I am looking for a way to compare coefficients obtained from quantile regression.
The two surveyed models are nested, estimated on the same sample and for the same quantile.
$$
Y = \beta_1X+\epsilon_2\\
Y = \beta_2X+\gamma Z + \epsilon_1
$$
Given the estimates and the respective standard errors I want to know if $\beta_1$ and $ \beta_2$ differ significantly.
All I could find up to now are tests for coefficient differences between regressions for different quantiles or for different groups (samples).
$\begingroup$
$\endgroup$
1
-
$\begingroup$ Regarding your last sentence, could you perhaps contribute to stats.stackexchange.com/questions/175587 and stats.stackexchange.com/questions/124796? $\endgroup$– Richard HardyCommented Sep 6, 2019 at 7:47
Add a comment
|