Timeline for Can I trust a significant result of a t-test if the sample size is small?
Current License: CC BY-SA 3.0
14 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Mar 7, 2017 at 23:33 | comment | added | Dave Harris | Just a comment, I don't want to add to the wonderful comments below; you do not trust the result of a t-test, you trust the procedure itself. An individual result is either correct or incorrect, but without further investigation, you will never know which. A t-test in either Fisher's methodology or Pearson and Neyman's methodology is trustable if the assumptions are met. If you set $\alpha<.05$ then it will deceive you, upon infinite repetition, no more than 5% of the time, possibly quite a bit less. The question you should ask is "are the assumptions met?" | |
| Mar 7, 2017 at 20:04 | comment | added | Silverfish | Very closely related: Is there a minimum sample size required for the t-test to be valid? | |
| Mar 7, 2017 at 19:49 | history | edited | amoeba | CC BY-SA 3.0 |
more general and more explicit title
|
| Mar 7, 2017 at 19:45 | answer | added | Patrick B. | timeline score: 6 | |
| Mar 7, 2017 at 16:06 | vote | accept | Eric | ||
| Mar 7, 2017 at 15:47 | answer | added | Placidia | timeline score: 9 | |
| Mar 7, 2017 at 14:27 | history | edited | amoeba |
edited tags
|
|
| Mar 7, 2017 at 13:12 | answer | added | Denziloe | timeline score: 4 | |
| Mar 7, 2017 at 9:25 | review | Close votes | |||
| Mar 7, 2017 at 12:21 | |||||
| Mar 7, 2017 at 1:09 | history | edited | Michael R. Chernick | CC BY-SA 3.0 |
edited title
|
| Mar 7, 2017 at 1:04 | answer | added | Glen_b | timeline score: 16 | |
| Mar 7, 2017 at 0:35 | answer | added | Jim Lewis | timeline score: 22 | |
| Mar 6, 2017 at 23:34 | answer | added | Hugh | timeline score: 17 | |
| Mar 6, 2017 at 23:13 | history | asked | Eric | CC BY-SA 3.0 |