Hi everyone!

I am looking for some advice on how to categorize into categorical variables of different magnitudes based on the cut-points where the observed distribution crosses over the Gaussian normal distribution.

Just like the following graph, the cut-points are on -82, -21, 16 and 96.
From: Park, A. Y., & Sapotichne, J. (2019). Punctuated equilibrium and Bureaucratic autonomy in American city governments. Policy Studies Journal, 48(4), 896-925. doi:10.1111/psj.12333

Array

I want to categorize into categorical variables by the following picture. Are there any solutions for me to get the value of the cut-points?
Array


Thanks,
May