-
@ Low Information Voter
2025-05-25 21:51:52
Doctor Malone's advice is spot on.
But researchers have unexpected results fairly often.
Options:
(a) Full-speed to publication, damn the torpedos. Only an option if you have a VERY permissive source of funding, or you're so high profile that THEIR funding depends on you.
(b) Include in results for your peers to see, but say the expected in the summary.
(c) As above, but hide them in an appendix
(d) "Massage" data, remove outliers, transform by various functions. The effect probably won't go away, but you can reduce its statistical significance to the point that it becomes debatable.
(e) Lie. Best for funding, but you risk your professional career.
(f) Don't publish. Not a great option for funding, actually.