Limited access to biostatistics expertise and statistical software is a major limitation for scientific use of registry data in Norway. Norwegian researchers often use outdated and overly simplistic statistical methods, which do not allow for adequate control of selection bias and confounding.
The main object of WP 2 is to develop new statistical methods and software specifically designed for health registries. The results will be communicated to the research community through peer reviewed publications in international journals. The work package will also offer courses and guidance to researchers.
Contact information
Rolv Terje Lie (leader WP 2)
Øystein Ariansen Haaland (WP2)
Magne Solheim (WP2)
Workshop on causal inference in health registry research
cOURSE IN REGISTER BASED EPIDEMIOLOGY IN COLLABORATION WITH NORDFORSK
Resources
BIOS
Statistical support service is available at the core facility BIOS.
Documentation and instructional material
Introduction
- Basic concepts in statistics (Part 1)
Covers random samples, types of variables, measures of central tendency.
- Basic concepts in statistics (Part 2)
Covers measures of dispersion and box plots.
- Statistical inference
Covers the normal distribution, central limit theorem, testing, confidence intervals, false positives and false negatives, and statistical power.
Data management in registry research
- Part 1
- Part 2
- Cover missing and duplicate IDs, missing data, contradictory information, impossible values and poor quality
Significance vs. relevance, fishing and multiple estimates
- Part 1
- Part 2
- Cover missing and duplicate IDs, missing data, contradictory information, impossible values and poor quality