a proliferation of soy products on the market. Scientists say that benefits can be seen with as little as 25 grams of soy protein daily, though 50 grams works better. Unfortunately, most soy products only have small amounts of soy protein. A soy burger might have 18 grams and a glass of soy milk only 8 grams.

1. Presumably each of the 38 trials used in the meta-analysis in some way measures the effect of eating soy protein on cholesterol levels. But what are the likely differences between the trials? How do you think such differences should be taken into account?

2. One way of performing a meta-analysis might be to treat each trial as part of a single larger trial and then "combine" all the results. What assumptions do you need to take this approach?


Study cites adult males for most teen-age births.
The New York Times, 2 August 1995, A10
Jennifer Steinhauer

A new national study by the Alan Guttmacher Institute has found that at least half of the babies born to teen-age girls are fathered by adults. The study implies that a startling number of teen-age girls are having sex with men who are breaking state laws on statutory rape.
The survey, the most comprehensive of its kind, draws from a survey of nearly 10,000 mothers between the ages of 15 and 49 who were interviewed from 1989 to 1991. Half of the fathers of babies born to mothers between ages 15 and 17 were 20 or older, and 20 percent of the fathers were 6 or more years older. In general, the younger the mother was, the greater the age difference between her and her partner. Such results, says the article, contradicts the idea that teen-age pregnancy is just limited to teenagers, but in fact adults bear a large portion of the responsibility.
Such trends have also been found on the state level in California and New York. Among California mothers ages 11 to 15, 51 percent of the fathers were adults.
The article stresses that such results highlight the fact that, while the whole political debate focuses on teen-age pregnancy as a result of planned, rational behavior, teen-age pregnancy in many cases results from sexual abuse and assault.


1. What would you like to know about the above survey to determine the reliability of its results?
2. What do you think is the significance of the reported findings?

Russian life expectancy puzzles Russia.
The New York Times, 2 August 1995, A1 and A6
Michael Specter

Russia has the lowest life expectancy rate of any developed country in the world. In the last four years, the Russian male life expectancy dropped from 64 to 57. Life expectancy for men born in Russia this year is lower than in India, Egypt, or Bolivia.

Deaths are now twice as common in Russia as births, and infant mortality has risen 15 percent in each of the last 2 years. If such a trend continues, nearly 50 percent of today's youth will not even make it to the retirement age of 55, for women, and 60, for men.

While death rates soar, the birth rate is lower in Russia than in any other country. In addition, there is a large increase in the number of babies born with birth defects. More than 10 percent of newborns have serious birth defects, and 50 percent of all school children suffer from chronic illnesses; these rates are growing every year.

What is even more baffling than the continued life expectancy decrease is that the cause behind the slide is a mystery. None of the experts seem to know what is causing it. Some say that epidemic rates of heart disease, cancer, and accidents account for much of the problem. Many researchers believe that the cause behind the declining birth rates and the rising numbers of birth defects lies in Soviet ecological abuse. For years, the Soviets held open testing of nuclear weapons, chemical plants spewed deadly materials into the country's most important rivers, and generations of factory workers and farmers were exposed to doses of dangerous pesticides and harmful chemicals.

Congenital abnormalities are at an all time high. People are beginning to examine the hypothesis that radiation produced by decades of nuclear irresponsibility is a major factor in the surge of illnesses and birth defects.

Others say that there are more obvious causes, like failure to vaccinate the population against preventable illnesses such as rubella, which can lead to the passing of genetic mutations from mother to child.


1. How can they be so sure that the explanations given cannot account for the drop in life expectancy?

2. The article states that the 1994 male life expectancy in Russia is 57.3 years. What does this mean? To whom does it apply? How could they have made this estimate? In particular, what data did they need?

2. Here is the way population books claim life expectancy is determined. Age dependent death rates for a specific census year, say 1990 are computed as follows: If at the middle of the year 1990 there were A people of age x alive and, during this year, B people of age x died, then the age specific death rate for age x is B/A. From these rates a life table is constructed. This life table considers a hypothetical population of 100,000 people born in 1990. It gives the number of these people that would be expected to be still alive at age x. How is this table constructed from the age specific rates? How, from such a table, could you find the probability that a women who is just 32 will live one more year? How could you find the probability that she will live 3 more years? How would you estimate the life expectancy for a women 32 years old?

