Digital paradigm shift: Unraveling students’ intentions to embrace Tablet-based Learning through an extended UTAUT2 model

Figures & data

Figure 1. UTAUT2 model.

Source: Venkatesh et al. (2012).

Figure 2. Conceptual framework.

Note: The round dot lines show the moderating path of gender.

Source: Authors’ construct.

Table 1. Constructs, items, IL, CA (α), rho_A, CR and AVE

Constructs	No. of Items	Item	IL	CA (α)	rho_A	CR	AVE
AT	Four (4)	AT1	0.844	0.818	0.830	0.879	0.645
		AT2	0.840
		AT3	0.775
		AT4	0.751
BI	Four (4)	BI1	0.788	0.751	0.755	0.843	0.573
		BI2	0.787
		BI3	0.751
		BI4	0.698
EE	Five (5)	EE1	0.784	0.858	0.864	0.898	0.639
		EE2	0.839
		EE3	0.857
		EE4	0.762
		EE5	0.751
FC	Four (4)	FC1	0.769	0.754	0.755	0.844	0.576
		FC2	0.772
		FC3	0.776
		FC4	0.716
HB	Three (3)	HB1	0.797	0.819	0.826	0.893	0.736
		HB2	0.888
		HB3	0.885
HM	Three (3)	HM1	0.784	0.795	0.804	0.880	0.710
		HM2	0.875
		HM3	0.866
PE	Four (4)	PE1	0.863	0.873	0.877	0.914	0.726
		PE2	0.871
		PE3	0.880
		PE4	0.791
PLO	Four (4)	PLO1	0.887	0.894	0.900	0.926	0.759
		PLO2	0.880
		PLO3	0.906
		PLO4	0.810
SI	Four (4)	SI1	0.784	0.781	0.793	0.858	0.601
		SI2	0.805
		SI3	0.721
		SI4	0.790
T	Five (5)	T1	0.791	0.826	0.828	0.878	0.591
		T2	0.790
		T3	0.804
		T4	0.769
		T5	0.684

Note: AT = Attitude; BI = Behavioural Intention; EE = Effort Expectancy; FC = Facilitating Conditions; HB = Habit; HM = Hedonic Motivation; PLO = Perceived Learning Opportunities; PE = Performance Expectancy; SI = Social Influence; T = Trust; IL = Item Loadings; CA (α) = Cronbach’s Alpha; CR = Composite Reliability and AVE = Average Variance Extracted.

Figure 3. PLS-SEM path coefficients and item loadings.

Table 2. Discriminant validity using Fornell-Larcker criterion

	AT	BI	EE	FC	HB	HM	PLO	PE	SI	T
AT	0.803
BI	0.530	0.757
EE	0.642	0.594	0.800
FC	0.594	0.593	0.610	0.759
HB	0.568	0.428	0.483	0.578	0.858
HM	0.680	0.620	0.583	0.604	0.461	0.843
PLO	0.701	0.546	0.558	0.567	0.637	0.584	0.871
PE	0.595	0.547	0.751	0.598	0.422	0.516	0.481	0.852
SI	0.629	0.634	0.556	0.634	0.445	0.565	0.503	0.568	0.775
T	0.511	0.601	0.578	0.571	0.484	0.565	0.495	0.615	0.568	0.769

Note: The squared inter-construct correlations for the Fornell-Larcker criterion are represented by the values that lie below the diagonal; Values (in bold) on the leading diagonal are the $\sqrt{AVE}$ (square root of the Average Variance Extracted [AVE]) for each construct.

Table 3. Heterotrait-Monotrait (HTMT, <0.90) ratio for discriminant validity

	AT	BI	EE	FC	HB	HM	PLO	PE	SI	T
AT
BI	0.666
EE	0.763	0.735
FC	0.764	0.785	0.752
HB	0.710	0.541	0.573	0.731
HM	0.838	0.798	0.700	0.779	0.566
PLO	0.827	0.659	0.629	0.688	0.736	0.685
PE	0.701	0.668	0.870	0.740	0.501	0.607	0.544
SI	0.775	0.812	0.671	0.831	0.566	0.709	0.600	0.685
T	0.617	0.762	0.680	0.722	0.590	0.690	0.574	0.725	0.701

Note: Values below the diagonal are HTMT correlation ratios; $\mathrm{H}\mathrm{T}\mathrm{M}{\mathrm{T}}_{0.85}or\mathrm{H}\mathrm{T}\mathrm{M}{\mathrm{T}}_{0.90}$ (Henseler et al., 2015).

