黃河泉 发表于 2020-8-12 07:55 
存在,请将指令与节果发出来,以供判断!
老师,麻烦您帮我看下命令和结果,我的门槛变量就是核心解释变量,Threshold effect test的p值小于0.1,说明存在单一的门槛值19.8652,小于19.8652时核心解释变量x增加1%y增加0.029%,大于19.8652时核心解释变量x增加1%y下降0.13%。这样解释对吗?但是系数0.029的p值大于0.1了,这说明什么呢?
. xthreg y2 hycwgg1 hhia , rx(x) qx(x) thnum(1) trim(0.05) grid(400) bs(1000)
Estimating the threshold parameters: 1st ...... Done
Boostrap for single threshold
.................................................. + 50
.................................................. + 100
.................................................. + 150
.................................................. + 200
.................................................. + 250
.................................................. + 300
.................................................. + 350
.................................................. + 400
.................................................. + 450
.................................................. + 500
.................................................. + 550
.................................................. + 600
.................................................. + 650
.................................................. + 700
.................................................. + 750
.................................................. + 800
.................................................. + 850
.................................................. + 900
.................................................. + 950
.................................................. + 1000
Threshold estimator (level = 95):
-----------------------------------------------------
model | Threshold Lower Upper
-----------+-----------------------------------------
Th-1 | 19.8652 19.8158 19.9082
-----------------------------------------------------
Threshold effect test (bootstrap = 1000):
-------------------------------------------------------------------------------
Threshold | RSS MSE Fstat Prob Crit10 Crit5 Crit1
-----------+-------------------------------------------------------------------
Single | 2.72e+07 4792.9834 4.16 0.0920 3.7638 5.5495 14.9747
-------------------------------------------------------------------------------
Fixed-effects (within) regression Number of obs = 5680
Group variable: ID Number of groups = 568
R-sq: within = 0.0014 Obs per group: min = 10
between = 0.0339 avg = 10.0
overall = 0.0039 max = 10
F(4,5108) = 1.82
corr(u_i, Xb) = 0.0582 Prob > F = 0.1217
------------------------------------------------------------------------------
y2 | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
hycwgg1 | -2.080311 1.429647 -1.46 0.146 -4.883031 .7224099
hhia | -4.273057 23.00395 -0.19 0.853 -49.37065 40.82453
|
_cat#c.x |
0 | .0286919 .0627921 0.46 0.648 -.0944076 .1517914
1 | -.1298331 .0627065 -2.07 0.038 -.2527647 -.0069015
|
_cons | 3.36338 3.193659 1.05 0.292 -2.89756 9.624319
-------------+----------------------------------------------------------------
sigma_u | 22.013656
sigma_e | 72.940783
rho | .08348053 (fraction of variance due to u_i)
------------------------------------------------------------------------------
F test that all u_i=0: F(567, 5108) = 0.91 Prob > F = 0.9361