NATS 1500 2017W

From MathWiki

(Redirected from NATS1500)

NATS 1500: Statistics and Reasoning in Modern Society 2017 W

Where there is no uncertainty there cannot be truth. -- Richard Feynman (
The best thing about being a statistician is that you get to play in everyone's backyard. -- John Tukey (
Statistical literacy is a necessary precondition for an educated citizenship in a technological democracy -- Gerd Gigerenzer et al. in Helping Doctors and Patients Make Sense of Health Statistics (
A certain elementary training in statistical method is becoming as necessary for anyone living in this world of today as reading and writing. – H. G. Wells in "The Informative Content of Education," The Presidential Address to the Educational Science Section of the British Association for the Advancement of Science, given on September 2nd, 1937.

Quick Links

Breaking News

Final Exam

The final exam will take place on Sunday, April 9 at 10:00 am in TC AVIVA (the Tennis Canada Aviva Centre) in rows 7 to 13. Plan to be there 30 minutes ahead of time to be sure you won't be late.

Here a tip sheet on exams from the Registrar (

Office hours

Fridays 8 am to 10 am or at other times by appointment.

Table of contents


Week Links and Assignments
1 Monday, January 9



  • Due Wednesday: Complete this survey (
Wednesday, January 11

Readings for Monday:

  • Read pp. 1-28 of Wainer: Truth or Truthiness

Assignment: Due: Monday

  • Right after this class you will receive an e-mail message inviting you to join the 'Piazza' class forum. Follow the instructions to join.
  • Post a message describing some decision you need to make that would be helped if you had access to valid and reliable evidence about the consequences of the decision. The decision could be an ongoing one like a decision about some aspect of diet or nutrition. Or, it could be a 'bigger' decision about your life. Describe the decision and discuss briefly how good evidence could have an impact on it. Note that the posting is public.
  • Finish the survey if you haven't done so already.
  • Join Piazza
2 Monday, January 16
  • Lies, Damned Lies ... (

Note: There was an error in the weights for course components and they embarrassingly didn't sum to 100%. The error has been corrected and you should have a new look at the weights.

Hans Rosling and global health

Which of the following pairs of countries has the higher child mortality:

Sri Lanka Turkey
Poland South Korea
Malaysia Russia
Pakistan Viet Nam
Thailand South Africa
  • Hans Rosling's TED talk on changes in global health (
    Watch and answer this question:
    Explain why Rosling says that the Swedish university professors who give out the Nobel Prize know the world approximately as well as chimpanzees ... and Swedish university students know significantly less about the world than chimpanzees.
  • The Gapminder site ( with links to the Gapminder software, data and more.

Team Project: Team Project Description Due: January 30, 2017.

  • Teams are already assigned on Moodle.
  • I'll send email messages to each team tonight so you can contact each other.
  • Each team name will be a 'group' on Piazza and you can 'join' your team and use Piazza for team discussion visible only to members of your team and instructors.

Assignment: due Wednesday: Join your team on Piazza.
For Wednesday: Bring your laptops and you'll work on the Team Project with your team during the tutorial.
In your spare time: Watch John Oliver's Last Week Tonight: Scientific Studies, May 8, 2016 ( (Beware: Very mature -- or immature?? -- language)

Wednesday, January 18



  • Read Chapter 4 of Wainer (hardest chapter in the book) and prepare to discuss on Monday.

Team Assignment: (due Monday)

  • Choose one question that involves some uncertainty about causality based on an observed association. Choose a topic that we have discussed in class or one that has been posted on Piazza or one that has been discussed in the textbook by Wainer. For example:
    • Cultures that practice circumcision exhibit a lower rate of cervical cancer. Does circumcision decrease the risk of cervical cancer?
    • The 1964 US Public Health survey showed that, controlling for age and sex, people who quit smoking were, on average, less healthy that people who continued. Is quitting smoking harmful to your health, on average?
    • People who use sunscreen lotion have a higher rate of skin cancer than those who don't. Does using sun screen lotion increase the chances of getting skin cancer?
  • Come up with at least three conjectures that would explain the association:
    • one involving a mediating variable(s) that would be consistent with the possibility that the 'X' variable causes the 'Y' variable
    • one involving a confounding factor(s) that provide alternative explanations for the association that don't imply causality, and
    • one involving reverse causality or involving selection.

Submit the assignment as a Word or PDF file by uploading it to 'Team Assignment 1' on Moodle. One one member of the group needs to submit for the group. At the top of the submission, list the names of those who participated.