Table 4. Multicollinearity test

		Variance Inflation Factor (VIF)
S/N	Variables	Inner Model	Outer Model
1	Attitude	3.142	1.834–2.237
2	Behavioural Intention	-	1.400–1.688
3	Effort Expectancy	2.878	1.730–2.423
4	Facilitating Conditions	2.679	1.337–1.504
5	Habit	3.184	1.529–2.294
6	Hedonic Motivation	3.683	1.489–2.041
7	Perceived Learning Opportunities	2.530	1.960–3.194
8	Performance Expectancy	2.813	1.828–2.556
9	Social Influence	2.262	1.472–1.614
10	Trust	2.131	1.447–2.012

Figure 4. PLS-SEM bootstrapping.

Table 5. Path coefficients and hypotheses testing

Hypotheses	Structural Paths	β	SD	T-values	p-value	2.5%	97.5%	Remarks
H1	AT ➔ BI	−0.144	0.053	2.707	0.007	−0.249	−0.038	S
H2	EE ➔ BI	0.154	0.062	2.474	0.014	0.030	0.270	S
H3	FC ➔ BI	0.131	0.058	2.259	0.024	0.007	0.233	S
H4a	HB ➔ BI	−0.034	0.054	0.626	0.531	−0.150	0.060	NS
H5a	HM ➔ BI	0.187	0.073	2.570	0.010	0.046	0.337	S
H6	PLO ➔ BI	0.156	0.059	2.661	0.008	0.051	0.273	S
H7	PE ➔ BI	0.021	0.055	0.389	0.697	−0.086	0.132	NS
H8	SI ➔ BI	0.257	0.048	5.328	<.001	0.164	0.352	S
H9	T ➔ BI	0.167	0.055	3.031	0.003	0.054	0.269	S

Note: S = Supported and NS = Not Supported.

Table 6. Coefficient of determination, effect size and predictive relevance

Structural Path	R-Square ($R^2$)	R-Square Adjusted	F-Square ($f^2$)	Q-Square ($Q^2$)
AT → BI	0.590	0.579	0.016	0.323
EE → BI			0.020
FC → BI			0.016
HB → BI			0.001
HM → BI			0.023
PLO → BI			0.023
PE → BI			0.000
SI → BI			0.071
T → BI			0.032

Note: $f^2$ = effect size; $R^2$ = R-square; $Q^2$ = predictive relevance.

Table 7. Moderating effect of gender

Hypotheses	Path	Β	SD	T-values	p-value	2.5%	97.5%	Remarks
H4b	Mod_G_HB → BI	−0.018	0.074	0.250	0.802	−0.179	0.118	NS
H5b	Mod_G_HM → BI	0.063	0.088	0.716	0.474	−0.109	0.247	NS

Note: Mod_G_HB → BI = Moderating role of Gender on the relationship between habit and behavioural intention; Mod_G_HB → BI = Moderating role of Gender on the relationship between hedonic motivation and behavioural intention; and NS = Not Supported.

Figure 5. Importance-Performance Map.

Table 8. Total effects and performance Index values for behavioural intention

Behavioural Intention (BI)	Total Effects (Importance)	Index Value (Performance)
AT	−0.144	77.511
EE	0.154	74.272
FC	0.131	73.669
HB	−0.034	72.497
HM	0.187	75.117
PLO	0.156	78.191
PE	0.021	73.850
SI	0.257	71.295
T	0.167	72.526

Figure 6. Revised conceptual framework.

Source: Authors’ construct.

