Switching from LPG to electricity in Indonesia: a study in the Java Bali grid

Figures & data

Table B1. Demographic and characteristics of respondents.

Variable		Count
Gender	female	146 [50.34%]
	male	144 [49.66%]
Age	15–24	38 [13.1%]
	25–34	48 [16.55%]
	35–44	74 [25.52%]
	45–54	83 [28.62%]
	55–64	39 [13.45%]
	Above 64	8 [2.76%]
Household category	1300 Volt Ampere (VA)	115 [39.66%]
	2200 Volt Ampere (VA)	112 [38.62%]
	3500 Volt Ampere (VA)	42 [14.48%]
	4400 Volt Ampere (VA)	9 [3.1%]
	5500 Volt Ampere (VA) and above	12 [4.14%]
LPG consumption	subsidized LPG (3kg cylinder)	100 [34.48%]
	non subsidized LPG	134 [46.21%]
	both	56 [19.31%]
City	Jakarta area	140 [48.28%]
	Outside Jakarta (other big cities: Bandung, Yogyakarta, Surabaya)	150 [51.72%]
Exposure level	partially use an induction stove for cooking	55 [18.97%]
	never use an induction stove for cooking	235 [81.03%]
Employment	Full-time employee in the office/ factory	129 [44.48%]
	students, college students	31 [10.69%]
	Full-time employee work from home	43 [14.83%]
	entrepreneur	27 [9.31%]
	not working on purpose (i.e., taking care of the house)	22 [7.59%]
	retired	18 [6.21%]
	others	20 [6.9%]
Education level	high school	51 [17.59%]
	bachelor	40 [13.79%]
	graduate	134 [46.21%]
	postgraduate	65 [22.41%]
Household/life situation	single and live alone	10 [3.45%]
	single and living with parents/ friends	50 [17.24%]
	partner without children	25 [8.62%]
	couple with children	198 [68.28%]
	others (prefer not to answer)	7 [2.41%]

Table B2. Means and standard deviation of constructs’ indicators.

Item	Ref	Mean	SD	Kurtosis	Skewness
y.1	(Wu et al., 2017)	3.417	1.064	−0.697	−0.101
y.2		3.310	1.067	−0.614	−0.046
y.3		3.317	1.042	−0.539	0.052
y.4		3.207	1.056	−0.552	0.072
y.5	(Chang & Chen, 2014)	3.334	1.068	−0.642	−0.033
x1.1	(Davis, 1989)	3.379	1.051	−0.530	−0.090
x1.2		3.245	1.101	−0.470	−0.201
x1.3		3.348	1.040	−0.362	−0.201
x1.4		3.441	1.050	−0.432	−0.258
x1.5		3.700	1.032	−0.193	−0.546
x1.6		3.845	0.976	0.238	−0.715
x2.1	(Davis, 1989)	4.021	0.958	0.406	−0.871
x2.2		4.007	0.928	0.496	−0.819
x2.3		4.034	0.894	0.810	−0.854
x2.4		3.938	0.904	−0.251	−0.525
x3.1	(Gross & Sheth, 1989; Hilmiyati-Mas’adah et al., 2023)	3.976	0.896	−0.699	−0.386
x3.2		3.734	0.930	−0.797	−0.147
x3.3		3.676	0.953	−0.620	−0.152
x3.4		3.762	0.911	−0.579	−0.199
x3.5		3.872	0.918	−0.432	−0.417
x3.6		3.717	0.930	−0.656	−0.158
z.11	(Zailani et al., 2019)	3.517	0.997	−0.274	−0.195
z.12		3.586	0.910	−0.579	0.004
z.13		3.276	1.079	−0.550	−0.121
z.14		3.317	1.035	−0.514	−0.066
z.15		3.690	0.961	−0.211	−0.350
z.21	(Zailani et al., 2019)	3.321	1.161	−0.585	−0.290
z.22		3.376	1.151	−0.582	−0.349
z.23		2.990	1.152	−0.592	0.020
z.31	(Mutum et al., 2021)	3.097	1.141	−0.540	−0.162
z.32		3.097	1.094	−0.497	−0.066
z.33	(Zailani et al., 2019)	3.334	1.041	−0.400	−0.207
z.41	(Zailani et al., 2019)	3.717	1.128	−0.577	−0.499
z.42		3.841	1.011	−0.591	−0.501
z.43		3.986	0.958	−0.063	−0.730
z.51	(Mutum et al., 2021)	3.886	0.938	−0.058	−0.577
z.52		3.831	0.959	−0.251	−0.505
z.53	(Zailani et al., 2019)	4.055	0.904	0.404	−0.784

Figure 1. Subsidized LPG for poor households.