Life expectancy then means the life expectancy computed from the life table for a child born in 1990. At the next census the birth rates will be recomputed for the year 2000 and a new life table provided given a new life expectancy. What assumptions are being made when you compute a life expectancy this way?

Editor's note: Surely insurance companies must use more sophisticated techniques than this to compute life expectancy. Does anyone know how they do it?

3. The article "Music charms may lengthen life (New York Times, 5 December 1978, Science Times) reports a study that found that famous male symphony conductors lived an average of 5 more years than the life expectancy computed in a year around the time that most of them were born. An astute reader, Douglas Carroll, (New York Times, 14 March, 1995, letter to the editor) observed that there was a subtle statistical flaw in the study. Can you spot it?


War on marijuana focuses on prenatal, other risks.
The Record, 20 July 1995, Pg. A17
Lauran Neergaard, AP

This article contains a number of claims about the supposed dangers of marijuana. White House drug policy coordinator Lee Brown is quoted as saying that marijuana is "a very dangerous drug that can well cause you to fight for your health and your very life in a hospital emergency room." He appears to be commenting on the results of a 1993 federal survey of 350 hospitals which found that 4,293 teens ages, 12 to 17, were treated in emergency rooms after using marijuana. This is more than twice the figure for heroin and cocaine combined. The article also mentions research on the effects of marijuana on children whose mothers smoked the drug during pregnancy. No effects were found up to age 3, but 4-year-olds who had been exposed exhibited slight impairments in memory and perception skills, the article says. There is no discussion of sustained effects, however. These and other issues surrounding the use of marijuana were part of a conference by the National Institute on Drug Abuse, at which a dozen protesters called marijuana a benign drug that eases pain and should be available for AIDS and cancer patients. The Clinton administration seems, however, to be on an opposite course: the article says that the above data "are part of a government campaign to change marijuana's image from that of a benign 1960's drug to an addictive killer that American children are using more and more often."


1. In the study on marijuana and pregnancy, critics have argued that the researchers did not take possible environmental factors into effect. For example, some of the impaired skills may be learned behaviors. How might this affect the validity of the study's results?

2. Suppose it is determined that THC, the active ingredient in marijuana, does not penetrate the placental barrier in pregnant women. (there is in fact no conclusive evidence either way.) How would you then interpret the above assertions?

3. What questions would you want to know about the hospital survey to be satisfied with the survey's results?

4. What are the possible interpretations of the statement from the article: "This is more than twice the figure for heroin and cocaine combined."


Minority policy -- gainers, losers.
The San Francisco Chronicle, 31 July 1995, Pg. B1
Jonathan Marshall

This article provides more fuel for the affirmative action debate by examining recent research on the overall benefits -- to the students themselves, their fellow classmates, and society as a whole -- of admitting under-qualified minority students to elite colleges and universities. "The tragic fact is that three-fourths of minority students are failing to graduate, often from colleges where they should never have been admitted in the first place", argues Thomas Sowell, a conservative economist at the Hoover Institution. He believes these same students would be in the top of their class at the "average American college." Others argue that a diverse student population benefits the entire college.

New research seems to support both views, at least in part. According to research by Kermit Daniel and two colleagues at the University of Pennsylvania's Wharton School, black students who attend colleges rated in the top 20 percent are likely to earn 20 percent higher wages than those who attend average schools, and whites who graduate from colleges where blacks comprise 8 to 17 percent of the student body are themselves likely to earn 15 percent higher wages than those who attend virtually all- white institutions. At the same time, research by two economists at Tufts University suggests that "as black students with comparable test scores and family income attend more selective institutions, their dropout rate tends to rise."

The article states that the typical black student, with a combined SAT score of between 701 and 850, had a 56 percent chance of graduating from a college where the average score was 900 but only a 39 percent chance where the average score is 1000. Moreover, those who dropped out earned, on average, $106 less than those who went to less selective schools and graduated. Also, the article gives somewhat comparable data for UC Berkeley: about 86 percent of students who entered in 1988 with scores between 1200 and 1300 (the average is 1225) graduated within 6 years, while 58 percent of students with scores between 700 and 800 graduated in that time. (Apparently the UC Berkeley data is for all students, not just blacks.)


