I am doing a regression with many dummy variables as controls. They all have high skews (>1 or <-1). To reduce the skew I generated logs of each variables but when I do so and run the command 'summarize var, detail' of the logged version, the new variable has zero for all percentiles, variance, SD and mean and the skew is unavailable. Why could this be the case?
Related Posts with Taking the log of dummy variables is setting everything to zero, why?
too many signficant variablesHello I am doing a multilevel regression using PASEC2014 data at the first level I included several …
Placebo test using an RD designHi all, I have a question regarding the performance of a placebo test when using a regression disco…
Placebo test using an RD designHi all, I have a question regarding the performance of a placebo test when using a regression disco…
Apply values across all in the same groupHi all, I have a dataset of about 10k participants across 600 groups (ParentSiteID in the dataset). …
Plotting weights in meta-analysisHello, I need to plot the study weights in forest plot. I am using the following command to create …
Subscribe to:
Post Comments (Atom)
0 Response to Taking the log of dummy variables is setting everything to zero, why?
Post a Comment