Normal distributions and log transformations

15 Views Asked by At

Lets assume I have a origin distribution, lets call it Omega, which is heavily skewed and does not seem to be normally distributed.

I now apply a log(n+1) function to it and get a normal distribution with little to no skew, lets call it Beta.

My question is: When I calculate confidence intervals for Beta, can I say something like "Point number 312 in Beta lies within the 95% confidence interval, so it can be said that point 312 in Omega also lies within a 95% confidence interval!"?

Can I assume Omega is normally distributed, because a transformation of the distribution is?

If I can not make those assumptions, which assumptions can be made about Omega which show in Beta and its confidence intervals?

Sorry probably these are very basic questions, but I often fail to grasp the meaning of very mathematical topics which are often described in a very formal way.

Best regards