Source: (Directorate General of Oil & Gas, 2023), (Central Agency on Statistics, 2023)

Figure 2. LPG import-demand ratio.

Source: (Indonesia Ministry of Energy & Mineral Resources, 2020)

Figure 3. Empirical research model.

Table C1. Cross loading of Indicators.

ITEMS	SI	USE	EOU	TOC	FUNC	RECV	EMOV	EPIV	CODV
y.1	0.838	0.516	0.393	0.552	0.595	0.467	0.530	0.435	0.359
y.2	0.910	0.540	0.390	0.541	0.624	0.535	0.631	0.471	0.401
y.3	0.868	0.529	0.437	0.503	0.549	0.567	0.548	0.488	0.385
y.4	0.924	0.586	0.362	0.563	0.622	0.585	0.658	0.446	0.387
y.5	0.920	0.594	0.363	0.515	0.604	0.526	0.643	0.441	0.456
x1.1	0.460	0.779	0.325	0.546	0.553	0.377	0.467	0.234	0.237
x1.2	0.533	0.865	0.350	0.619	0.649	0.502	0.564	0.324	0.309
x1.3	0.581	0.913	0.427	0.641	0.669	0.482	0.573	0.352	0.342
x1.4	0.557	0.902	0.369	0.680	0.645	0.460	0.578	0.359	0.340
x1.5	0.558	0.855	0.502	0.757	0.659	0.465	0.569	0.393	0.368
x1.6	0.440	0.745	0.465	0.626	0.534	0.433	0.499	0.371	0.379
x2.1	0.359	0.367	0.911	0.489	0.411	0.303	0.271	0.517	0.412
x2.2	0.394	0.433	0.942	0.537	0.458	0.337	0.323	0.498	0.408
x2.3	0.437	0.460	0.949	0.600	0.505	0.385	0.372	0.573	0.442
x2.4	0.409	0.509	0.892	0.615	0.507	0.353	0.375	0.511	0.436
x3.1	0.482	0.568	0.686	0.800	0.609	0.406	0.490	0.477	0.467
x3.2	0.533	0.710	0.551	0.898	0.680	0.452	0.546	0.452	0.339
x3.3	0.492	0.696	0.458	0.887	0.640	0.441	0.523	0.386	0.312
x3.4	0.546	0.740	0.492	0.893	0.694	0.472	0.569	0.475	0.398
x3.5	0.533	0.643	0.524	0.838	0.620	0.437	0.523	0.447	0.371
x3.6	0.534	0.630	0.481	0.896	0.657	0.502	0.559	0.433	0.357
z.11	0.554	0.740	0.469	0.734	0.816	0.608	0.622	0.515	0.471
z.12	0.560	0.626	0.462	0.671	0.873	0.495	0.598	0.462	0.516
z.13	0.600	0.596	0.351	0.598	0.827	0.496	0.614	0.344	0.410
z.14	0.572	0.538	0.337	0.498	0.813	0.441	0.555	0.339	0.421
z.15	0.492	0.515	0.499	0.574	0.815	0.418	0.495	0.500	0.511
z.21	0.530	0.496	0.401	0.478	0.547	0.938	0.668	0.550	0.451
z.22	0.542	0.506	0.431	0.466	0.526	0.942	0.669	0.554	0.441
z.23	0.574	0.472	0.194	0.480	0.567	0.858	0.722	0.447	0.380
z.31	0.611	0.600	0.314	0.541	0.646	0.657	0.935	0.387	0.432
z.32	0.643	0.614	0.335	0.574	0.633	0.774	0.943	0.434	0.437
z.33	0.651	0.599	0.380	0.620	0.692	0.687	0.940	0.459	0.548
z.41	0.285	0.257	0.281	0.299	0.406	0.337	0.418	0.409	0.834
z.42	0.489	0.398	0.450	0.417	0.546	0.482	0.522	0.530	0.917
z.43	0.330	0.330	0.435	0.377	0.480	0.352	0.336	0.486	0.839
z.51	0.462	0.335	0.458	0.441	0.440	0.526	0.375	0.921	0.466
z.52	0.535	0.447	0.549	0.504	0.544	0.580	0.517	0.901	0.534
z.53	0.376	0.299	0.536	0.439	0.425	0.420	0.324	0.896	0.508

