Table 2 Coefficient from GPSM.

From: The impact of human capital and social capital on entrepreneurship entry: the threshold of human capital-social capital coupling

variables

MD1 (sd)

MD2 (sd)

MD3 (sd)

variables

Coef. (std. err.)

cir_1

−0.00389

(0.00752)

0.03145***

(0.00707)

−0.04863***

(0.01227)

coupling

−45.22047***

(1.983772)

cir_2

0.03161***

(0.0087)

−0.07192***

(0.00834)

0.04814***

(0.01429)

Coupling square

101.1823***

(4.382458)

cir_3

−0.05171***

(0.00829)

0.05944***

(0.00766)

0.04007**

(0.01368)

coupling cubic

−71.24281***

(3.052647)

mar

−0.01232

(0.00672)

−0.00993

(0.00631)

0.04306***

(0.01085)

gps_flog

31.28376***

(2.556385)

age

−6.1217***

(0.24991)

3.3218 9***

(0.23087)

7.1589***

(0.43139)

gps_flog_sq

−89.72813***

(5.491524)

age2

−573.59***

(23.226)

326.63***

(20.49)

658.52***

(40.192)

gps_flog_3

68.0014***

(3.28398)

heal

−0.28362***

(0.02369)

0.19624***

(0.02154)

0.27916***

(0.03941)

coupling_gp~g

0.490786

(2.959742)

hum

0.68381***

(0.01309)

−0.37526***

(0.01005)

−0.98523***

(0.02767)

_cons

4.03992***

(0.2662798)

soc

0.27051***

(0.01255)

0.5922***

(0.00942)

−1.4518***

(0.01937)

  
  1. Notes:
  2. 1. cir_1, cir_2, and cir_3 are dummy variables from cir, sd is standard deviation.
  3. 2. Mean difference was estimated by Generalized Propensity Score Matching(GPS). MD1 was Mean Difference when human capital-social capital coupling was less than or equal to 0.3, MD2 was Mean Difference when human capital-social capital coupling was from 0.3 to 0.7, MD3 was Mean Difference when human capital-social capital coupling was larger than 0.7;coupling is a variable, coupling square is square function of coupling, coupling cubic is cubic function of coupling, they were significant at 0.01.
  4. 3. gps flog is propensity score of treatment variable by GPS, gps flog sq is it’s square, gps_flog_3 is cubic of it, coupling gp~g is the product of treatment variable and it’s propensity score.
  5. ***Regression coefficient is significant at the 1% level; **regression coefficient is significant at the 5% level.
Back to article page