Timeline for How do you classify based on percentile ranking when most scores are the same?
Current License: CC BY-SA 3.0
9 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| S Jun 12, 2015 at 17:36 | history | bounty ended | tumultous_rooster | ||
| S Jun 12, 2015 at 17:36 | history | notice removed | tumultous_rooster | ||
| Jun 12, 2015 at 17:36 | vote | accept | tumultous_rooster | ||
| Jun 7, 2015 at 21:42 | answer | added | MrMeritology | timeline score: 2 | |
| Jun 5, 2015 at 18:14 | comment | added | Nick Cox | Much of the point of percentiles is to make precise where some value falls in a distribution. This should work best when ties are absent (although it has limitations even then) and, as you point out, is stretched mightily by ties. I can't say that any analysis will identify precisely when it becomes absurd; that's a matter depending on practical purposes. What you should tell testees is (ideally) just the frequency distribution when ties are extreme. | |
| Jun 5, 2015 at 18:07 | history | edited | Nick Cox | CC BY-SA 3.0 |
added 2 characters in body
|
| S Jun 5, 2015 at 18:04 | history | bounty started | tumultous_rooster | ||
| S Jun 5, 2015 at 18:04 | history | notice added | tumultous_rooster | Draw attention | |
| Jun 3, 2015 at 17:56 | history | asked | tumultous_rooster | CC BY-SA 3.0 |