Reading & listening
Things that shaped how I think about doing research. Just what I’ve found interesting and useful.
How research is and could be done
Nullius in Verba
This is a great podcast on “What science is, and what it could be”. Smriti and Daniël discuss topics such as integrity, values, replicability, open science, and much more.
Critical Metascience
On his substack, Mark Rubin discusses topics such as “theory aversion”, “open science”, and “preregistration”. Great discussions and reflections on central concepts in metascience!
Cargo Cult Science
A great talk on scientific integrity and the risks of doing research ‘wrong’ and fooling ourselves (and others). A famous line from his speech: “The first principle is that you must not fool yourself—and you are the easiest person to fool”. The analogy to the cargo cults on the islands in the southern part of the pacific is also great of course.
Science, Scholarship and intellectual virtues
Paper on how higher education should not be value free, and instead attempt to foster values that are internal to the sciences. E.g., love of truth (discovering it is the point of our efforts, and because knowing the truth matters), and wisdom (balance, because virtues may be conflicting)
Methods and inference
The Effect
Nick explains complex concepts clearly with examples, illustrations, and code. The effect consists of the main principles of quantitative methods, i.e. descriptives and regression, and on causal principles such as understanding the data generating process, identification and the more specific causal models and designs.
Causal Inference: The Mixtape
A great alternative to The effect by Huntington-Klein. Scott Cunningham delves a little more into the nitty gritty with more statistics and equations than what Nick does in his book. Anyhow, it’s a great book on what causal inference is, and a toolbox consisting of designs such as IV, DiD, RDD.
Statistical Rethinking
This is really a course with recorded videos, and a book. I have only seen some of his YouTube videos on causal inference (not on Bayesian statistics). He talks about The Golems of Prague (statistical models are stupid), causal inference, and how science needs to come before statistics. His video from a 3h workshop is a good start!
Improving Your Statistical Inferences
In his book, Daniël Lakens helps us understand, for instance, P-values and how to interpret confidence intervals correctly. The section about philosophical approaches to p-values is a nice soft introduction! Towards the end he also presents some metascientific concepts, such as integrity.
The book of why
The book of why is a great introduction to causality and understanding what we really ask when we ask causal questions, and how we can approach them. This is not a ‘statistical toolbox’, but instead a more abstract way of approaching the topic.
Not available for free:(