Predicting Oscar
winners
Each year, hundreds of millions of people worldwide watch the televised Oscars
ceremony. Can one predict which films and which directors, actors and actresses will
win the Oscars? Iain Pardoe believes that he can.
Predicting Oscar
Cox Proportional-Hazards Regression for Survival Data
Appendix to An R and S-PLUS Companion to Applied Regression
John Fox
Februrary 2002
1
Introduction
Survival analysis examines and models the time it takes for events to occur. The prototypical such eve
STA 442/2101F Homework 2.
due Wednesday, November 3, 2010 before 4 pm
When answering questions requiring numerical work, the results are to be reported in a narrative summary, in your own words. Tables and Figures may
be included, but must be formatted al
STA 442/2101F Homework 3.
due Wednesday, December 8, 2010 before 4 pm
When answering questions requiring numerical work, the results are to be reported in a narrative summary, in your own words. Tables and Figures may
be included, but must be formatted al
R version 2.11.1 (2010-05-31)
Copyright (C) 2010 The R Foundation for Statistical Computing
ISBN 3-900051-07-0
R is free software and comes with ABSOLUTELY NO WARRANTY.
You are welcome to redistribute it under certain conditions.
Type 'license()' or 'lice
Road Trauma in Teenage Male Youth with Childhood
Disruptive Behavior Disorders: A Population Based
Analysis
Donald A. Redelmeier1,2,3,4*, William K. Chan1,3, Hong Lu3
1 Department of Medicine, University of Toronto, Toronto, Canada, 2 Clinical Epidemiolog
STA 442/2101F Homework 1. Solutions
1. STA 442 and 2101 The UN data on homicide rates and income inequality have been
saved in the le spirit and can be read into R using the following:
load(url("http:/www.utstat.utoronto.ca/reid/sta442f/spirit").
The vari
Administration
HW 3 due Dec 8, before 4 pm
Practice questions for nal have been posted
Final will have:
four questions
one question from HW
one question from Practice Qs
some theory questions
some applied questions
one question for STA 2101 only
one quest
Administration
HW 3 due Dec 8, before 4 pm
Wednesday, December 1: nish survival data, questions
on HW 3 or Practice Qs
Tuesday, December 7: random and mixed effects models
(9.4); review/overview
Exam on Dec 17, 2 5 pm: STVLAD Auditorium B, St.
Vladimir In
STA 442/2101F Homework 2.
due Wednesday, November 3, 2010 before 4 pm
When answering questions requiring numerical work, the results are to be reported in a narrative summary, in your own words. Tables and Figures may
be included, but must be formatted al
Survival Data: single sample
Model: f (t), h(t), 1 F (t), H(t)
density, hazard, survivor function, cumulative hazard
Data: (t1 , 1 ), . . . , (tn , n )
tj an observed time
j = 1 if tj a true failure time, 0 if tj is a censoring time
random censorship assu
Annals of Internal Medicine
Academia and Clinic
Do Oscar Winners Live Longer than Less Successful Peers?
A Reanalysis of the Evidence
Marie-Pierre Sylvestre, MSc; Ella Huszti, MSc; and James A. Hanley, PhD
In an article published in Annals of Internal Med
www.inventionhome.com
Ads by Google
November 15, 2010
Male teens with ADHD at greater risk of traffic accidents, a study
says
By CARLY WEEKS
From Monday's Globe and Mail
The findings may signal the need to adopt more stringent requirements to ensure peopl
Example G Cost of construction of nuclear power plants
Description of data
Table G.1 gives data, reproduced by permission of the Rand Corporation, from a report (Mooz, 1978)
on 32 light water reactor (LWR) power plants constructed in USA. It is required t
Example E
A before and after study of blood pressure
Description of data
Table E.1 gives , for 15 patients with moderate essential hypertension, supine systolic and diastolic
blood pressures immediately before and after taking 25 mg of the drug captopril.
