Significant Results
From MgmtWiki
Definition
Selecting data to prove a result that can get a scientific paper published.
Context
Reputational Crisis
Problems
Statistical Inference = hacking P-Values or confidence intervals
- Fisher said "No scientific researcher will ever have the same level of significance from experiment to experiment"
- Neyman-Pearson emphasized the role of judgement
- Baysian is the only way to get real numbers.
- Nunally has a more valid book
Moral courage, Philip Zimbardo who is famous for his stain for prison experiment, and two colleagues, asked theri participants "do you feel that there is a difference between altruism and heroism?"
Solution
- Stop funding so many post-doc researchers in colleges.