Interpretation of regression with Box Cox transformed Y (square root Y, cube root Y etc)

Box Cox applied to Y in a regression can make the interpretation of the model harder. Often the interest is on Y not the transformed Y. This video explains why a back transformation method that takes you from, say, a square rooted, or cube rooted Y to Y in terms of X is wrong. Examples are given when the transformed Y is meaningful. A bit exasperated by a number of textbooks that simply warn to round lambda to the nearest interpretable value, but not explain what that meant, I had to make this video.

0:32 right to way interpret Box Cox transformed regression 1:06 examples of when transformed Y has an interpretation 1:41 when it doesn't matter whether to work with transformed Y 1:54 a wrong way to interpret a Box Cox regression 2:37 references to do a back transformation on Box Cox

Main presentation of Box Cox https://youtu.be/zYeTyEWt7Cw ... https://www.youtube.com/watch?v=vzyW-JhfJ4g