0

I am trying to replicate a research paper as part of my Applied Econometrics course, and I came across a particularly vague statement in the reference paper.

"Following Malmendier and Tate (2005), we take the natural logarithm of the variable and add a constant of one for all variables to avoid outliers instead of following other approaches (e.g., winsorizing or trimming) without discarding information."

I could not find the relevant section in Malmendier and Tate (2005), and the sources I consulted online do not say much about dealing with outliers without discarding information using logistical transformations. Furthermore, some of the explanatory variables in my panel data have both negative and positive values. How do I deal with this situation?

Helix123
  • 1,467
Madhav Bajaj
  • 1
  • 1
    The log-plus-one or "logp1" or some other constant is occassionally added so that no special case is needed for values that are exactly (or close to) 0. The result is still a monotonic transformation. There's a few ways to go about it. – AdamO Apr 28 '22 at 22:38
  • 2
    There are plenty of questions with good answers about this here on cross-validated, e.g., https://stats.stackexchange.com/questions/30728/how-small-a-quantity-should-be-added-to-x-to-avoid-taking-the-log-of-zero – Helix123 May 01 '22 at 12:44

0 Answers0