Presentations.html

Chance Workshop
Group Presentations
Dartmouth College
July 1998

I. Black Smokers and Nicotine--An Absorbing Issue
Melissa Cass, Carol Janik, Josephine Rodriguez, Steve Terry,
Beth Walters

II. Non-Cents: A Simulation Activity
Kal Godbole, Leona Mirza, Alice Richardson, Mark Rizzardi

III. Clinical Trials
Charlotte Buffington, Angela Hare, Ellen Musen, Nancy Roper

IV. Therapeutic Touch
Eli Brettler, Rita Kolb, Raj Prasad

V. Chasing the Home Run Record
Michael Dutko, Richard Iltis, Steve Samuels, Linda Thiel,
John Wasik

VI. Where Have All the Boys Gone?
Steve Givant, Eunice Goldberg, Ellen King, Barbara Stewart

I. Black Smokers and Nicotine--An Absorbing Issue
Melissa Cass, Carol Janik, Josephine Rodriguez, Steve Terry, Beth Walters

Workshop Handouts

Studies suggest black absorb more nicotine: Link to higher cancer rate is hinted. The Boston Globe, 8 July 1998, p. By Richard Saltus.

Studies show that black smokers absorb more nicotine. The Valley News, 8 July 1998, pB1. By John Schwartz.

Black smokers retain more nicotine, ABC News Online, July 10, 1998.
http://www.abcnews.com:80/sections/living/DailyNews/
[links updated daily--it does not appear that archives can be accessed!]

Discussion questions based on above articles.
Additional References
The group also distributed copies of the following three articles from the July 8 issue of JAMA:

Caraballo, Ralph S. et. al., Racial and ethnic differences in serum cotinine levels of cigarette smokers. Journal of the American Medical Association 1998, 280: 135-139.

Prez-Stable, Elisio J. et. al., Nicotine metabolism and intake in black and white smokers. Journal of the American Medical Association 1998, 280: 152-156.

(Editorial), Pharmacogenetics and ethnoracial differences in smoking. Journal of the American Medical Association 1998, 280: 179-180.

II. Non-Cents: A Simulation Activity
Kal Godbole, Leona Mirza, Alice Richardson, Mark Rizzardi

Workshop Handouts
Exceprt from article in Milwaukee Journal, May 1992.

Worksheets for classroom simulation activities and discussion questions.

III. Clinical Trials
Charlotte Buffington, Angela Hare, Ellen Musen, Nancy Roper

Workshop Handouts

Magazine ad for ZYRTEC (allergy medication) & discussion questions

Quiz on terms from medical experiments

Additional References

Jessica Utts, Seeing Through Statistics. Chapter 5 "Experiments and Observational Studies"

IV. Therapeutic Touch
Eli Brettler, Rita Kolb, Raj Prasad

Workshop Handouts

Hallelujah! Science looks at prayer for friend and fungus. The New York Times on the Web, 5 April 1998. By Jeff Stryker.
http://search.nytimes.com/search/daily

Rosa, Linda et. al., A close look at therapeutic touch. Journal of the American Medical Association, 1998, 279, 1005-1010.

Chance class discussion handout to accompany above

Additional References

Eli Brettler posted the group's handout to his web page, including links to the articles online!

V. In Pursuit of the Home Run Record
Michael Dutko, Richard Iltis, Steve Samuels, Linda Thiel, John Wasik

Workshop Handouts
McGwire gets better, and a record looks more vulnerable. The New York Times, 9 July 1998, A1. By Buster Olney
Discussion questions based on the above article

Additional References
VegasINSIDER online sports betting site report on wagers regarding MdGwire
SportingNews online statistics on McGwire

Addendum.
Players who have hit 30 home runs before the All-star game.
Plaryes who hit 30 or more homeruns after the All-star game.
Side-by-side boxplots of McGwire's home run distances, home and away
and time series plot of home run distances, with home/away plotting symbols

VI. Where Have All the Boys Gone?
Steve Givant, Eunice Goldberg, Ellen King, Barbara Stewart

Workshop Handouts
¥ Alpert, Mark. Where have all the boys gone? Scientific American, July 1998 issue online:
http://www.sciam.com/1998/0798issue/0798scicit3.html
¥ Discussion questions based on the article

Additional References
¥ Online JAMA abstract. Davis. D.L. et. al., Reduced ration of male to female births in several industrial countries. A sentinel health indicator? Journal of the American Medical Association 1998, 279: 1018-1023.
¥ Vital Statistics. No. 89. Birth and Birth Rates: 1970-1992. (Photocopy from US Statistical Abstract)
¥ Report: Environmental factors may support dip inmale birth rates.
http://www.sportserver.com/newsroom/ntn/health/033198/health19_4144_noframes.html
Addendum
The group sent us an extended write-up after the meeting. It is reproduced on the following pages.

