Timeline for Why is "statistically significant" not enough?
Current License: CC BY-SA 3.0
5 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Apr 13, 2017 at 12:44 | history | edited | CommunityBot |
replaced http://stats.stackexchange.com/ with https://stats.stackexchange.com/
|
|
| Dec 11, 2013 at 5:57 | vote | accept | Jim Von | ||
| Dec 11, 2013 at 5:57 | |||||
| Dec 11, 2013 at 5:30 | comment | added | gung - Reinstate Monica | What I mean is, presumably there was some legitimate reason to run the study in the 1st place. Current theoretical knowledge &/or recent studies suggested your hypothesis might be true. Your hypothesis isn't likely to "lead to meaningless data analysis" unless it is incoherent. Potentially interesting surprises / features of your data could very well be discovered post-hoc; the fact that they are surprises implies you didn't know they would occur when you planned the study. The issue regarding "post-hoc" is whether to believe the surprises--they need to be confirmed by future research. | |
| Dec 11, 2013 at 5:05 | comment | added | Jim Von | You mean the hypothesis should be reasonable? And how to judge whether my hypothesis will lead meaningless data analysis? “Potentially interesting surprises” should be revealed by Post-hoc? | |
| Dec 11, 2013 at 4:53 | history | answered | gung - Reinstate Monica | CC BY-SA 3.0 |