INT: Intention to switch cooking energy; USE: Usefulness; EOU: Ease of use; TOC: Time-oriented Convenience; FUNC: Functional Value; RECV: Recognition Value; EMOV: Emotional Value; EPIV: Epistemic value; CODV: Conditional Value

Table C2. Validity and reliability.

Construct	Cronbach’s Alpha	ρA	Composite Reliability	Average Variance Extracted (AVE)
Intention to switch cooking energy	0.936	0.939	0.951	0.797
Usefulness	0.919	0.924	0.937	0.714
Ease of use	0.943	0.947	0.959	0.854
Time-oriented Convenience	0.935	0.937	0.949	0.756
Functional Value	0.886	0.890	0.916	0.687
Recognition Value	0.900	0.900	0.938	0.834
Emotional Value	0.934	0.934	0.958	0.883
Epistemic value	0.891	0.899	0.932	0.821
Conditional Value	0.835	0.912	0.898	0.747

Table C3. Discriminant validity: Heterotrait-Monotrait.

Construct	USE	EOU	TOC	FUNC	RECV	EMOV	CODV	EPIV
USE
EOU	0.515
TOC	0.824	0.648
FUNC	0.805	0.556	0.814
RECV	0.591	0.405	0.567	0.664
EMOV	0.693	0.386	0.658	0.764	0.820
CODV	0.431	0.504	0.477	0.640	0.519	0.555
EPIV	0.438	0.617	0.558	0.580	0.627	0.489	0.633
SI	0.666	0.462	0.640	0.735	0.655	0.722	0.480	0.554

Table C4. Inner variance inflation factor (VIF) of structural model constructs.

Constructs	VIF
Usefulness	2.418
Ease of use	1.593
Time-oriented Convenience	2.955
Functional Value	2.317
Recognition Value	2.702
Emotional Value	3.041
Conditional Value	1.735
Epistemic value	1.824

Table   C5. Quadratic effects on unsupported linear correlation paths.

Path	Path coefficient (β)	t statistics	Quadratic effect result
H1d: USE → EPIV	0.139	3.005	Significant (R^2=0.390)
H2a: EOU → FUNC	−0.056	0.881	Not significant
H2b: EOU → RECV	0.073	0.876	Not significant
H2c: EOU → EMOV	−0.078	1.296	Not significant
H5: RECV → INT	0.108	2.110	Significant (R^2=0.579)

Notes: t-values test: 1.96 at sig level 5%

Table C6. Predictive power of structural model

Indicators	Q²_predict	RMSE PLS	RMSE LM	RMSE PLS-LM
y.1	0.298	0.895	0.929	−0.034
y.2	0.310	0.889	0.920	−0.031
y.3	0.288	0.882	0.899	−0.017
y.4	0.327	0.870	0.893	−0.023
y.5	0.313	0.888	0.869	0.019

Table 1. Structural model estimates.

Hypotheses	Path coefficient	t-statistics	Result
H1a: USE → FUNC	0.387	4.964	Accepted
H1b: USE → RECV	0.336	4.026	Accepted
H1c: USE → EMOV	0.415	5.332	Accepted
H1d: USE → EPIV	0.013	0.180	Rejected
H2a: EOU → FUNC	0.079	1.475	Rejected
H2b: EOU → RECV	0.084	1.148	Rejected
H2c: EOU → EMOV	−0.026	0.454	Rejected
H2d: EOU → EPIV	0.409	5.807	Accepted
H3a: TOC → FUNC	0.404	5.155	Accepted
H3b: TOC → RECV	0.212	2.189	Accepted
H3c: TOC → EMOV	0.314	3.834	Accepted
H3d: TOC → EPIV	0.253	2.556	Accepted
H4: FUNC → INT	0.315	4.228	Accepted
H5: RECV → INT	0.081	1.095	Rejected
H6: EMOV → INT	0.327	4.253	Accepted
H7: EPIV → INT	0.151	3.007	Accepted
H8a: CODV x FUNC → INT	0.106	1.684	Rejected
H8b: CODV x RECV → INT	0.019	0.323	Rejected
H8c: CODV x EMOV→ INT	−0.160	2.180	Accepted

