1 medhdfe簡介運(yùn)用Stata進(jìn)行中介效應(yīng)分析時(shí),，你是否曾經(jīng)為如何加入時(shí)間固定效應(yīng)和個(gè)體固定效應(yīng)而煩惱呢,？ 全新編寫的Stata命令medhdfe重磅首發(fā),，支持在進(jìn)行中介效應(yīng)分析時(shí)控制多維固定效應(yīng),，讓你和煩惱說拜拜,。 2 medhdfe特點(diǎn)medhdfe將中介效應(yīng)命令sgmediation2與多維固定效應(yīng)命令reghdfe相結(jié)合，完美解決在中介效應(yīng)分析時(shí)如何加入多維固定效應(yīng)的問題,。
相比在sgmediation2命令中的控制變量部分引入多個(gè)維度虛擬變量,，medhdfe通過absorb加入多維固定效應(yīng)，極大提高了運(yùn)行速度,。 medhdfe還支持加入穩(wěn)健標(biāo)準(zhǔn)誤,，包括異方差穩(wěn)健標(biāo)準(zhǔn)誤robust和聚類穩(wěn)健標(biāo)準(zhǔn)誤cluster。
3 medhdfe語法medhdfe depvar [if exp] [in range] , iv(indepvar) mv(medvar) cv(ctrlvar) absorb(absvars) [options]
其中,，depvar為因變量,，indepvar為自變量，medvar為中介變量,，ctrlvar為控制變量,，absvars為固定維度。 4 medhdfe示例以grunfeld.dta數(shù)據(jù)集為例,，對medhdfe命令的運(yùn)用進(jìn)行演示,。 示例以invest為因變量，mvalue為自變量,，kstock為中介變量,，同時(shí)控制company公司固定效應(yīng)和year年份固定效應(yīng)。 webuse 'grunfeld', clear
medhdfe invest, iv(mvalue) mv(kstock) absorb(company year)
具體結(jié)果如下： Model with dv regressed on iv (path c)
reghdfe invest mvalue , absorb(company year) cluster() vce()
(MWFE estimator converged in 2 iterations)
HDFE Linear regression Number of obs = 200
Absorbing 2 HDFE groups F( 1, 170) = 76.08
Prob > F = 0.0000
R-squared = 0.8808
Adj R-squared = 0.8604
Within R-sq. = 0.3092
Root MSE = 81.0287
------------------------------------------------------------------------------
invest | Coefficient Std. err. t P>|t| [95% conf. interval]
-------------+----------------------------------------------------------------
mvalue | .1799679 .0206334 8.72 0.000 .1392372 .2206986
_cons | -48.70963 23.0425 -2.11 0.036 -94.19591 -3.223356
------------------------------------------------------------------------------
Absorbed degrees of freedom:
-----------------------------------------------------+
Absorbed FE | Categories - Redundant = Num. Coefs |
-------------+---------------------------------------|
company | 10 0 10 |
year | 20 1 19 |
-----------------------------------------------------+
Model with mediator regressed on iv (path a)
reghdfe kstock mvalue , absorb(company year) cluster() vce()
(MWFE estimator converged in 2 iterations)
HDFE Linear regression Number of obs = 200
Absorbing 2 HDFE groups F( 1, 170) = 15.30
Prob > F = 0.0001
R-squared = 0.7127
Adj R-squared = 0.6637
Within R-sq. = 0.0826
Root MSE = 174.6154
------------------------------------------------------------------------------
kstock | Coefficient Std. err. t P>|t| [95% conf. interval]
-------------+----------------------------------------------------------------
mvalue | .1739291 .0444647 3.91 0.000 .0861551 .2617031
_cons | 87.88131 49.65618 1.77 0.079 -10.14081 185.9034
------------------------------------------------------------------------------
Absorbed degrees of freedom:
-----------------------------------------------------+
Absorbed FE | Categories - Redundant = Num. Coefs |
-------------+---------------------------------------|
company | 10 0 10 |
year | 20 1 19 |
-----------------------------------------------------+
Model with dv regressed on mediator and iv (paths b and c')
reghdfe invest kstock mvalue , absorb(company year) cluster() vce()
(MWFE estimator converged in 2 iterations)
HDFE Linear regression Number of obs = 200
Absorbing 2 HDFE groups F( 2, 169) = 217.44
Prob > F = 0.0000
R-squared = 0.9517
Adj R-squared = 0.9431
Within R-sq. = 0.7201
Root MSE = 51.7245
------------------------------------------------------------------------------
invest | Coefficient Std. err. t P>|t| [95% conf. interval]
-------------+----------------------------------------------------------------
kstock | .3579163 .022719 15.75 0.000 .3130667 .4027659
mvalue | .1177158 .0137513 8.56 0.000 .0905694 .1448623
_cons | -80.16378 14.84402 -5.40 0.000 -109.4674 -50.86019
------------------------------------------------------------------------------
Absorbed degrees of freedom:
-----------------------------------------------------+
Absorbed FE | Categories - Redundant = Num. Coefs |
-------------+---------------------------------------|
company | 10 0 10 |
year | 20 1 19 |
-----------------------------------------------------+
Sobel-Goodman Mediation Tests
| Est Std_err z P>|z|
---------------------+------------------------------------------------
Sobel | 0.062 0.016 3.796 0.000
Aroian | 0.062 0.016 3.789 0.000
Goodman | 0.062 0.016 3.804 0.000
Indirect, Direct, and Total Effects
| Est Std_err z P>|z|
---------------------+------------------------------------------------
a_coefficient | 0.174 0.044 3.912 0.000
b_coefficient | 0.358 0.023 15.754 0.000
Indirect_effect_aXb | 0.062 0.016 3.796 0.000
Direct_effect_c' | 0.118 0.014 8.560 0.000
Total_effect_c | 0.180 0.021 8.722 0.000
Proportion of total effect that is mediated: 0.346
Ratio of indirect to direct effect: 0.529
Ratio of total to direct effect: 1.529
5 medhdfe獲取
|