1. According to the article, as recently as 1980 only 4 in 10 black students at Berkeley graduated within 6 years, while the graduation rate now is 62 percent. (The national average for whites is 56 percent.) This improvement is a result of increased programs like tutoring and counseling for disadvantaged kids, the article says. What other data would you want to see to test this conclusion? Is it reasonable to define "graduation rate" to mean within 6 years?

2. Do you think the article helps you decide if it is worthwhile to attend a school with an average SAT score that is above your own?

3. There is a chart included with the article, titled "Graduation rate by score", that displays the graduation rate data for Berkeley. Thus for each 100-point range of scores for the combined SAT (i.e. 700-799, 800-899, through 1500- 1600) a percentage is given, starting with 58 percent for the 700-799 range, rising to 88 percent for the 1300-1399 range, and then dropping again to 84 percent for the 1400-1499 range, and 79 percent for the 1500-1600 range. How would you explain the drop in graduation rate for the top SAT scores?


Mark Galit put us on the trail of the following story.
The bra and cancer: researcher says long-term use increases the risk, but physicians say it's not that simple.
The Cincinnati Enquirer, 31 May 1995, Pg. E01
Sue MacDonald

In this article we learn that wearing a bra for many hours a day, especially constrictive ones, may significantly increase a woman's chance of developing breast cancer. The results of a survey of 4700 women by medical anthropologist Sidney Ross Singer and his wife Soma Grismaijer are published in their book, "Dressed to Kill".

Among the findings listed in the article:

Women who wear a bra more than 12 hours a day, but not to sleep, increase their chances of developing breast cancer by 11 percent, compared to the general population.

Women who wear bras 24 hours a day have a six-fold greater incidence of breast cancer than women in the general population, a 113-fold greater incidence compared to women who wear their bras less than 12 hours a day, and a 125-fold increase compared to women who go bra-less.

While Singer concedes that follow-up studies are necessary, he says women shouldn't have to wait many years for the results, and should "just do their own experiments by quit wearing bras." His critics say that, although "many of the identified risk factors for breast cancer -- high-fat diet, obesity, family history, etc. -- account for only about 30 percent of the actual disease", the bra connection is too simple an explanation. Predictably, the bra industry goes one step further. "It's very difficult to take anything like this seriously," says Karen Bromley of the Intimate Apparel Council in New York City.


1. How do you think Singer and Grismaijer came up with the findings given above? What do they mean?

2. Do you think women should take Singer's advice and go bra- less? Why?

3. What is meant by the statement in the article that many of the risk factors for breast cancer account for only 30 percent of the disease?

4. In the article Singer says: "What's really broken is the notion that women aren't acceptable in their natural state." What do you think of this remark?

5. Suppose that the same women who wear bras 24 hours a day also tend to have larger than average breasts. How might this affect Singer's conclusions?

County heat deaths will Surpass 400; coroner's methods continue to be questioned.
Chicago Tribune, 19 July 1995, Page 1
Joel Kaplan & Cindy Schreuder with Peter Gorner and Anita Srikameswaran

As the total of heat-related deaths in Chicago reached 376, doubling earlier figures, more questions surfaced about the criteria used by the Cook County medical examiner to blame hot weather for the staggering toll. The number of deaths recorded so far from the heat wave is almost as many as the number that occurred (454 deaths) throughout the country in the summer of 1988, one of the hottest summers of the century.

Dr. Edmund R. Donoghue, the Cook County medical examiner, defends the large number of deaths listed as heat-related and says that his criteria are excellent. Here are the 3 determining factors:

1) Body temperature of the deceased is 105 degrees or above or
2) The deceased is found at home and the investigating officers report that the temperature is unusually high. or
3) The deceased is found decomposed in a home with an unusually high temperature.

