epidemiology

Data manipulation, visualization, and reproducible documents with R and the Tidyverse

Recent developments by the R community have revolutionized the data analysis pipeline in R, from manipulating and visualizing data to communicating results. Our workshop will provide hands-on training in tools from the tidyverse ecosystem, using real …

When interaction is not interaction: confounding and measurement error

Last week, I presented ggdag at JSM in Vancouver. As you can imagine, I had a lot of conversations with people about DAGs, confounding, colliders, and all the types of bias that can arise in research. One strange type of bias came up a couple of times that I don’t see discussed very often: measuring either the effect you are studying (x) or a variable along a confounding pathway (z) incorrectly can make it appear as if there is an interaction between x and z, even if there isn’t one.