Good Day,
I am currently using Dynamic Panel Data Threshold regression to estimate the model. A command for the estimation is xthenreg.
Here is my results:
1. first model without variable control
xthenreg SR WUI, endogenous (WUI)
N = 10, T = 38
Panel Var. = id
Time Var. = period
Number of moment conditions = 1368
------------------------------------------------------------------------------
SR | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
Lag_y_b | .6178324 5.727219 0.11 0.914 -10.60731 11.84298
WUI_b | 423.7773 1977.905 0.21 0.830 -3452.845 4300.4
cons_d | -92.59644 626.7891 -0.15 0.883 -1321.08 1135.888
Lag_y_d | -1.314451 3.582833 -0.37 0.714 -8.336674 5.707772
WUI_d | -104.9881 3109.682 -0.03 0.973 -6199.853 5989.876
r | .2102576 1.620444 0.13 0.897 -2.965754 3.386269
------------------------------------------------------------------------------
2. Second model with variabel control EPU.
xthenreg SR WUI EPU, endogenous (WUI)
N = 10, T = 38
Panel Var. = id
Time Var. = period
Number of moment conditions = 1404
------------------------------------------------------------------------------
SR | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
Lag_y_b | 1.080042 3.677588 0.29 0.769 -6.127898 8.287983
EPU_b | -.4016385 .5782757 -0.69 0.487 -1.535038 .731761
WUI_b | 665.8928 1338.011 0.50 0.619 -1956.561 3288.346
cons_d | -13.12387 112.6 -0.12 0.907 -233.8158 207.568
Lag_y_d | -1.847607 2.828002 -0.65 0.514 -7.39039 3.695176
EPU_d | .4576925 .4374148 1.05 0.295 -.3996248 1.31501
WUI_d | -641.8407 1292.78 -0.50 0.620 -3175.642 1891.961
r | .0747121 .4422641 0.17 0.866 -.7921095 .9415338
------------------------------------------------------------------------------
Here is my questions:
1. For the first model, Does r coeficient represent a threshold for variable WUI?
2. For the second model, Does r coeficient represent a threshold for variable WUI and EPU? How can I find the treshold coeficient when I employ threshold regression with variable control like model 2?
3. If the r does not represent the threshold coeficient for the transition variable, how should I find the threshold coeficient?
Thank you
I am currently using Dynamic Panel Data Threshold regression to estimate the model. A command for the estimation is xthenreg.
Here is my results:
1. first model without variable control
xthenreg SR WUI, endogenous (WUI)
N = 10, T = 38
Panel Var. = id
Time Var. = period
Number of moment conditions = 1368
------------------------------------------------------------------------------
SR | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
Lag_y_b | .6178324 5.727219 0.11 0.914 -10.60731 11.84298
WUI_b | 423.7773 1977.905 0.21 0.830 -3452.845 4300.4
cons_d | -92.59644 626.7891 -0.15 0.883 -1321.08 1135.888
Lag_y_d | -1.314451 3.582833 -0.37 0.714 -8.336674 5.707772
WUI_d | -104.9881 3109.682 -0.03 0.973 -6199.853 5989.876
r | .2102576 1.620444 0.13 0.897 -2.965754 3.386269
------------------------------------------------------------------------------
2. Second model with variabel control EPU.
xthenreg SR WUI EPU, endogenous (WUI)
N = 10, T = 38
Panel Var. = id
Time Var. = period
Number of moment conditions = 1404
------------------------------------------------------------------------------
SR | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
Lag_y_b | 1.080042 3.677588 0.29 0.769 -6.127898 8.287983
EPU_b | -.4016385 .5782757 -0.69 0.487 -1.535038 .731761
WUI_b | 665.8928 1338.011 0.50 0.619 -1956.561 3288.346
cons_d | -13.12387 112.6 -0.12 0.907 -233.8158 207.568
Lag_y_d | -1.847607 2.828002 -0.65 0.514 -7.39039 3.695176
EPU_d | .4576925 .4374148 1.05 0.295 -.3996248 1.31501
WUI_d | -641.8407 1292.78 -0.50 0.620 -3175.642 1891.961
r | .0747121 .4422641 0.17 0.866 -.7921095 .9415338
------------------------------------------------------------------------------
Here is my questions:
1. For the first model, Does r coeficient represent a threshold for variable WUI?
2. For the second model, Does r coeficient represent a threshold for variable WUI and EPU? How can I find the treshold coeficient when I employ threshold regression with variable control like model 2?
3. If the r does not represent the threshold coeficient for the transition variable, how should I find the threshold coeficient?
Thank you
