Study Design: the best model for a cat is... a cat!
summary
Claim: the best model for a cat is a cat, so designs use real cats not proxies. Specifics include a controlled feeding experiment comparing wet vs dry vs both food weekly weights, observational case-control and cohort studies on outdoor cats with an intervention of fixing, and R tooling: pwr and sampling packages, SSC diagram from Six Sigma, NA handling and dummy coding in R, data.table and vectorization to generate and summarize sample datasets. Ethics: IRB/animal welfare and consent procedures are required.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 0:47Types of Data
- 5:25Observing
- 8:04CaseControl
- 9:26Fish Diagram
- 12:48Study Cohort
- 13:16Consent
- 14:57Conclusion
related talks
40mMachine learning 101: Where to begin?
01h 19mModel criticism with ArviZ
18mThe Importance of Modeling Data Collection
57mGood Practices for Applied Machine Learning
20mBetter Care and Feeding of Machine Learning Models
33mHidden Biases In Data: Moving Beyond Black Box Models
42mHow am I going to deal with all of these cats
31mA Gentle Introduction To Causal Inference