Language of instruction
The course provides general knowledge about common statistical methods that are applicable in epidemiological and health research. The target group is all PhD students at the Faculty of Medicine who have epidemiology or health sciences as an essential part in their thesis work.
Passed course in applied statistics I is required together with basic knowledge in the statistical package used in the same course.
Weeks 19-20 The course will be given in Lund or Malmö, if possible due to the covid situation. Any changes will be posted here and will also be communicated in the welcome letter for the course.
Weeks 45-46 in English - the course will be given at CRC, Malmö, if possible due to the covid situation. Any changes will be posted here and will also be communicated in the welcome letter for the course.
Malmö - Lecture room will be announced to the students.
The course will provide practical knowledge concerning the management and analysis of empirical data in the chosen epidemiological and health research project.
The course will also help the participants to be able to critically examine various types of empirical research in medical science, also outside their own field of research.
Content of the course
The course contains four content areas:
1) Design and analysis of studies of binary outcomes
- Cohort and case-control studies
- Incidence and risk calculations
- The measure of association, odds ratios, absolute and relative comparisons
- Simple survival analysis (time to event)
- Power calculation
2) Introduction to correlation and regression analysis
- Spearman rank and Pearson's correlation coefficient
- Simple linear regression
- Simple logistic regression
3) How to control for confounding effects
- Overview of the different approaches
- Overview of common regression models
- Using multiple logistic regression
4) Reliability analysis
- Association vs consistency
- Limits of agreement
- ICC for continuous data
- Kappa calculations for categorical data, introduction to more advanced methods
This course in applied statistics with specialization in Epidemiology (Public Health) and Health Sciences provides participants with the necessary tools to plan, analyze, and evaluate common types of observational studies in these research fields. Emphasis is placed on the use of regression methods as a way to manage confounding effects.
The course provides an overview of commonly used regression models in this context, but with primary focus on logistic regression analysis of case-control data. The course addresses the difference between correlation, association, and consistency, and also provides an introduction to reliability analysis of both quantitative measurements and qualitative assessment instruments.
Take-home examination. To pass the course it is not only required to have a passing grade on the examination, but also to have completed the group assignment, including active participation in the discussions concerning both their own group work as well as the work of other groups.
Vittinghoff E, Glidden DV, Shiboski SC, McCulloch CE. Regression methods in biostatistics. Springer, 2nd edition, 2012. Chapters 3.1-3.5, 4.1–4.4 & 5.1-5.4.
(The book is available as e-book at Lund University)
As a reference to the section on the design and analysis of binary outcomes (theme No. 1) the same book is used as in the course
Epidemiology I – Introduction till epidemiology:
Rothman KJ. Epidemiology – an introduction. Oxford University Press, 2nd edition, 2012.
The articles that deal with the topic of reliability analysis (theme No. 4) will be handed out at the start of the course.