##### Child pages
• Statistical methods in medicine and epidemiology, spring 2014
Go to start of banner

# Statistical methods in medicine and epidemiology, spring 2014

### News

An error has been spotted (thanks to Tanja) in the answers to exercise 6. In task 1, the standard error for the relative

hazard should read 1/sqrt(D) (and not sqrt(D)). The numeric solution was nevertheless ok, i.e. based on the correct

expression. The corrected answer is now available on below.

The second course exam will take place in classroom C123 on Thursday, May 8, at 2 p.m. (- 5 p.m.). The exam covers the following

lectures, exercises and book chapters:

(A) Lectures 18-25; (B) Exercises 6-9 (see the note on R exercises below); (C) The following chapters from

the textbook: 12, 17.3-17.4, 19, 21, 22, 23, 25.4 - 25.6, 26.1-26.5.

N.B. You will not be required to write R code. However, there may be questions about how to interpret output

from statistical software (such as R). Ref: Hands-on exercises 7 and 8.

N.B. Lecture 14 contains some material for individually matched case control studies, included in the 2nd exam.

### Scope

5-10 sp. It will be possible to complete only the first part of the course (5 study points).

### Prerequisites

The course builds up the basic methods for statistical inference, along with applications to epidemiological study designs.

Some pre-knowledge of statistical inference is helpful. It is recommended that students with statistics as their major subject would

have completed the basic courses of statistical inference.

### Lectures

Weeks 4-9 and 11-18, Thursday 14-17 (CK111 & B210), after lectures exercises in class B120.

Easter holiday 17-23.4.

### Course content

The lectures largely follow the previous version, Fall 2012, content

### Exams

The second course exam will take place on Thursday, May 8 (C123, 2 p.m. sharp - 5 p.m.). The exam covers the following

lectures, exercises and book chapters:

(A) Lectures 18-25; (B) Exercises 6-9 (see the note on R exercises below); (C) The following chapters from

the textbook: 12, 17.3-17.4, 19, 21, 22, 23, 25.4 - 25.6, 26.1-26.5.

N.B. You will not be required to write R code. However, there may be questions about how to interpret output

from statistical software (such as R). Ref: Hands-on exercises 7 and 8.

N.B. Lecture 14 contains some material for individually matched case control studies, included in the 2nd exam.

The first course exam will take place on Thursday, March 13. The exam covers the following lectures, exercises and book chapters:

(A) Lectures 1-17 (i.e., lecture 18 on sample size is not included in the first exam); (B) Exercises 1-5; (C) The following chapters from the textbook:

1-10,11.1-11.5,11.7-11.8, 13.1-13.3, 14.1-14.3, 15.1-15.4,15.6-15.8, 16.1-16.5,16.7-16.8,17.1-17.2,18.1,18.3-18.4.

N.B. Specific regression models (although already referred to in some of the lecture material) are not included in the first exam.

### Bibliography

Clayton and Hills: Statistical Models in Epidemiology, Oxford University Press, 2013 (N.B. Older editions will be fine, too.)

### Lecture summaries

1. Lecture 1 (Binary model. Conditional probability. Bayes formula) (Chapters 1 and 2)
2. Lecture 2 (Likelihood) (Chapter 3)
3. Lecture 3 (Estimation of survival. Hazard function) (Chapters 4 and 5)
4. Lecture 4 (Time. Censoring. Competing risks.) (Chapters 6 and 7)
5. Lecture 5 (Approximating likelihood functions. Statistical paradigms) (Chapters 8, 9 and 10)
6. Lecture 6 (P values). 6b Lecture 6b (P values continue) (Chapter 11)
7. We continued with material from lectures 5 and 6
8. See the material from lectures 5 and 6
9. See the material from lecture 5 and 6
10. Lecture 10 (Chapter 13)
11. Lecture 11 (Chapter 14)
12. See the material from lectures 10 and 11.
13. Lecture 13 (Chapters 16-19)
14. Lecture 14 (Chapters 16-19)
15. See the material from lectures 13 and 14.
16. We continue with the material from last week (case-control studies)
17. Case-control studies cont.
18. Lecture 18 (Chapter 21)
19. Lecture 19 (Poisson regression)
20. Lecture 20 (Logistic regression)
21. Lecture 21 (Hypothesis testing in regression models. Interaction. Deviance. Individually randomised case-control studies.)
22. Lecture 22 (Model choice. More about interaction.)
23. Lecture23 (Small samples)
24. Fisher exact test (example). Article 3
25. Article1 Article 2

### Data sets

1. The data set for task 2/2: atresia.dat(4 columns: individual ID, age at diagnosis, age at death or censoring,

failure indicator (1=death/0=censoring); the ages are given in days).

2. BCG.dat
3. BCGcasecontrol.dat
4. BCGmatched.dat
5. CHD.dat
6. doseresponse.dat

### Errata

Lecture 3, slide 25: The notation for S(t_i) on the last line is erroneous and not needed here.

### Registration

Did you forget to register?  What to do?

RyhmäPäiväAikaPaikkaPitäjä
1.Thursday 16 (or 17)-19C128Christopher Solomon
• No labels