Learning objectives:
a. Use and explain the basic concepts used in epidemiology
b. Understand mechanisms giving rise to missing data
c. Describe key assumptions and apply imputation methods to deal with missing data
d. Know when to apply a mixed model in practice
e. Be able to perform mixed model analyses using statistical software (R)
f. Understand the advantages, limitations and key characteristics of IPD-MA in intervention, diagnostic and prognostic research
g. Understand the relevance of between-study heterogeneity, and be familiar with statistical methods for investigating and reporting this
h. Be familiar with statistical methods for summarizing relative treatment effects, and for developing and validating clinical prediction models using IPD from multiple sources

Assessment
Exams (midterm and final).

Each exam counts for 50% of the final grade, and both exams will consist of two parts (making up 12.5%, 37.5%, 25% and 25% of the final grade).

To pass the course, student’s final grade must be greater than or equal to 5.5, and the grade for each of the two exams must greater than or equal to 4.
Students with final grades between 4.0 and 5.4 will have a chance to attend one or both of the retake exams.
To be able to attend the retake exams, your grades for each of the two exams should be greater than or equal to 4.

Prerequisities
INFOMDWR Data Wrangling.

This course is for students in the master Applied Data Science only.

This course provides insight into the basic principles used in epidemiology, such as bias and confounding and students will learn statistical methods to address missing data and correlated data.
Possible mechanisms for data being missing, their potential impact in terms of bias, and methods to handle missing data will be discussed.
Correlated data may occur because response variables are observed more than once per individual, or when there is a hierarchical (multilevel) structure in the data, e.g. patients within hospitals, pupils within classrooms, etcetera.
Mixed models are one way of analyzing this kind of data.
Systematic reviews and meta-analyses are methods to summarize published aggregate data, but it is increasingly common that individual participant data (IPD) are obtained from multiple primary studies, leading to “big data”.
Meta-analysis involving IPD, a special application of mixed models, will therefore also be discussed.
Emphasis will be on meta-analysis for interventions, though it is also possible to use meta-analysis to investigate the accuracy of diagnostic tests, to develop clinical prediction models, and to externally validate such models.

Students will apply all these methods in R during computer labs and assignments

Course form
Lectures, tutorials, practicals.

Study materials
• Lecture notes Epidemiology and Big Data
• Reader
• Data sets for assignments and case study