Analysing the impact of post-pandemic factors on entrepreneurial intentions: the enduring significance of self-efficacy in student planned behaviour

Figures & data

Figure 1. Proposed research model.

Table 1. Profile of respondents.

Variable	Categories	Frequencies	Percentage
Gender	Male	129	41.60%
	Female	181	58.40%
Education Qualification	Postgraduate	159	51.30%
	Undergraduate	151	48.70%
Area	Urban	175	56.50%
	Rural	135	43.50%

Table 2. Constructs reliability and validity.

Constructs	Items	Factor Loadings	Cronbach’s Alpha	CR	AVE	VIF
Attitude Towards Behavior	ATT1	0.861	0.911	0.934	0.738	2.518
	ATT2	0.848				2.394
	ATT3	0.844				2.393
	ATT4	0.854				2.444
	ATT5	0.887				2.939
Entrepreneurial Intentions	EI1	0.861	0.920	0.940	0.759	2.551
	EI2	0.883				2.925
	EI3	0.860				2.614
	EI4	0.879				2.874
	EI5	0.871				2.690
Perceived Behavior Control	PBC1	0.763	0.841	0.893	0.676	1.687
	PBC2	0.849				2.157
	PBC3	0.850				2.137
	PBC4	0.824				1.750
Self-efficacy	SE1	0.768	0.851	0.893	0.627	1.741
	SE2	0.796				1.836
	SE3	0.811				1.890
	SE4	0.770				1.771
	SE5	0.810				1.983
Subjective Norms	SN1	0.826	0.921	0.939	0.718	2.333
	SN2	0.849				2.595
	SN3	0.849				2.632
	SN4	0.809				2.181
	SN5	0.870				2.840
	SN6	0.880				3.021

Table 3. Fornell-Larckers criterion and HTMT ratio.

Constructs	1	2	3	4	5
1. ATT	0.859
2. EI	0.609	0.871
3. PBC	0.343	0.524	0.822
4. SE	0.632	0.661	0.401	0.792
5. SN	0.574	0.637	0.457	0.635	0.848
HTMT ratio
1. ATT	_
2. EI	0.663	_
3. PBC	0.386	0.589	_
4. SE	0.717	0.743	0.470	_
5. SN	0.625	0.691	0.517	0.717	_

Figure 2. Structural model.

Table 4. Model fit results.

	Estimated Model
SRMR	0.044
d_ULS	0.642
d_G	0.292
Chi-Square	529.664
NFI	0.902
RMS_theta	0.12

Table 5. Results of hypotheses.

Hypothesis	Path	Path Co-efficient Value	Standard Deviation	t-values	P Value	F2	Remarks
H1	ATT -> EI	0.198	0.086	2.294	0.022**	0.064	Supported
H2	PBC -> EI	0.255	0.056	4.546	0.000**	0.100	Supported
H3	SN -> EI	0.185	0.091	2.035	0.042**	0.061	Supported
H4	SE -> EI	0.351	0.155	3.056	0.002**	0.096	Supported

**p < 0.05.

Table 6. Multi-group hypothesis testing (gender).

		Path Coefficient
Hypotheses	Relationship	Female	Male	Difference	Confidence Interval (2.5%; 97.5%)	p-Value	Results
H1a	ATT -> EI	0.209	0.230	−0.020	(−0.250; 0.256)	0.433	Not Supported
H2a	PBC -> EI	0.245	0.216	0.029	(−0.195; 0.185)	0.378	Not Supported
H3a	SN -> EI	0.230	0.211	0.020	(−0.256; 0.255)	0.439	Not Supported
H4a	SE -> EI	0.287	0.302	−0.015	(−0.298; 0.283)	0.446	Not Supported

Table 7. Multi-group hypothesis testing (area).

		Path Coefficient
Hypotheses	Relationship	Rural	Urban	Difference	Confidence Interval (2.5%; 97.5%)	p-Value	Results
H1b	ATT -> EI	0.175	0.229	−0.054	(−0.249; 0.256)	0.316	Not Supported
H2b	PBC -> EI	0.147	0.307	−0.161	(−0.193; 0.175)	0.044**	Supported
H3b	SN -> EI	0.245	0.230	0.015	(−0.253; 0.265)	0.468	Not Supported
H4b	SE -> EI	0.360	0.231	0.128	(−0.289; 0.286)	0.169	Not Supported

Figure 3. IPMA results.

Table 8. PLS predict results.

PLS Predict (PLS)					PLS Predict (LM)
	RMSE	MAE	MAPE	Q²_predict	RMSE	MAE	MAPE	Q²_predict
EI4	1.225	0.972	26.8	0.426	1.263	0.991	27.255	0.389
EI3	1.274	0.997	27.77	0.387	1.322	1.053	29.46	0.340
EI5	1.213	0.958	25.616	0.470	1.249	0.992	26.016	0.438
EI1	1.145	0.909	23.377	0.428	1.186	0.934	24.023	0.386
EI2	1.236	0.971	28.67	0.437	1.265	0.978	28.704	0.410

Figure 4. PLS-SEM and ANN normalized importance.

Table 9. ANN- RMSE results.

Training			Testing
INPUT: ATT, PBC, SN, SE OUTPUT: EI			INPUT: ATT, PBC, SN, SE OUTPUT: EI
N	SSE	RMSE	N	SSE	RMSE	Differences (Training-Testing)	Total Sample
212	44.595	0.458643	98	21.122	0.464253	−0.0056	310
225	37.617	0.408885	85	12.822	0.38839	0.0205	310
213	38.714	0.426328	97	14.924	0.392244	0.0341	310
213	34.718	0.403727	97	17.387	0.423376	−0.0196	310
216	34.801	0.401392	94	14.396	0.391342	0.0100	310
223	41.861	0.433264	87	13.322	0.391314	0.0420	310
205	31.502	0.392005	105	15.759	0.387409	0.0046	310
215	32.977	0.391639	95	14.444	0.389926	0.0017	310
221	43.289	0.442581	89	18.103	0.451004	−0.0084	310
228	45.639	0.4474 05	82	17.15	0.457325	−0.0099	310
MEAN		0.420587			0.413658	0.0069
SD		0.024243			0.032149	−0.0079

Note: SSE = sum square of errors, RMSE = root mean square of errors, N = sample size, SE = self-efficacy, PBC = perceived behavior control, ATT = attitude, SN = subjective norms and EI = entrepreneurial intentions.

Figure 5. ANN model.

Table 10. Sensitivity analysis.

Sensitivity Analysis Importance	PBC	SN	SE	ATT
ANN1	100.0%	86.9%	63.8%	59.1%
ANN2	64.2%	36.9%	100.0%	67.8%
ANN3	73.2%	47.4%	100.0%	83.4%
ANN4	49.9%	38.4%	100.0%	68.1%
ANN5	81.8%	23.6%	100.0%	61.1%
ANN6	56.0%	32.3%	100.0%	53.7%
ANN7	75.8%	19.0%	100.0%	72.2%
ANN8	65.0%	54.3%	100.0%	48.6%
ANN9	89.6%	55.8%	98.0%	100.0%
ANN10	61.0%	67.3%	100.0%	70.7%
Average Importance	0.716	0.462	0.962	0.685
Normalized Importance	74.50%	48.00%	100.00%	71.20%
Rank	2	4	1	3