Notes: t-values test: 1.96 at sig level 5%

Table 2. Coefficient of determination of endogenous variables: impact of quadratic effect.

Endogenous variables	Linear			Quadratic
	R^2 excl. moderating effect	R^2 incl. moderating effect	Effect size (f^2)	R^2	Effect size (f^2)
Emotional value	0.451	0.451
Epistemic value	0.368	0.368		0.390	3.70%
Functional value	0.627	0.627
Recognition value	0.323	0.323
Intention to switch cooking energy	0.558	0.568	2.31%	0.578	4.66%

f² 0.020-0.150:low; 0.150-0.350: medium; > =0.350: high (Benitez et al., 2020).

for social science research, R²=0.75 high; R²=0.5 medium; R²=0.25 low (Hair et al., 2021)

Table 3. Multi group analysis.

Correlations	Female vs. Male		Jakarta vs. other cities		Under 45 vs. 45 above
	β-diff	p-value	β-diff	p-value	β-diff	p-value
USE → FUNC	−0.128	0.412	0.101	0.474	−0.005	0.967
USE → RECV	0.204	0.238	0.155	0.352	0.173	0.297
USE → EMOV	−0.209	0.188	−0.034	0.817	−0.096	0.532
USE → EPIV	0.131	0.365	0.139	0.351	−0.197	0.182
EOU → FUNC	0.092	0.385	0.134	0.176	−0.007	0.971
EOU → RECV	0.059	0.699	−0.113	0.423	−0.136	0.355
EOU → EMOV	−0.050	0.694	0.015	0.901	0.000	0.978
EOU → EPIV	0.330	0.009	0.137	0.306	0.023	0.855
TOC → FUNC	0.042	0.792	−0.194	0.167	−0.031	0.830
TOC → RECV	−0.295	0.145	−0.114	0.551	−0.126	0.498
TOC → EMOV	0.192	0.285	0.091	0.569	0.100	0.551
TOC → EPIV	−0.400	0.036	−0.319	0.095	0.025	0.917
FUNC → INT	0.112	0.460	0.132	0.368	−0.081	0.605
RECV → INT	−0.317	0.057	−0.240	0.167	−0.078	0.642
EMOV → INT	0.083	0.592	0.132	0.398	0.319	0.055
EPIV → INT	−0.067	0.554	−0.030	0.786	−0.105	0.350
CODV x FUNC → INT	−0.126	0.325	−0.055	0.661	−0.039	0.791
CODV x RECV → INT	−0.160	0.257	−0.052	0.689	−0.229	0.122
CODV x EMOV→ INT	0.116	0.466	0.086	0.581	0.148	0.374

Notes: β-diff = path coefficient difference between two groups; t-values test: 1.96 at sig level 5%

Other provincial cities: Bandung (West Java), Yogyakarta (in the Central Java area), Surabaya (East Java)

Table 4. Multi-group analysis by gender.

