Good morning to everyone,
I am analyzing the gender wage gap and I noticed that differentials are different if I compute raw calculations (with salaries as they are) or if I use their logarithmic form for more complex analyses.
I know that the logarithmic transformations normalized distributions with higher skewness and hence change the "mean allocation" by groups depending on the outliers.
But I was thinking about how can I justify this divergence?
For example, If I need to communicate to a firm with a raw salary gap of 4% and a gap in log(wages) of 1% . How I can simply justify this divergence and the consistency of results?
Many thank in advance for your time, wishing you a great rest of the week
I am analyzing the gender wage gap and I noticed that differentials are different if I compute raw calculations (with salaries as they are) or if I use their logarithmic form for more complex analyses.
I know that the logarithmic transformations normalized distributions with higher skewness and hence change the "mean allocation" by groups depending on the outliers.
But I was thinking about how can I justify this divergence?
For example, If I need to communicate to a firm with a raw salary gap of 4% and a gap in log(wages) of 1% . How I can simply justify this divergence and the consistency of results?
Many thank in advance for your time, wishing you a great rest of the week