Chance Workshop
July 1998
An Addendum to the Group Presentation:
Ellen King
Eunice Goldberg
Barbara Stewart
Steven Givant

from Eunice Goldberg, in consultation with Ellen King

Article:
Albert, Mark
Where Have All the Boys Gone
Scientific American -July 1998
www.sciam.com/1998/0798issue/0798scicit3.html

Other Articles and Sources:

Tanner, Lindsey
Report: Environmental factors may support dip in male birth rates
Chicago, Associated Press, 1998

Davis, Gottlieb, Stamplinitzky
Reduced Ratio of Male to Female Births in Several Industrial Countries
JAMA Abstracts, April 1, 1998

U.S. National Center for Health Statistics: Vital Statistics of the United States (Birth rates by gender and ethnicity, infant mortality, ages of mothers and fathers, etc.)

JAMA, New England Journal of Medicine

Recommended searches

Information on male-female ratios in the rest of the world
Government policies that promote one gender over the other (China, India, Middle East, Latin America, etc.)
Long range effects of unbalanced ratios
Environmental factors that may be gender biased.
War, etc.

Recommended Activity--Population Simulation Problem

Several versions of this problem have circulated:

One version is that a country prefers males and families keep having children until a male is born. A second version says that females are preferred and therefore families keep having children until they have a female.
Other versions say that a country does not want to take a chance on having too many girls-- so once a girl is born, the family must quit having children. The government does not care how many boys are born.

What can be learned?

1.The analysis of the article helps students become quantitatively literate. By asking questions and doing research, the students will learn to evaluate and make sense of data.

2. The problem helps students learn about issues of probability, simulation, and distributions of data.

3. There can be meaningful integration of subject areas: social studies and mathematics and/or science.

For the analysis of the article students could possibly: learn to define the issues being presented and investigate whether the data actually explains or enlightens

Think about ways to obtain needed information either through further research or investigation
ask other questions about this issue that are compelling or need clarification and should be investigated, ie. explain the relationship, if any, between a declining ratio and a declining birth rate

interpret, analyze, question, and evaluate data: enough, convincing, misleading, contradictory, etc.
think about what information is needed to inform the issue: demographics, original studies, census data, etc.
learn to find backup data: internet, journals, newspapers, experts, etc.

gain experience looking at large quantities of information and sifting out what is important and/or meaningful
learn to ask "what ifs": what if the ratio became very disparate? what if it continued for many generations? what if science changes the way babies are reproduced?
etc.

By doing the problem students could:

generate their own data

have the experience of dealing with a data set that is messy yet still shows the theoretical patterns

look at measures of central tendency: mean, median, mode

look at spread: standard deviation and eyeballing the distribution

learn to simulate two-choice problems using different two choice simulations

learn the need for lots of data before patterns appear--learn that any one sample may not show the pattern-- or even a few samples may not--

compare the results of experimental simulations to the theoretical model of this problem

compare the variation between the number of boys or girls (depending on the version) in individual families-- and the average for many families.

look at the shape of the distributions for average number of boys vs. girls, and the distribution of the number of families with zero girls, 1 girl, 2 girls, 3 girls, etc.

compare real life to the simulation:

Is there a physical limit on the number of children one family can have?

Is the ratio really 50-50?

Can we simulate other ratios using these methods?

What do those other ratios say for future generations?

Are the simulations reasonable models?
etc.

Time frame: At least four hours of class time

materials:

(Copies of the article, questions presented in class, additional articles and data can be included here)
I am adding a copy of the problem as I used it in a workshop. It is similar to what Ellen said she did.

Males Preferred
Does it Work?
There are countries in the world where wives are considered to be a failure unless they produce a male child. King Henry the VIII beheaded several wives because they didn't produce a male heir.
There are still countries in the world that express a preference for boys. A certain unnamed country prefers boys to girls and thinks it has come up with a way to guarantee more boys. The leaders of the country developed this family-planning scheme:
A couple will continue to bear children until a son is born, at which time they will stop having children.
Do you think this family-planning scheme will produce more boys or more girls?
Explain how you made your decision.

