Py学习  »  Git

Stata学习:如何构建面板Logit模型?xtlogit

Stata与R学习 • 1 年前 • 213 次点击  

文献来源

  1. Wang, X., et al. (2024). Social Media Alleviates Venture Capital Funding Inequality for Women and Less Connected Entrepreneurs. Management Science.

示例代码

cd "C:\Download\mnsc.2023.4728\replication_package_MS-INS-21-00189"
use "VC_Twitter_data.dta", clear
*
egen std_gender = std(genderdiversity)
egen std_constraint = std(constraint) if location != 0 & year > 2006 & isolated == 0
g hastwt_gender = std_gender * hastwitter
g twt_engage_gender = std_gender * twt_engage
*
g leadgetfunded = .
replace leadgetfunded = 0 if logleadfunding == 0
replace leadgetfunded = 1 if logleadfunding > 0
xtlogit leadgetfunded yearsold totalfundingrounds websiterank googletrends ///
        founder_alumni_degree_max hastwitter twt_engage hastwt_gender ///
		twt_engage_gender year07 year08 year09 year10 year11 year12 ///
		year13 year14 year15 if location != 0 & year > 2006 & isolated == 1, fe

得到结果

note: multiple positive outcomes within groups encountered.
note: 2,100 groups (4,008 obs) omitted because of all positive or
      all negative outcomes.

Iteration 0:  Log likelihood = -11535.635  
Iteration 1:  Log likelihood = -7832.8501  
Iteration 2:  Log likelihood = -7443.2594  
Iteration 3:  Log likelihood = -7437.0157  
Iteration 4:  Log likelihood = -7436.9903  
Iteration 5:  Log likelihood = -7436.9869  
Iteration 6:  Log likelihood = -7436.9863  
Iteration 7:  Log likelihood = -7436.9862  
Iteration 8:  Log likelihood = -7436.9862  
Iteration 9:  Log likelihood = -7436.9862  

Conditional fixed-effects logistic regression      Number of obs    =   71,017
Group variable: newid                              Number of groups =   14,737

                                                   Obs per group:
                                                                min =        2
                                                                avg =      4.8
                                                                max =       10

                                                   LR chi2(18)      = 35362.49
Log likelihood = -7436.9862                        Prob > chi2      =   0.0000

------------------------------------------------------------------------------
leadgetfun~d | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
    yearsold |  -1.008618   .0499464   -20.19   0.000    -1.106511   -.9107248
totalfundi~s |  -2.690731   .0481554   -55.88   0.000    -2.785113   -2.596348
 websiterank |   2.09e-08   3.72e-08     0.56   0.574    -5.21e-08    9.39e-08
googletrends |   .0167098   .0019523     8.56   0.000     .0128833    .0205363
founder_al~x |   .0068908   .0006679    10.32   0.000     .0055817    .0081999
  hastwitter |   .3935411   .0730682     5.39   0.000     .2503301    .5367522
  twt_engage |   .2263231   .0442594     5.11   0.000     .1395763      .31307
hastwt_gen~r |   .1129485   .0619693     1.82   0.068     -.008509    .2344061
twt_engage~r |   .0093881   .0388061     0.24   0.809    -.0666704    .0854465
      year07 |  -52.89135   828.3843    -0.06   0.949    -1676.495    1570.712
      year08 |  -50.97488   828.3843    -0.06   0.951    -1674.578    1572.629
      year09 |  -49.19419   828.3843    -0.06   0.953    -1672.798    1574.409
      year10 |  -47.40924   828.3843    -0.06   0.954    -1671.013    1576.194
      year11 |  -45.57147   828.3843    -0.06   0.956    -1669.175    1578.032
      year12 |  -44.05268   828.3843    -0.05   0.958    -1667.656    1579.551
      year13 |  -42.25346   828.3843    -0.05   0.959    -1665.857     1581.35
      year14 |   -40.8178   828.3844    -0.05   0.961    -1664.421    1582.786
      year15 |  -39.72342   828.3844    -0.05   0.962    -1663.327     1583.88
------------------------------------------------------------------------------

期刊排版

见原文表A7列2。

(完)

Python社区是高质量的Python/Django开发社区
本文地址:http://www.python88.com/topic/169234
 
213 次点击