Green human resource management and millennial retention in Indonesian tech startups: mediating roles of job expectations and self-efficacy

Figures & data

Figure 1. Research framework.

Table 1. Profile respondents.

Demographic	Details
Total respondents	436
Age range	24–39 (Millennials born between 1981 and 1996)
Gender distribution	55% Male, 45% Female
Education level	71% Bachelor’s Degree, 24% Master’s Degree, 5% Other
Job position	63% Mid-Level, 31% Entry-Level, 6% Senior/Management
Company size	Average of 50–200 employees per company
Experience in current company	Less than 2 years: 40%, 2–5 years: 35%, More than 5 years: 25%
Industry	Technology Startups
Location	Spread across major cities in Indonesia

Table 2. Common method bias test.

Test conducted	Statistical technique	Result (variance explained)	Interpretation
Harman’s single-factor test	Principal component analysis (PCA)	22.7%	Indicative of minimal common method bias as the variance explained is below the 50% threshold.

Table 3. Measurement model.

Construct/variable	Indicators	Outer loading (convergent validity)	AVE	Composite reliability	Cronbach’s alpha
GHRM	G1, G2, G3, G4, G5, G6, G7, G8	0.761, 0.722, 0.765, 0.761, 0.825, 0.722, 0.768, 0.752	0.578	0.916	0.895
Job expectations	JE1, JE2, JE3	0.865, 0.766, 0.776	0.645	0.845	0.725
Employee retention	ER1, ER2, ER3, ER4, ER5, ER6, ER7, ER8	0.853, 0.787, 0.818, 0.837, 0.840, 0.798, 0.903, 0.809	0.673	0.954	0.904
Self-efficacy	SE1, SE2, SE3, SE4, SE5, SE6, SE7	0.830, 0.885, 0.896, 0.885, 0.813, 0.798, 0.901, 0.896, 0.706, 0.887	0.726	0.963	0.946

Note: The indicators (e.g., G1, JE1, ER1, and SE1) are representative labels for the respective items in each construct’s measurement scale.

Table 4. HTMT ratio for discriminant validity.

Constructs	GHRM	Job Expectations	Employee Retention	Self-efficacy
GHRM	—
Job expectations	0.82	—
Employee retention	0.78	0.75	—
Self-efficacy	0.70	0.68	0.65	—

Table 5. PLS predict.

Construct	Q2 (predictive relevance)	Root mean square error (RMSE)	Mean absolute error (MAE)	Interpretation
GHRM	0.55	0.12	0.09	Good predictive relevance and accuracy; Q2 > 0 and low error rates.
Job expectations	0.62	0.10	0.08	Indicates strong predictive relevance and accuracy for Job expectations.
Employee retention	0.48	0.15	0.11	Satisfactory predictive relevance and accuracy for Employee retention.
Self-efficacy	0.59	0.11	0.09	Demonstrates good predictive relevance and accuracy for Self-efficacy.

Table 6. Path coefficient, R-square, direct, and indirect effects.

Analysis type	Path/variable	Path coefficient/R-square	T-statistic	P-value	Information/interpretation
Path coefficient test	Ghrm → self efficacy	0.762	7.113	0.000	Strong relation
	Ghrm → job expectation	0.654	5.180	0.000	Medium relation
	Ghrm → employee retention	0.488	3.542	0.000	Medium relation
	Job expectations → employee retention	0.347	2.206	0.028	Medium relation
	Self-efficacy → employee retention	−0.314	1.292	0.197	Medium relation, but not significant
R-square value	Job expectations	0.428	—	—	42.8% variance explained
	Employee retention	0.676	—	—	67.6% variance explained
	Self-efficacy	0.580	—	—	58.0% variance explained
Indirect effect test	Ghrm → job expectation → employee retention	0.227	1.158	0.247	Not significant
	Ghrm → self efficacy → employee retention	0.239	1.822	0.069	Not significant

Data availability statement

The raw data supporting the conclusions of this article will be made by request to the corresponding authors.