In 1982, the Centers for Disease Control published a landmark study that seemed to foreshadow what happened in Chicago. Government scientists retrospectively examined the sickness and death associated with the July 1980 heat wave in St. Louis and Kansas City, Missouri. Heat-related illnesses and deaths were identified, by a review of death certificates and hospital, emergency room and medical examiners' records in the 2 cities and compared with similar statistics compiled from the previous 2 years, when there were no heat waves. Some findings of the study---

1)Deaths from all causes increased by 57 percent and 64 percent in St. Louis and Kansas City, respectively, as compared to only 10 percent in less populated areas of Missouri.

2)Access to air-conditioning was the most important lifesaver--those without it faced a 49.4 percent greater chance of dying.

3)The CDC blamed the difference in mortality rates between urban and less populated areas on a phenomenon called "heat island effect." (Heat retained is by buildings and pavement, and given off by large numbers of people, cars, and buildings. Winds aren't as brisk in the city, so the heat isn't dissipated as efficiently.)

Donoghue said that many of those who suffered heat-related deaths were facing imminent death. He was required, however, to list heat as a contributing factor because of the circumstances under which they were found. Donoghue said heat was a secondary or contributing factor in a majority of the deaths. He says that it would be difficult to determine how many deaths were caused primarily by heat since autopsies weren't performed on all the bodies.


1. How would you define "heat-related death"? Do you agree that the three criteria used by the Cook County medical examiner are excellent?

2. Suppose that during the heat wave a person traveled to Chicago from a rural area, and died there. Would you call such a death heat-related or urban-related?


Science of species law called Into question; best research often full of uncertainty.
Chicago Tribune, 21 July 1995, Page 1
Hugh Dellios

The article by Hugh Dellios is one in a series of articles dealing with the Republican attack on environmental regulation.

The article deals with research done by scientists to identify endangered species. It focuses on the current debate over the 22-year-old Endangered Species Act, and whether the science behind the act was sufficient to determine environmental regulations.

Dellios uses the plight of the coho salmon as an example. The National Marine Fisheries Service used "the best available science" in its research to determine the status of the salmon. Its experts proposed listing coho from California, the Oregon Coast, and the Columbia River basin as "threatened." The experts, however, voiced doubts about their results since they were unable to resolve some basic issues such as the cause of the decline of the fish population.

The author writes that the mysteries of saving imperiled plants and animals have made the Endangered Species Act vulnerable to attacks on its scientific integrity. In addition, listing decisions that later seemed wrong have only aggravated the situation. Critics contend the Act restricts land use and favors species that do not deserve protection. Supporters, on the other hand, argue that the research behind the Act is based on "sound science".

Compared to such activities as measuring air and water quality, calculating the risk of extinction or the chance of recovery of a particular species is far more difficult. At the end of the article, the six articles of the series are listed with a short description of each. The entire text file can be downloaded from the Tribune Reference Desk Library in the Tribune area from America Online.


1. How do you think the "chance of extinction" is determined?

2. Assuming such a probability can be computed, what would be your cut-off point for protecting a plant or animal species?

3. What other factors would you include as criteria for species protection?


Census report finds Hispanic Americans lagging further in college degrees.
The New York Times, 27 July 1995
Steven A. Holmes

In a new report, "Hispanics-Latinos: Diverse People in a Mulitcultural Society", Jorge del Pinal, former chief of the Ethnic and Hispanic Statistics Branch of the Census, reported that the percentage of Hispanic Americans who have college degrees is falling behind the percent of non-Hispanic- Americans who do.

Only 9 percent of Hispanic-Americans 25 or older held bachelor's or advanced degrees in 1994, while 24 percent of non-Hispanic Americans have one. In 1970, the Hispanic figure was 5 percent and the non-Hispanic figure was 11.6 percent. However, there was little gain for Hispanic- Americans since 1980, when the figure was 8 percent. The report also notes that the number of younger and older Hispanics who have degrees are about equal as compared to the broader population in which degrees are held by a higher percent of younger people than older.

Some factors cited for the trend were poverty, discrimination, and lack of university recruitment. The results of the Census excludes recent immigrants, whom Mr. del Pinal describes as having a greater antipathy towards higher education.

The study comes in the wake of the backlash against affirmative action and the recent decision of California's Board of Regents to end affirmative action programs in public colleges and universities.


