| Date |
Topic
[omitted] |
Topic (as in \WMS5)
Notes(Link) |
Assignment due on
this date.
[F] indicates final version
|
Resources |
| |
|
|
|
|
| T 1 |
Introduction |
1 |
|
Excel Resources:
basic tutorial (from USD)
quickstart (from F&Marshall
College)
statistical
tables via Excel (U Coventry)
homegrown (by JH
in early 90's)
General
Excel Links (J Walk and Co)
Data: Breast Cancer Surgery
500_92_98 all92-98 freq distn CMAJ
Table: Population by knowledge of official languages, Canada 1996 Census |
| W 2 |
Probability |
2.1-2.6 |
(web) SURVEY
Q1 and Q3(b) from
exercises_around_Ch1
[F] |
Cards: first_ace
Lotteries: Loto-Québec
Coding Theory: hat problem
Epidemic: simplified
Queues: |
| J 3 |
Probability |
2.7-2.9
jh_notes
on MM_Ch4 |
Exercises
2.8, 2.9,
2.14, 2.15
from WMS5
Q3 (homegrown)
from p2 of
exercises around Ch2
[F] |
Birthdays: c607 1981-93 simulate
Gambling 17th century: deMéré
Text 100
games even more (Macro)
Lotteries:
Massachusetts Mass_NH Canada
Electric circuits: WMS5 2.61 2.62
Basketball: Cold Facts about Hot Hand
Statistics students 20th/21st century: random
sequences
NASA on Probabilities: 1
2 3
Tree Diagrams: Binary Tree(2) Binary Tree(3) |
| F 4 |
Probability |
2.10-2.13
jh_notes
on MM_Ch4 |
Exercises
2.48, 2.58,
2.61a, 2.62
from WMS5
Use a tree diagram to verify the statement "three-fourths of the time..."
underlined on p3 of NY Times article on hat problem
Q1 (homegrown / la Quotidienne 3) from p2 of
exercises around Ch2
[F] |
Molecular Biology: transgenic mice
text simulate
Genetics: haemophilia
Screening for HIV: Can we afford the False Positive Rate? Meyer and Pauker (NEJM1987)
pages 238 239 240 241
Monty Hall (Gameshow) Problem
link1 link2
Claude E. Shannon (1916-2001) |
| |
|
|
|
|
| M 7 |
Discrete Random Variables
[3.5,
3.6,
3.9-3.11] |
3.1-3.3 |
Exercises
2.89, 2.90,
2.117, 2,120 (Solution)
from WMS5
Q3 (lifetable) from
exercises_around_MM_Ch4
Draw a tree diagram (and use probability notation) to represent the assumptions and
calculations in testing
all Americans for HIV
Solution
Data Analysis
(Write up may be from group of up to 3)
Q2 (Breast Cancer data) from
exercises_around_Ch1
NOTE (added Fri 4th, at 1 pm) The data analysis
part may take some time and so it can be handed in up to Wednesday 9th. The rest
of the assignment (everything else mentioned in this box) is still due on Monday
7th.
NOTE (added Mon 7th, at 1 pm)
Wording of Q2 a) not as clear as it could be. Q2 a) asks for a histogram for '92,
and another for '98; it also asks for your impression of how much the '98 distribution
is different from the '92 one; back up your impression by reporting and comparing
some relevant descriptive statistic from '92 with its counterpart for '98.
[F] |
Expected Values
position of 1st_ace
longevity (life expectancy)
position of 1st duplicate birthday
Returns on Lotteries:
Waiting for an elevator:
|
| T 8 |
Discrete Random Variables |
3.4(Binomial)
3.5(Geometric) |
|
Binomial(n=10, various
p)
Time spent inside
GM vs Microsoft |
| W 9 |
Discrete Random Variables |
3.7(Hypergeometric)
3.8(Poisson)
jh's notes from
course 626
most relevant pages are 2-5, right 1/2 of
9, 10-14, 17-18 |
Exercises
3.13, 3.19, 3.21, 3.24
from WMS5
Question 1 (cluster of miscarriages) from Exercises on Counts and Proportions at
end of jh's notes on 3.4(binomial)
Use Excel (or your calculator) to make a probability histogram for the Binomial(n=145,
p=0.528) situation described in Question 7 (sex ratio) in the exercises at end of
jh's notes. Does the shape of the diagram suggest a faster way to assess the 'extremeness'
of the observed proportion.
[F] |
Hypergeometric_P_E_V.xls
Tamoxifen
Labour dispute text Example 2.10p39
Rhino
Politics and small sample sizes
Banco pdf Excel
Keno[20/80] pdf Excel
Cell Occupancy Excel cell_occupancy_macro |
| J 10 |
Continuous Random Variables |
4
in class
on blackboard
and via projector |
|
Class Data Excel |
| F 11 |
Continuous Random Variables
[4.7,
4.9,
4.10,
4.11] |
4.5(Normal) |
Exercises
3.75 [calculate Prob(1 or 0 African American)], 3.77, 3.88,
3.91[for 3.91 use both the binomial (the appropriate one) and the Poisson
(approximation to binomial], 3.97 from WMS5
SOLUTIONS
See the A vs C comparison in the politics and small sample sizes story(last page).
