Timeline for Interpretation of non-significant results as "trends"
Current License: CC BY-SA 4.0
11 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jan 4 at 23:55 | history | edited | Ben Bolker | CC BY-SA 4.0 |
added 12 characters in body
|
| Dec 11, 2023 at 1:09 | history | edited | Ben Bolker | CC BY-SA 4.0 |
added 7 characters in body
|
| Jul 8, 2019 at 6:53 | comment | added | amdex | Thanks for the great response. I agree that the discrepancy between the actual participants and the population is what led to the confusion. | |
| Jul 8, 2019 at 6:53 | vote | accept | amdex | ||
| Jul 7, 2019 at 18:05 | comment | added | Dave | @boscovich Sure, I was talking in absolutes when we’re doing statistics, but I think of it as an insignificant p-value meaning that you really haven’t a clue what’s happening with the population. At least with a significant p-value you have reached some established threshold of evidence to suggest that you know something. But definitely it’s possible to get a significant p-value when it’s misidentified the direction. That error should happen from time to time. | |
| Jul 7, 2019 at 17:57 | comment | added | boscovich | @Dave: even if the presence of "statistical evidence" (statistically significant p-value?), "the observed differences may very well be the opposite of what’s going on with the population" | |
| Jul 5, 2019 at 23:51 | comment | added | Dave | +1 I think many people fail to realize that in the absence of statistical evidence, the observed differences may very well be the opposite of what’s going on with the population! | |
| Jul 5, 2019 at 23:00 | comment | added | Ben Bolker | thanks, partially incorporated (trying to balance brevity/clarity and accuracy ...) | |
| Jul 5, 2019 at 23:00 | history | edited | Ben Bolker | CC BY-SA 4.0 |
added 61 characters in body
|
| Jul 5, 2019 at 15:34 | comment | added | Isabella Ghement | Beautiful response, Ben! I wonder if your second example statement could be modified for clarity to reflect the gist of the first example statement: "although group A used X 13% more often than group B IN OUR EXPERIMENT, the difference IN USAGE OF X BETWEEN GROUPS IN THE GENERAL POPULATION was not clear: the plausible range OF THAT DIFFERENCE went from A using X 5% less often than group B to A using X 21% more often than group B." | |
| Jul 5, 2019 at 8:11 | history | answered | Ben Bolker | CC BY-SA 4.0 |