请教利用xtptm程序作门限回归时F临界值为负数是什么原因-经管之家官网！

坛友互助群

扫码加入各岗位、行业、专业交流群

. xtptm lngdp g s u e f,rx(t) thrvar(lngdp) iters(1000) regime(3)

================ Fundamental information: ===================

Number of Regime independent variables: 5

Number of Regime dependent variables: 1

Number of individuals in panel:257

Number of periods in panel: 12

============ Triple threshold regession: stage = 1

Minimized Sum of Squared Residuals:213.0979

Standard error of residuals: 0.2750

Threshold estimate: 9.3768

95% conf. int. of threshold: 9.3759 9.3953

Three Thresholds: 9.3768 8.7434 9.8584

Threshold regression(Ordinary Std. Err.):

----------------------------------------------------

| Coef Std t prob

----+-----------------------------------------------

****

----------------------------------------------------

Threshold regression(Robust Std. Err.):

-----------------------------------------------------

| Coef Std_Robust t prob

----+------------------------------------------------

****

-----------------------------------------------------

============ Triple threshold regession: stage = 2

Minimized Sum of Squared Residuals:206.5939

Standard error of residuals: 0.2708

Threshold estimate: 10.1816

95% conf. int. of threshold: 10.1805 10.1841

Three Thresholds: 10.1816 9.3768 8.7434

Threshold regression(Ordinary Std. Err.):

----------------------------------------------------

| Coef Std t prob

----+-----------------------------------------------

****

----------------------------------------------------

Threshold regression(Robust Std. Err.):

-----------------------------------------------------

| Coef Std_Robust t prob

----+------------------------------------------------

****

-----------------------------------------------------

============ Triple threshold regession: stage = 3

Minimized Sum of Squared Residuals:205.9329

Standard error of residuals: 0.2703

Threshold estimate: 8.6134

95% conf. int. of threshold: 8.6067 8.6152

Three Thresholds: 8.6134 10.1816 9.3768

Threshold regression(Ordinary Std. Err.):

----------------------------------------------------

| Coef Std t prob

----+-----------------------------------------------

1 | 2.5947 0.1215 21.3543 0.0000

2 | -0.4587 0.1038 -4.4193 0.0000

3 | 1.5073 0.1019 14.7871 0.0000

4 | -6.1648 0.5753 -10.7152 0.0000

5 | -0.9076 0.2573 -3.5277 0.0004

6 | -0.3048 0.0142 -21.5063 0.0000

7 | -0.0714 0.0100 -7.1588 0.0000

8 | 0.0981 0.0093 10.5691 0.0000

9 | 0.2781 0.0108 25.7941 0.0000

----------------------------------------------------

Threshold regression(Robust Std. Err.):

-----------------------------------------------------

| Coef Std_Robust t prob

----+------------------------------------------------

1 | 2.5947 0.8631 3.0064 0.0027

2 | -0.4587 0.1675 -2.7380 0.0062

3 | 1.5073 0.1531 9.8478 0.0000

4 | -6.1648 0.7636 -8.0730 0.0000

5 | -0.9076 0.2561 -3.5438 0.0004

6 | -0.3048 0.0332 -9.1840 0.0000

7 | -0.0714 0.0235 -3.0364 0.0024

8 | 0.0981 0.0305 3.2166 0.0013

9 | 0.2781 0.0330 8.4217 0.0000

-----------------------------------------------------

Thresholds in triple model:

1

+---------------+

1 |9.376786232|

2 |10.18164921|

3 |8.613411903|

+---------------+

Ho: Double threshold; Ha: Triple threshold

Number of bootstrap:1000

F-stat & Prob:523.5828 0.0000

F-critical value of 90% 95% 99%:

1

+----------------+

1 |-703.1379895|

2 |-677.8133244|

3 |-647.8346884|

+----------------+

扫码或添加微信号：坛友素质互助

「经管之家」APP：经管人学习、答疑、交友，就上经管之家！
免流量费下载资料----在经管之家app可以下载论坛上的所有资源，并且不额外收取下载高峰期的论坛币。
涵盖所有经管领域的优秀内容----覆盖经济、管理、金融投资、计量统计、数据分析、国贸、财会等专业的学习宝库，各类资料应有尽有。
来自五湖四海的经管达人----已经有上千万的经管人来到这里，你可以找到任何学科方向、有共同话题的朋友。
经管之家（原人大经济论坛），跨越高校的围墙，带你走进经管知识的新世界。
扫描下方二维码下载并注册APP