Skip to main content
9 events
when toggle format what by license comment
May 9, 2022 at 1:47 vote accept ML33M
May 8, 2022 at 14:58 answer added EdM timeline score: 1
May 8, 2022 at 14:42 comment added Christian Hennig If you know that data are truly distributed log-normally, you can log-transform your data and then normal theory (as probably used for confidence intervals; I don't know how these are computed for that particular score) will apply. However in a real situation you will not know that, and normal theory may be fine without transformation if data don't look very asymmetric (e.g., when concentrated far away from zero).
May 8, 2022 at 7:42 comment added ML33M @ChristianHennig I have made some changes and now showing what I had in my head. It is more like a question if all that matters in the calculation and final interpretation of data. I hope this make sense now :)
May 8, 2022 at 7:40 history edited ML33M CC BY-SA 4.0
added 664 characters in body
May 7, 2022 at 10:07 comment added Christian Hennig Your title mentions confidence intervals but your actual question doesn't. Is there some background to your question that you may want to add, that brings them in?
May 7, 2022 at 10:05 answer added Christian Hennig timeline score: 2
S May 7, 2022 at 6:23 review First questions
May 7, 2022 at 10:04
S May 7, 2022 at 6:23 history asked ML33M CC BY-SA 4.0