STA 442/2101F Notes on generalized linear models.
We can think of the linear model
y i = xT + i ,
i
ind
, i (0, 2 ),
(1)
in two parts. First we specify the mean structure
E(yi | xi ) = xT ,
i
and then we specify the distribution of Yi :
Yi N (xT , 2 ),
i
STA 442/2101F Homework 1.
due Wednesday, October 13, 2010 before 4 pm
When answering questions requiring numerical work, the results are to be reported in a narrative summary, in your own words. Tables and Figures may
be included, but must be formatted al
Example G Cost of construction of nuclear power plants
Description of data
Table G.1 gives data, reproduced by permission of the Rand Corporation, from a report (Mooz, 1978)
on 32 light water reactor (LWR) power plants constructed in USA. It is required t
Administration
HW 1 due October 13, before 4 pm
Ofce Hours Tuesday 4-5 and Wednesday 4-5 SS 6002A
Wednesday: Questions re HW 1
Question of the week
STA 442F: Oct 5, 2010
1/9
From last week
Davison, 8.3, 8.5, 8.6, 8.7
you should be able to do linear regres
Administration
TA Chunyi Wang,
Part time undergraduates: class reps during the break
Statistics Canada information session: Monday Oct 4,
10-12, SS1084
absence policy: record through ROSI and give me a
printout
course notes on web page
www.utstat.utoronto
Administration
HW 3 due Dec 8, before 4 pm
Practice questions for nal have been posted Problem 3
to be replaced!
Wednesday, November 17: return HW 2, questions on HW
3 or Practice Qs
Exam on Dec 17, 2 5 pm
STA 442F: Nov 16, 2010
1/1
From last week: genera
Administration
Statistics Canada information session:
Monday Oct 4, 10-12, SS1084
Samuel Beatty in-course scholarships: application due
November 12
http:/www5.physics.utoronto.ca/students/
undergraduate-program/samuel-beatty-scholarship
HW 1 due October 1
STA 442/2101F Notes on prediction in linear regression.
The interpretation of the matrix expressions for the linear model are more intuitive in the
case of simple linear regression. For this, the model is
y j = 0 + 1 xj + j ,
j = 1, . . . , n,
and we usua
Technical Note: Survivor treatment selection bias and all that
The article about Oscar winners living longer mentioned lead time bias and survivor treatment selection bias. Both of these are usually dened with reference
to studies of new screening tests f
Sta 442/2101 F Methods of Applied
Statistics/Applied Statistics I
http:/www.utstat.utoronto.ca/reid/442F10.html
STA 442/2101F: Methods of Applied Statistics, I
Tuesdays, 2-4 pm, Wednesdays 3-4 pm (Note Time Change)
Fall, 2010
Course description: This cour
J. R. Statist. Soc. A (2008)
171, Part 2, pp. 375394
Applying discrete choice models to predict Academy
Award winners
Iain Pardoe
University of Oregon, Eugene, USA
and Dean K. Simonton
University of California at Davis, USA
[Received September 2005. Revis
Single-Factor
Experiments
Coding Schemes
Testing Contrasts
Post-Hoc Comparisons
Multiple Testing
1
Indicator Variables
Indicator variables can be included in a
regression model
= 0 + 1 1 + 2
The effect of an indicator variable is to change
the level (in
Methods of Applied
Statistics I
STA442 / STA2101
Craig Burkett
1
About me
Craig Burkett
Formerly
Aerospace engineer
High-school teacher
Lecturer at UBC
Lecturer at UTM
Now a lecturer at U of T and a statistical
consultant
2
Logistics
Lecture 2-5pm every F
General Linear Models
with General Linear
Hypothesis Tests
and Likelihood Ratio Tests
1
Background
Linear combinations of Normals are Normal
~ , ~ ,
A sum of squared, standardized Normals is
Chi-square
~
Regression estimates have a Normal distn
~ , 2