P-value discourse seems to focus on the hackable nature of significance testing. But shouldn't the bigger point that p-values are necessary but insufficient criteria? Being distinguishable from noise is the lowest possible form of scientific rigor.
Also weighting data vs. analysis. As Alan Baddeley (memory researcher) said: "I resolved that if I were to develop a theory, it would be based very closely on the evidence, which would survive even if the theory proved totally wrong."
cognitive psychology. PhD