Complement your statistical skills with expert methods in R and Stata
Learn to programme two statistical packages in order to use advanced methods that complement the statistical techniques taught on our other modules. In each of the two statistical packages, students begin with 20 essential commands and progress towards computer-intensive statistical methods such as simulation, advanced regression modelling techniques, multiple imputation, cross-validation and bootstrapping. Led by Dr Jason Oke, senior statistician in the NDPCHS statistics group, an experienced teaching team guide students from the basics to advanced topics in R and Stata.
This course is delivered and assessed wholly online over an intensive 8 weeks.
The last date for receipt of complete applications is 5pm Friday 7th April 2023. Regrettably, late applications cannot be accepted.
Aims of the module
The overall aims of this module are to enable students:
- To gain confidence in two high-level professional statistics packages, and complement the techniques learnt on other modules with advanced techniques such as multiple imputation to overcome missing data.
Intended learning outcomes are:
- Learn fundamental programming techniques such as loops, and apply them in contexts such as Monte-Carlo simulation power calculations.
- To develop the ability to complement the techniques learnt on other modules with computer-intensive techniques such as multiple imputationand resampling methods such as the bootstrap.
Students should leave the course with confidence that in the future they could manage challenging datasets with state-of-the-art R packages; address missing data with multiple imputation; use simulation to evaluate statistical power, or check model assumptions; use bootstrap and permutation methods to calculate confidence intervals and p-values in non-standard situations; apply multi-level statistical models to bigger data sets, in which each individual contributes repeated outcome measurements.
Students will also have an introductory view of Bayesian statistical modelling; an overview of statistical learning methods (“machine learning”, or “algorithms” in the popular press); and the ability to understand when each of these might be useful for a problem in evidence-based medicine.