Multiverse analysis: a method to assess the robustness of effect estimates. By Judith Rijnhart, research associate at the department of Epidemiology and Data Science.
Multiverse analysis: a method to assess the robustness of effect estimates
There is an increasing awareness that replication should become common practice in empirical studies. However, replication studies are only useful if published study results are robust. The relatively new multiverse-analysis methodology assesses the robustness of effect estimates against arbitrary data analytical decisions, i.e., researcher degrees of freedom. Application of this methodology involves the identification of all reasonable data analytical decisions, and data analysis based on all combinations of reasonable data analytical decisions. In this tutorial I will provide a step-by-step explanation of the multiverse-analysis methodology and show its application to a real-life data example. In the workshop, participants will gain hands-on experience with performing a multiverse analysis. To get the most out of the workshop, participants are asked to bring a laptop with R and Rstudio software installed on it.
| Start date: | |
|---|---|
| End date: | |
| Time: | |
| End time: | |
| Location: | Agora 5 |
| Directions: | From the main entrance in the Main Building, you go left to to the 6-elevators. Go to the third floor and you'll see the Agoracomplex on the right side. |
| Costs: | Free |
|---|---|
| Register now |
| Organisor |
APH Methodology program |
|
|---|---|---|
| Contact | Simone van der Riet | aph.methodology@amsterdamumc.nl |