求助关于门限回归的问题 - Stata专版

0关注
0粉丝

小学生

64%

还不是VIP/贵宾

-

0%

威望: 0 级
论坛币: 10 个
通用积分: 0
学术水平: 0 点
热心指数: 0 点
信用等级: 0 点
经验: 175 点
帖子: 8
精华: 0
在线时间: 5 小时
注册时间: 2019-12-19
最后登录: 2021-10-29

楼主

ab769488716

发表于 2021-4-9 14:22:55 |AI写论文

是否 +2 论坛币

k人参与回答

经管之家送您一份

应届毕业生专属福利!

求职就业群

赵安豆老师微信：zhaoandou666

经管之家联合CDA

送您一个全额奖学金名额~ !

立即领取

感谢您参与论坛问题回答

经管之家送您两个论坛币！

+2 论坛币

我的门限回归，P值是显示双门槛显著，三门限不显著，但是在画图的时候第一个门限的图看着不显著，我觉得是因为第一个门限值的F值过大导致的，我这还能做门限回归吗，一般对F值来说，不是越大越好吗，我能继续按照这个数据写毕业论文吗，求求大神救救我，我想毕业

，我的数据整理和画出的图都上传到图片了

Threshold estimator (level = 95):
-----------------------------------------------------
   model | Threshold       Lower       Upper
-----------+-----------------------------------------
   Th-1 |    34.8500    34.2100    45.9800
   Th-21 |    34.8500    34.2100    45.9800
   Th-22 |    17.4300    17.1100    17.4800
   Th-3 |    9.8300       9.4000    10.1000
-----------------------------------------------------

Threshold effect test (bootstrap = 300 300 300):
-------------------------------------------------------------------------------
Threshold |    RSS       MSE    Fstat Prob Crit10 Crit5 Crit1
-----------+-------------------------------------------------------------------
Single | 1170.6114    0.3185 1614.29  0.0000  12.3550  14.6663  21.6455
Double | 1146.5107    0.3120    77.25  0.0000  12.3943  13.9499  18.7512
Triple | 1129.0969    0.3072    56.68  0.5633  82.2548  87.3758  96.8494
-------------------------------------------------------------------------------

Fixed-effects (within) regression             Number of obs    =    3680
Group variable: Code                         Number of groups =    736

R-sq:  within  = 0.4876                      Obs per group: min =       5
   between = 0.4349                                     avg =    5.0
   overall = 0.4562                                     max =       5

                                             F(9,2935)       = 310.32
corr(u_i, Xb)  = 0.0303                      Prob > F          = 0.0000

------------------------------------------------------------------------------
      Perf |    Coef. Std. Err.    t P>|t|    [95% Conf. Interval]
-------------+----------------------------------------------------------------
      Top3 | .0102886 .0031952    3.22 0.001    .0040236 .0165536
Sizeemployee | .0535741 .0445249    1.20 0.229 -.0337292 .1408773
      Age |  -.0484163 .0085084 -5.69 0.000 -.0650994 -.0317332
   state | .0144249 .0644674    0.22 0.823 -.1119811 .1408309
boardsize | .0036599 .0165053    0.22 0.825 -.0287033    .036023
         |
  _cat#c.CSR |
      0  | .1524541 .0075299 20.25 0.000    .1376898 .1672185
      1  | .1041655 .0028837 36.12 0.000    .0985112 .1098198
      2  | .0882833 .0018546 47.60 0.000    .0846469 .0919197
      3  | .0306992 .0012005 25.57 0.000    .0283454    .033053
         |
   _cons |  -1.802742 .4265244 -4.23 0.000 -2.639059 -.9664242
-------------+----------------------------------------------------------------
   sigma_u |  .70107606
   sigma_e |  .62024254
      rho | .5609483 (fraction of variance due to u_i)
------------------------------------------------------------------------------
F test that all u_i=0: F(735, 2935) = 4.74                Prob > F = 0.0000