(As presented by Eunice Goldberg at Barat College, 1998)

SAMPLE 1: Using coin tosses simulate the generation of 20 families. Let heads represent boys and tails represent girls.

FAMILY NUMBER
TALLY BOYS
TALLY GIRLS
# GIRLS
1

TOTAL BOYS:

TOTAL GIRLS:

SAMPLE 2: Using coin tosses simulate the generation of 20 families. Let heads represent boys and tails represent girls.

SAMPLE 3: Using THE RANDOM NUMBER TABLE to simulate the generation of 20 families. Let even numbers stand for boys and odd numbers stand for girls.

SAMPLE 4: Using THE RANDOM NUMBER TABLE to simulate the generation of 20 families. Let even numbers stand for boys and odd numbers stand for girls.

SAMPLE 5: Using a DIE simulate the generation of 20 families. Let 1,2, or 3 be a boy and 4, 5, and 6 be a girl.

SAMPLE 6: Using DICE simulate the generation of 20 families. Let 1,2, or 3 be a boy and 4,5, and 6 be a girl.

Answer the following questions:
1. How many families did you generate altogether?____________
2. How many families had zero girls?_______
3. How many families had exactly 1 girl?_____
4. How many families had exactly 2 girls?_____
5. How many families had exactly 3 girls?____
6. Do you see any patterns?_______
7. What number occurred the most often?____
8. What is the average number of girls in a family?_______
Notes to me:
Make charts:

1. Make a spread sheet of samples derived from coin tosses, random number tables, and tosses of die. Subject name on the side, type of sample across the top. (This can be done in EXCEL-- with a split screen and a graph so the graph will change as data is added in-- or else it can be done on the board where students make dots along the distribution values.)

2. Look at the shape of the data after each set of data has been input. Does it gradually get a more normal shape?
Choices: Make separate plots for coins, random number tables, and die-- all should have similar shape--and put all on one plot to see how the shape normalizes.

CONCEPT:
IN THE LONG RUN THE THEORETICAL DISTRIBUTION WILL START TO APPEAR. INDIVIDUAL SAMPLES CAN BE ALL OVER THE PLACE. YOU MUST LOOK AT THE PATTERN IN THE LONG RUN.

3. What is the overall mean?

4. Find the mean for each subject's six samples. Then find the mean of those means. How does it compare to the overall mean? Look at the histogram of that data. How does it compare to the overall histogram? Look at the spread.
CONCEPT: THE MEAN OF THE MEANS SHOULD BE CLOSE TO THE OVERALL (POPULATION) MEAN.
THE SPREAD SHOULD BE TIGHTER. LARGER SAMPLES YIELD TIGHTER, MORE PRECISE DATA.
5. Find the mean for each of the six samples. How do the different methods of sampling compare to each other?
CONCEPT: TWO CHOICE METHODS OF SIMULATION SHOULD REVEAL SIMILAR RESULTS.

What is the mean of the six samples?

How does it compare to the overall mean?

How does the graph compare to the graph of the overall data and the data of the means of the subjects?

CONCEPT: THE MEAN OF THE MEANS SHOULD BE SIMILAR TO THE OVER ALL MEAN. THIS GRAPH SHOULD BE EVEN TIGHTER BECAUSE THE SAMPLE SIZES ARE LARGER.

Make a chart of individual families.

How many times were there zero girls?

How many times was/were there 1 girl? 2 girls? 3 girls? 4 girls?

CONCEPT: ZERO GIRLS IN A FAMILY WAS THE MODE, AND EACH INCREASE OF A GIRL WAS HALF AGAIN AS MANY FAMILIES AS THE PREVIOUS NUMBER.

THE SHAPE OF THE GRAPH IS SKEWED AND ASYMPTOTIC-- UNTIL YOU REACH A POINT WHERE IT WOULD BE PHYSICALLY IMPOSSIBLE FOR A WOMAN TO HAVE THAT NUMBER OF CHILDREN.

What does the distribution of boys look like? What is the mean, median, and mode?

CONCEPT: WHAT DOES A GRAPH LOOK LIKE WITHOUT ANY SPREAD.

Compare that to the distribution of girls. Does this method produce more boys?

LOOK AT THE THEORETICAL MODEL.