Today was the orientation for the PhD program in Statistics at UC Davis. Professor Debashis Paul gave a few words of advice that are worth writing down. It went along these lines:

There are two components of knowledge when solving problems in statistics - conceptual and technical. Conceptual means the broad, overarching ideas of say a most likely estimator, while technical means the details of the mathematical techniques applied, say integration by parts. You will run into problems that you can’t solve right away, but later on you figure them out. Take the time to understand why you couldn’t solve it right away. What was lacking that prevented you from seeing the solution? Then you know what you need to work on.