Computational methods and programming
If used effectively, modern computers boost the variety of available methods as well as productivity.
My main language is R, but I also offer support for JMP® and other software packages. I use software to enable analyses that would have been (almost) impossible to conduct a decade or two ago, but also to automate routine yet complex analyses to take the burden from the researcher along with the risk of making mistakes.
Computationally intense methods
Nowadays’ computing power allows for effective usage of methods requiring many iterations or replications. These methods are particularly useful if classical assumptions (e.g. normally distributed error terms) cannot be made or when classical approaches do not provide a useful solution.
Computationally intense alternatives ▼
- Randomization or permutation tests
- Bootstrapping
- Simulations
Webtool development
Running the same analyses over and over again, repetitively clicking through the menus of the software of your choice? Tailor-made webtools for complex analyses or even extended reports can significantly increase your productivity and reduce the risk of mistakes.
I use the R package shiny to develop customized webtools for routine yet complex analyses.
Considerations ▼
- Specification of user requirements
- Development of tools against these specific needs
- Validation
- Deployment – locally for individual use or based on an external server with authentication
I also develop customized workflows in JMP®.