Table

	AT	BI	EE	FC	HB	HM	PLO	PE	SI	T
AT1	0.844	0.485	0.543	0.456	0.408	0.570	0.548	0.482	0.560	0.440
AT2	0.840	0.457	0.548	0.451	0.374	0.593	0.537	0.531	0.545	0.450
AT3	0.775	0.362	0.506	0.500	0.580	0.461	0.637	0.461	0.434	0.414
AT4	0.751	0.383	0.460	0.523	0.508	0.551	0.557	0.433	0.466	0.330
BI1	0.473	0.788	0.490	0.509	0.399	0.509	0.509	0.514	0.508	0.465
BI2	0.411	0.787	0.460	0.485	0.379	0.509	0.432	0.446	0.410	0.494
BI3	0.351	0.751	0.446	0.392	0.250	0.395	0.366	0.380	0.539	0.406
BI4	0.364	0.698	0.397	0.402	0.258	0.461	0.335	0.301	0.465	0.454
EE1	0.458	0.503	0.784	0.530	0.376	0.446	0.479	0.556	0.490	0.519
EE2	0.519	0.507	0.839	0.506	0.402	0.527	0.554	0.617	0.456	0.505
EE3	0.581	0.499	0.857	0.556	0.412	0.493	0.447	0.642	0.480	0.473
EE4	0.552	0.450	0.762	0.429	0.402	0.464	0.397	0.559	0.400	0.447
EE5	0.453	0.403	0.751	0.397	0.336	0.392	0.329	0.640	0.385	0.347
FC1	0.455	0.447	0.425	0.769	0.428	0.446	0.388	0.507	0.463	0.442
FC2	0.463	0.462	0.467	0.772	0.512	0.377	0.481	0.472	0.500	0.396
FC3	0.474	0.453	0.526	0.776	0.454	0.450	0.482	0.492	0.471	0.504
FC4	0.410	0.437	0.430	0.716	0.356	0.566	0.366	0.341	0.489	0.390
HB1	0.448	0.342	0.359	0.435	0.797	0.314	0.444	0.282	0.334	0.446
HB2	0.498	0.362	0.427	0.503	0.888	0.385	0.534	0.412	0.395	0.401
HB3	0.513	0.396	0.452	0.543	0.885	0.476	0.647	0.387	0.412	0.404
HM1	0.543	0.476	0.441	0.445	0.381	0.784	0.398	0.281	0.401	0.379
HM2	0.574	0.507	0.473	0.516	0.373	0.875	0.493	0.475	0.517	0.491
HM3	0.599	0.578	0.552	0.558	0.410	0.866	0.571	0.528	0.504	0.546
PE1	0.527	0.507	0.676	0.454	0.298	0.494	0.423	0.863	0.476	0.492
PE2	0.461	0.459	0.593	0.503	0.368	0.419	0.388	0.871	0.489	0.537
PE3	0.557	0.463	0.667	0.552	0.362	0.446	0.422	0.880	0.485	0.580
PE4	0.478	0.430	0.620	0.539	0.423	0.391	0.407	0.791	0.486	0.489
PLO1	0.666	0.532	0.536	0.537	0.615	0.552	0.887	0.472	0.479	0.446
PLO2	0.557	0.451	0.411	0.479	0.543	0.446	0.880	0.353	0.371	0.407
PLO3	0.609	0.478	0.472	0.477	0.534	0.513	0.906	0.388	0.459	0.436
PLO4	0.607	0.434	0.522	0.478	0.519	0.520	0.810	0.462	0.439	0.435
SI1	0.452	0.539	0.475	0.459	0.276	0.460	0.333	0.465	0.784	0.462
SI2	0.554	0.552	0.432	0.481	0.306	0.471	0.424	0.427	0.805	0.429
SI3	0.426	0.366	0.382	0.491	0.378	0.402	0.356	0.379	0.721	0.381
SI4	0.510	0.474	0.426	0.551	0.451	0.411	0.450	0.483	0.790	0.485
T1	0.268	0.423	0.350	0.396	0.316	0.326	0.217	0.410	0.368	0.791
T2	0.489	0.488	0.533	0.507	0.415	0.492	0.424	0.626	0.560	0.790
T3	0.325	0.488	0.405	0.447	0.397	0.418	0.348	0.378	0.410	0.804
T4	0.347	0.465	0.411	0.362	0.308	0.428	0.426	0.369	0.382	0.769
T5	0.531	0.437	0.517	0.477	0.421	0.502	0.477	0.579	0.455	0.684

Note: Values (in bold) are the reflective indicator loadings obtained in PLS-SEM and the other values are the cross-loadings.

Venkatesh, V., Thong, J. Y., & Xu, X. (2012). Consumer acceptance and use of information technology: Extending the unified theory of acceptance and use of technology. MIS Quarterly, 157–178.

 Web of Science ®Google Scholar

Henseler, J., Ringle, C. M., & Sarstedt, M. (2015). A new criterion for assessing discriminant validity in variance-based structural equation modeling. Journal of the Academy of Marketing Science, 43, 115–135.

 Web of Science ®Google Scholar