1. What is the best way to measure the difference between the percentages given above?

2. What are the possible cause for these differences? How might you determine which causes are more likely?

3. How do you think the figures cited will be used in the affirmative action debate?


Researcher's find another Alzheimer's link: flawed gene combination; chances of developing the disease increases exponentially.
The Fresno Bee, 1 August 1995, Pg. C6
Claudia Coates, Associated Press

This article provides a good example of the need for careful reading. Something is wrong with the figures given below -- can you find the error(s)?

A pair of genes has been identified as linked to the development of Alzheimer's disease, and, according to the article, of the 3.6 million people in the US with sporadic Alzheimer's, about one in five carry both flawed genes. In the general population, the figure given is one in 600.

Another pair of genes has already been shown to make the development of Alzheimer's 11 times more likely. In the presence of this new pair, the risk more than triples to 34 times as likely. Alone, the new pair increases the risk 1.5 times.

Nevertheless, age is cited as the greatest risk factor for Alzheimer's. Nearly one half of all people who reach age 85 will develop the disease, the article says.


Drugs are no. 1 concern of youths, survey says.
The Arizona Republic, 18 July 1995, Pg. A4
Scripps Howard

This article describes some of the highlights of a recent survey sponsored by Columbia's Center on Addiction and Substance Abuse (CASA). The survey of 2000 American adults and 400 youths, aged 12 to 17, was conducted by telephone, and, to select the adult sample, pollsters at Luntz Research used a random digit dialing technique "to construct a nationally representative sample," according to the report released by CASA. For the adolescent sample the same technique was used but calls were centered in "areas of high incidence of at-home youths of the target ages," the Center reports, and "as such the youth sample can be considered to be illustrative but not strictly nationally representative." Curiously the article says that survey director Frank Luntz calls the 400-member youth sample small but still a "representative cross section of the country's youth."

The title of the article reflects answers to a survey question which asks the teenagers to identify the problem that concerns them the most-- drugs was the most common response (32 percent), followed by crime and violence in school (13 percent) and social pressures like popularity (10 percent). The article also states that Joseph Califano Jr., head of CASA, "blasted" the entertainment industry for encouraging drug use, especially among adolescents. A careful look at the survey results shows that just three percent of the kids questioned and 4 percent of adults picked Hollywood glamorization as the main reason young people start taking illegal drugs. At the same time, when asked to evaluate just how much movies and T.V. encouraged illegal drug use in general, 24 percent of teenagers and 16 percent of adults surveyed said a great deal.


1. Do you think that the above "illustrative" sample is representative of some larger population of American youth? Would you be prepared to generalize the results obtained from such a sample to this larger population?

2. The article states that the sample of 2000 adults has a margin of error of 2 percent, while the youth sample of 400 has a margin of error of 5 percent. What do these numbers mean? Is 400 a "large enough" sample?

3. Does the title of this article correctly reflect the meaning of the 32 percent figure found through the poll? From this figure would you agree with Califano that "drugs frighten young people more than anything else."?

4. The article states that the survey indicates that 81 percent of teens believe their worst temptation is marijuana, which they regard as the "gateway" drug to later use of cocaine, heroin, lsd and other drugs.

This remark appears to be based on the responses to the following questions.

Please tell me how using each of the following items will effect one's health if at all.
1. The use of marijuana -- is it:

(45%) Very dangerous
(35%) Fairly dangerous
(14%) Not too dangerous
(3%) Not dangerous at all
(3%) Don't know-no response

2. Do you think that if a person your age is using marijuana, he or she will be more likely to eventually use other illegal drugs like heroin or cocaine or does using marijuana not make using other illegal drugs more likely?

(81%) More likely
(16%) Not more likely
(3%) Don't know or know response.

Are these questions reasonably worded? Do you think the author of the article summarized the results of these two questions reasonably?

5. The article refers to the gateway theory. What proportion of the teen-agers who smoke marijuana go on to other drugs such as cocaine and heroin? What proportion of those who use cocaine or heroin smoked marijuana before doing so? Which of these would easiest to determine. How are these two percentages related?



CHANCE News 4.11
(21 July 1995 to 18 August 1995)