Correlations	Original path	β –diff by gender	Female		Male
			Path coef.	p-value	Path coef.	p-value
USE → FUNC	0.387	NS	0.304	0.014	0.432	0.000
USE → RECV	0.336	NS	0.455	0.001	0.250	0.023
USE → EMOV	0.415	NS	0.301	0.011	0.510	0.000
USE → EPIV	0.013	NS	0.076	0.529	−0.056	0.512
EOU → FUNC	0.079	NS	0.123	0.153	0.031	0.634
EOU → RECV	0.084	NS	0.092	0.379	0.033	0.756
EOU → EMOV	−0.026	NS	−0.048	0.520	0.002	0.984
EOU → EPIV	0.409	S	0.536	0.000	0.206	0.032
TOC → FUNC	0.404	NS	0.451	0.000	0.409	0.000
TOC → RECV	0.212	NS	0.061	0.673	0.356	0.013
TOC → EMOV	0.314	NS	0.410	0.001	0.218	0.089
TOC → EPIV	0.253	S	0.092	0.500	0.491	0.000
FUNC → INT	0.315	NS	0.364	0.001	0.252	0.020
RECV → INT	0.081	NS	−0.021	0.807	0.297	0.031
EMOV → INT	0.327	NS	0.364	0.001	0.281	0.017
EPIV → INT	0.151	NS	0.133	0.124	0.200	0.006
CODV x FUNC → INT	0.106	NS	0.031	0.718	0.157	0.092
CODV x RECV → INT	0.019	NS	−0.032	0.708	0.128	0.251
CODV x EMOV→ INT	−0.160	NS	−0.133	0.213	−0.248	0.038
R^2 Intention to switch	0.568		0.517		0.656

Notes: original path coefficient applied on all data; t-values test: 1.96 at a significance level of 5%

Wu, K., Vassileva, J., & Zhao, Y. (2017). Understanding users’ intention to switch personal cloud storage services: Evidence from the Chinese market. Computers in Human Behavior 68(68), 300–314. https://doi.org/10.1016/j.chb.2016.11.039

 Google Scholar

Chang, C.-C. L., & Chen, K. (2014). The push, pull and mooring effects in virtual migration for social networking sites. Information Systems Journal 24(4), 323–346. https://doi.org/10.1111/isj.12030

 Web of Science ®Google Scholar

Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly 13(3), 319–340. https://doi.org/10.2307/249008

 Web of Science ®Google Scholar

Gross, B. L., & Sheth, J. N. (1989). Time-oriented advertising: A content analysis of United States magazine advertising, 1890–1988. Journal of Marketing 53(4), 76–83. https://doi.org/10.1177/002224298905300406

 Web of Science ®Google Scholar

Hilmiyati-Mas’adah, H., Sudiro, A., Rohman, F., Yuniarinto, A., Hakam, D. F., & Nugraha, H. (2023). Assessing the impact of urban lifestyle and consumption values on conversion intention: A study towards energy sustainability. Sustainability 15(8), 6549. https://doi.org/10.3390/su15086549

 Web of Science ®Google Scholar

Zailani, S., Iranmanesh, M., Sean Hyun, S., & Ali, M. (2019). Applying the theory of consumption values to explain drivers’ willingness to pay for biofuels. Sustainability 11(3), 668. https://doi.org/10.3390/su11030668

 Web of Science ®Google Scholar

Mutum, D. S., Ghazali, E. M., & Wei‐Pin, W. (2021). Parallel mediation effect of consumption values and the moderation effect of innovativeness, in predicting the influence of identity on green purchasing behavior. Journal of Consumer Behaviour 20(3), 827–844. https://doi.org/10.1002/cb.1913

 Web of Science ®Google Scholar

Directorate General of Oil and Gas. (2023). Performance Report 2022. Jakarta.

 Google Scholar

Central Agency on Statistics. (2023). Number of poor people. Www.Bps.Go.Id. Retrieved July 28, 2023 , from https://www.bps.go.id/site/resultTab

 Google Scholar

Indonesia Ministry of Energy and Mineral Resources. (2020). Strategic planning ministry of energy and mineral resources 2020-2024 (Vol. 16). Jakarta.

 Google Scholar

Benitez, J., Henseler, J., Castillo, A., & Schuberth, F. (2020). How to perform and report an impactful analysis using partial least squares: Guidelines for confirmatory and explanatory IS research. Information & Management 57(2), 103168. https://doi.org/10.1016/j.im.2019.05.003

 Web of Science ®Google Scholar

Hair, J. F., Hult, G. T. M., Ringle, C. M., Sarstedt, M., Danks, N. P., & Ray, S. (2021). Partial least squares structural equation modeling (PLS-SEM) using R: A workbook. Springer.

 Google Scholar

Data availability statement

Data available on request