This one is directed at my Readers who have supervised graduate students through the PhD in their laboratory.
Have you ever, at any point, thought that you should not train more than replacement value (n=1)?
If not, why not? What influences in your life have shaped your decision to train graduate students? What is YOUR purpose here?