Statistical and epidemiological methods 2

The course requires basic knowledge of quantitative methods in medicine and health sciences. The course will introduce more advanced statistical methods and epidemiological concepts. The teaching will emphasize correspondence between research questions, study planning and design, data collection and analytical approach. A central topic will be performing and interpreting multivariable regression analyzes to estimate causal relationships, including linear and logistic regression, as well as survival analyses. Topics such as handling missing data, random and systematic errors, effect modification and non-parametric tests are also covered.

Knowledge:

After completing and passing the course, the student should:

• Have in-depth knowledge of various regression models and basic methods for assessing effect modification
• Have knowledge of different methods for survival analyses
• Have in-depth knowledge of systematic errors related to confounding, selection bias and measurement error, and have knowledge of statistical methods to control for confounding
• Have in-depth knowledge of random errors, with a focus on the use and interpretation of confidence intervals and p-values.
• Have knowledge of the terms association, prediction and causality

Skills:

After completing and passing the course, the student should be able to:

• Assess the consequences of systematic errors
• Summarize results of statistical analyzes for handling confounding
• Perform and interpret simple and multiple regression analyses
• Perform and interpret simple analyzes of lifetime data

General competence:

After completing and passing the course, the student should be able to:

• Critically assess various studies with a focus on study design, effect size and precision, as well as sources of systematic errors
• Apply knowledge and skills in statistical and epidemiological methods to carry out simple research projects

For most topics, introductions will be given, followed by practical exercises in Stata/SPSS, partly with ready-made syntax, but this will vary somewhat between the different topics.