2

Is it possible to run a hypothesis test to test for a significant difference in calculated coefficients of the same independent variable in two different subsets of one population, after you have performed a LASSO selection process on each subset individually. (Assuming the variable is selected as relevant in both subsets)

For example, if I had run a LASSO regression for Males in a population and then for Females in a population, could I then test for a statistically significant difference in any of the independent variables selected in both groups? If possible, how should one proceed with this in R?

kjetil b halvorsen
  • 63,378
  • 26
  • 142
  • 467
Green90
  • 45
  • 2
  • Possible duplicate of [Using regularization when doing statistical inference](https://stats.stackexchange.com/questions/291622/using-regularization-when-doing-statistical-inference) – kjetil b halvorsen Apr 07 '18 at 09:16

0 Answers0