Assuming that these 3 rhinos would have died anyway, and that survival has nothing
to do with de-horning, reconstruct the probabilities of observing each each of the
4 possible 2 x 2 tables.
By back-calculation from the confidence intervals, one can deduce that the headline
'Leukemia rate triples' on page 24 of jh's notes from course 626 is based
on just 2 cases, and that the expected number is E=0.57. Assuming a Poisson distribution,
what are the chances of observing 2 or more if E =0.57 (the situation at Douglas
Point)? If you like, think of the 0.57 as the average number of cases for non-nuclear
communities of this size.
What are the chances of getting 18 or more when the expected number is 12.8 (the
situation at Pickering)? Which is statistically more extreme, the ratio 2/0.57=3.5
or the ratio 18/12.8=1.4?
Draw the probability histograms for the Poisson distributions having means (expected
values) of 0.57 and 12.8 [Excel has the Poisson probability function build in, just
like the binomial and the hypergeometric] What lessons can you draw as to when we
could approximate the Poisson distribution by the Gaussian (Normal/Bell shaped) distribution.
You might want to visualize Poisson distributions for several values of E [the mean
is known as E-- for Expected no. -- in Epidemiology, as lambda (for ??) in
statistics]. See distributions for several other E's on pages 18-19 of jh's notes.
[F] |
|
| |
|
|
|
|
| M 14 |
Continuous Random Variables |
4.6(exp_gamma) |
QUIZ for last 1/2 hour of Class on Monday 14th.
Grades on Quiz
Solutions Q 1
2 3 4 5 6 |
Random Events in Time time_macro |
| T 15 |
Multivariate Distributions
[5.9,
5.10,
5.11] |
5 |
Exercises
4.7, 4.9, 4.13, 4.19, 4.22, 4.47[good for practice on Gaussian distribution], 4.55,
4.64, 4.66, 4.108, 4.109, 4.112 from WMS5
[PS: draw the pdf and cdf for example 4.112 and decide if the question is likely
to have been 'made up' or real]
SOLUTIONS
[self-correct the above exercises
using solutions at back of book or to be provided on web site later this week; DO
NOT HAND IN]
Exercise 1 (iodine level in salt) from page 5 of jh's notes on 4.5(NormalIf you can do this problem,
you know your way around the Gaussian Distribution pretty well.
[self-correct the above exercise
using solution provided on right hand side of the same page 5; DO NOT HAND IN] |
|
| W 16 |
Multivariate Distributions |
5 |
Exercises (self-correct)
5.6, 5.9, 5.11 from WMS5
SOLUTIONS
5.18, 5.21, 5.36, 5.37,
5.42, 5.43 from WMS5
SOLUTIONS |
Heights
of parents and children
Galton on marriage selection
p84-85 p86-87
p88-89
Heights
of individual parents |
| J 17 |
Multivariate Distributions |
5 |
|
|
| F 18 |
Functions of Random Variables |
6 |
Q1, Q2 and Q3 (homegrown -- Galton data) from exercises around Ch5
[F] |
The Trial of the Pyx
1799 1993 1998 photos
Stigler Article
Book |
| |
|
|
|
|
| M 21 |
HOLIDAY - NO CLASS |
- |
|
|
| T 22 |
Functions of Random Variables |
6
Notes on Ch 6 |
QUIZ for second hour of Class on Tuesday 22nd.
THE QUESTIONS ON THE QUIZ WILL BE SIMILAR TO THOSE IN THE FOLLOWING..
Exercises around Chapters 4
and 5(updated 18:30 Sunday May 20)
Solutions to Q1-Q5 of Quiz
(updated 23:10 Saturday May 26) |
|
| W 23 |
Sampling Distributions |
7
Student-t distribution |
|
If Gossett had had Excel... |
| J 24 |
Sampling Distributions |
7
sampling variability of means
JH's notes from_MM_ch5 |
Q1 (homegrown) from
Exercises around Chapters
6 and 7) |
word length and average word
length... |
| F 25 |
Sampling Distributions |
7 |
|
|
| |
|
|
|
|
| M 28 |
Review |
|
Q2 and Q3 (homegrown) from
Exercises around Chapters
6 and 7 (updated) |
|
| T 29 |
.. |
|
THE QUESTIONS ON THE FINAL EXAM WILL BE SIMILAR TO THOSE IN
THE FOLLOWING..
Exercises around Chapters 1 to 7
(added to, 23:15 Sunday May 27)
CHECK AGAIN LATER.. More Q's to be added.. |
|
| W 30 |
FINAL
EXAM |
|
|
|
