Timeline for Sample drawn from a Cauchy distribution; is it possible to do a t-test?
Current License: CC BY-SA 3.0
9 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Sep 16, 2015 at 1:06 | answer | added | Glen_b | timeline score: 5 | |
| Sep 12, 2015 at 2:59 | history | tweeted | twitter.com/#!/StackStats/status/642532986433478656 | ||
| Sep 12, 2015 at 0:31 | comment | added | Glen_b | Of course it is possible to perform a t-test -- you take the sample values and plug them in. It's not like the t-statistic becomes impossible to calculate The question of whether it means anything is different; it certainly doesn't test the hypothesis of zero-mean. You could use it to test a hypothesis of zero center of symmetry (NB if you use the ordinary t-tables you'll have to put up with a much lower significance level than the nominal), but it wouldn't be my first choice for a test of location for the Cauchy. There are tests with better power properties across a range of distributions | |
| Sep 11, 2015 at 18:00 | answer | added | Aksakal | timeline score: -1 | |
| Sep 11, 2015 at 17:38 | answer | added | RUser4512 | timeline score: 8 | |
| Sep 11, 2015 at 17:28 | history | edited | emmy | CC BY-SA 3.0 |
edited title
|
| Sep 11, 2015 at 17:28 | comment | added | Xi'an | The Cauchy has no mean. | |
| Sep 11, 2015 at 17:21 | review | First posts | |||
| Sep 11, 2015 at 17:22 | |||||
| Sep 11, 2015 at 17:19 | history | asked | emmy | CC BY-SA 3.0 |