Propensity score matching in estimating the effect of managerial education on academic planning behavior. Study design: a cross-sectional study

Table 2 Results of the structural equation model and propensity score analysis of APB

Variables	Description % or mean (range)	Model 1 Logistic regression (probit procedure)			Model 2 Propensity Analysis
		Recall (Probit coef.)	Ideation (Probit coef.)	APB (Probit coef.)	Recall (Probit coef.)	APB (Probit coef.)
Age (older vs. younger)	21.04 (18-27)	1.22***	0.20***	-	1.22***	-
Gender (female vs. male)	53.00	0.82***	0.17	-	0.80***	-
Parental occupation (white vs. blue collar)	41.09	0.07	0.29*	-	0.06	-
Place of permanent residence (urban vs. rural)	47.27	0.19*	-	-	0.18	-
Ideation (higher vs. lower)	18.05			0.75***		0.75*
Recall (higher vs. lower)	29.22		0.50*	0.57*		0.47*
Academic year (senior vs. junior)	51.54	-	-	0.19*	-	0.10
Living stipend level (million Vietnamese Dong)	1.36 (0.3-10)	0.14	-	0.11	0.15	0.11
Religion (yes vs. no)	2.14	-	0.60	-	-	-
Number of cases	421	421	421	421	421	421
Variance explained (adjusted R²)		0.27	0.22	0.32	.20	0.12
Exclusion test (model fitness) (p)¹		NS	NS	NS	NS	NS

Test for endogeneity - Biprobit rho for Model 2: 0.07 (-0.30-0.43)
rho-based p: Chi = 0.1378; p > 0.710 ²
Statistical significance: *p < .05; **p < .01; ***p < .001; NS = not significant
Exclusion of a variable for model identification is indicated by (-).
¹The likelihood ratio test was used for binary dependent variables.

ISSN: 1472-6920