Tutorial: Work on Team Project.
3 Monday, January 23


  • On Causality (
  • Chapter 1 (

Assignment: by Wednesday: Participate on Piazza to define and discuss:

  • What's the difference between studying 'causes of effects' or 'effects of causes'? Why does Rubin prefer to focus on one over the other?
  • What is a counterfactual? What is its relevance for causality?
  • What is the relevance of a 'gedanken experiment' to causality?
  • What can it mean to say "exercise causes improvements in health"? How would you be able to tell whether this is true or not?
  • What can it mean to say that "higher interest rates cause inflation"? How could you tell whether this is true?
  • Discuss the meaning of 'confounding factor', 'mediating factor', 'moderator variable', 'internal validity', 'external validity', 'reliability', 'validity', 'accuracy', 'precision'.
Wednesday, January 25
  • Slides: Chapter 2 (

Work on team project

4 Monday, January 30
  • Slides: Chapter 2 (
Wednesday, February 1
  • Slides: Chapter 3 (

Individual Assignment 5 Due Monday, Feb. 13 (this assignment has the weight of 3 regular assignments)

  • Answer these questions in the textbook by Utts and Heckard. Submit your answers in one PDF or Word file through Moodle. PDF is preferred.
5 Monday, February 6

Chapter 2 slides with notes (
Chapter 3 slides with notes (

Wednesday, February 8

Chapter 3 slides with notes (

6 Monday, February 13

Individual Assignment 6 due Wednesday February 15:
Have a look at the Wikipedia page on Simpson's Paradox ( On a sheet of paper, draw a Paik-Agresti diagram (as described in class) to show the proportion of applicants admitted by department and overall. Use Gender for the X-axis and the proportion admitted for the Y-axis. Draw lines connecting the proportion of women admitted and the proportion of men admitted within each department. Use a different color to draw a line for the overall proportion. Label each point with the percentage admitted represented by that point. Label each line with the department represented by that line.
Hand in your diagram (make sure to include your name and student number) at the beginning of class on Wednesday.

Wednesday, February 15

Individual Assignment 7: Due Monday, February 27:

  • Read Chapter 4. Find an interesting link commenting on polling problems either in the recent U.S. election or in the last Canadian election. Post the link on Piazza along with your comments on what principles discussed in Chapter 4 are illustrated in the link you have posted
  • I'm sorry the following text was entered by mistake. The actual assignment is the Piazza assignment in the previous bullet.
  • Assignment: due Monday, February 27 at 9 am: Read Chapter 4 of textbook. If your student number ends with an even|odd number, do the even|odd numbered questions in the notes on the textbook (see the quick links above). Do this all in a R script file with Markdown. Email me the .R file on Monday morning.
Reading Week: February 20 to February 24
7 Monday, February 27
  • Description of individual project due April 5 (
  • Sample Mid-Term Test ( I will take this up during the tutorial on Wednesday.
  • Chapter 4 slides (
  • Visualizing Regression (
Wednesday, March 1
Mid-term test: Saturday, March 4, 8 pm to 9 pm
8 Monday, March 6
  • Chapter 5 slides ( | with notes (
Wednesday, March 8
  • Chapter 6 Slides ( Relationships between categorical variables
March 10: Last day to drop a course without a grade
9 Monday, March 13

Individual Assignment 8 Due Wednesday, March 22 at 11 pm (this assignment has the weight of 3 regular assignments)

  • Answer these questions in the textbook by Utts and Heckard. Submit your answers in one PDF or Word file through Moodle. PDF is much preferred.
Wednesday, March 15

Chapter 6 slides with notes (

10 Monday, March 20
Wednesday, March 22
11 Monday, March 27
Wednesday, March 29
12 Monday, April 3
Wednesday, April 5 (last class)

We'll take up the sample exam. (

  • See the files folder ( for annotated versions of the sample final.
  • See the video files for a file call 'Sun.mp4' for a 'clearer' explanation of the XKCD cartoon in the sample exam. The pdf file is Sun.pdf (
Final exam period: April 7 to 24

Interesting links

Some places to look for Statistics in the News

  • (, a data journalism site founded by Nate Silver who is famous for his accurate predictions of U.S. elections.
  • Toronto Star (
    Many articles in daily newspapers use data or make claims that benefit from a critical analysis.
  • The Guardian (
    The Guardian is one of two international newspapers that have a special reputation for the effective use and presentation of data. The other is:
  • The New York Times ( (limit of 10 articles/month without a subscription)
  • New York Time Opinion Pages ( (doesn't require a subscription)
  • New York Times online access through York (
  • This is Statistics (, a site created by the American Statistical Association.
  • Significance ( published by the Royal Statistical Society.
  • Stats Chat (, a blog by Thomas Lumley
  • Statistical Modeling, Causal Inference and Social Science ( a blog by Andrew Gelman. This one is quite advanced, even graduate students find it challenging.

A few interesting articles

Causality and Climate Change

Relevant Books