Some random observations today:
* So much of evaluating students is based on random luck; grades are often based on exam performance. But a good student can have a bad day, or even have guessed wrong on exactly how to optimize study time. And vice versa for a less good student. I guess that is why you should look to consistency in the student’s record.
* The same is probably also true in terms of evaluating teaching performance.
* The above statements are also true for reading other researcher’s papers or seeing their seminars. You observe different people infrequently; it therefore is difficult sometimes to get the entire record. Hence the importance of the advisor in hiring new faculty, especially faculty who have just finished their degrees.
* It is difficult to measure academic output. Working on important problems is a good sign, if you are making progress. For many researchers, the risk/return tradeoff is hard: solving tough problems has high value, but might generate little intermediate output. That leads to ‘counting lines on the vita’ and ‘salami slicing’ research strategies.
*Different researchers have different strengths. There is no one dominant strategy for everyone.
(Gosh I sound pompous today. But I need to empty out the old noggin.)