## Steps 1. Define the study units and population. 3. Select the response variable(s) you are going to measure. 4. List all possible explanatory variates which might explain variations in the response and organize them in some fashion. 5. Identify how much data is needed to provide a robust answer. 6. Decide how to measure the selected response and consider what measurement error that approach implies. 7. Define a protocol for sampling or collecting the data that minimizes sampling error and biases. 8. Identify where the raw data will come from and how much it will cost. 9. Consider how the ideal data would look and compare it to what you can expect to get. 10. Check for any ethical or legal concerns associated with this data or its collection. Next up is [[4. Get the data]]. Ref: [Scientific Method, Statistical Method and the Speed of Light by R. J. MacKay and R. W. Oldford, Statistical Science, 2000](https://projecteuclid.org/journals/statistical-science/volume-15/issue-3/Scientific-Method-Statistical-Method-and-the-Speed-of-Light/10.1214/ss/1009212817.full)