ICT-Enabled Education for Sustainability Justice in South East Asian Universities

Abstract

This study aims to investigate the role of Information and Communication Technologies-enabled Education for Sustainability (ICTeEfS), critical reflection, and transformative teaching and learning beliefs in predicting students’ attitudes about seeking sustainability justice. A total of 1497 students from seven universities in Indonesia (374), Malaysia (426), and Vietnam (697) trialed four new scales measuring (a) knowledge of merging ICT with education for sustainability, (b) critical reflective practice, (c) sustainability justice attitudes, and (d) transformative teaching and learning beliefs. The findings show that the four scales are reliable and could be used in other research on education for sustainability. Differences were observed for gender, year of study, subject of study, ICT skills, and knowledge of education for sustainability. Regression analysis highlighted that sustainability justice is a multidimensional concept composed of several constructs with a specific reference to critical reflection, transformative teaching and learning beliefs. The implications for education, practice and further research are discussed.

1. Introduction

In the last few decades, the climate has changed dramatically, caused by actions such as the increasing quantities of carbon that are being released into the atmosphere threatening human existence [1,2,3]. Climate change concerns not only our ecosystems but also undermines our basic rights, increases inequalities, and creates new forms of injustice, turning the climate crisis into an ethical and human rights issue related to justice [4,5,6]. Education has played its role both as a part of and a solution to the climate change crisis. Although environmental education has been integrated into school curricula for more than three decades, human actions regarding climate change and sustainability are still insufficient [7,8]. In particular, the so-called “knowledge–action gap”, “attitude-behavior gap” or even the “value–action gap” have persisted for generations [9,10,11]. An attitude–knowledge–behavior gap concerning sustainability refers to a situation in which people exhibit positive attitudes but fail to merge attitudes with knowledge, reflection, and action. In general, there is an increased need for knowledge and evidence concerning the role of higher institutions in promoting sustainability, including teaching methodologies such as place-based pedagogy; methodologies that give students “voice and choice” to tackle real-life problems and confront sustainability issues in their local environments [12,13,14].

The sustainability challenges have caused humanity to realize that the climate change crisis is not only dependent on techno-scientific solutions and transmissive pedagogies but also on political solutions and above all transformative pedagogies. It is necessary to bring in current discourse concepts, such as climate justice connected to power, human agency, ethics, and human rights. Climate justice means placing equity and human rights in the front line, underlining the responsibility that countries and communities bear concerning the climate crisis [15,16]. There is a need for a shift from discourse on instrumental approaches to tackle climate change to raising socio-political consciousness and civic engagement, especially for people and communities exposed to the impacts of climate change [17,18]. It has been documented that climate change disproportionately affects poor and vulnerable people in both low-income and high-income countries [19,20], turning environmental justice and social justice into an urgent issue locally and globally. This matter has prompted Makrakis [21,22] to advance the concept of sustainability justice in an attempt to integrate the four pillars (environment, society, economy, and culture) of sustainable development into their corresponding dimensions of justice. Justice and sustainability are at the center of ethics and human rights, and as such they are highly contestable and contested [23,24].

Cultivating pro-sustainability justice attitudes is one of the first steps to transforming our world towards a more sustainable and just society enabled by Information and Communication Technologies (ICTs). In this process, the knowledge to contextualize ICT with sustainability along with critical transformative practices and beliefs could help predict sustainability justice attitudes. With the increasing proliferation of ICT across all aspects of social, economic, and cultural life, it is perceived as a powerful enabler for promoting sustainability justice despite the teacher–school–system barriers [25]. In this context, sustainability issues such as poverty (SDG1), food insecurity (SDG2), health accessibility (SDG3), gender inequality (SDG5), and climate change (SDG13) can provide meaningful and challenging opportunities for developing a wide range of ICT skills along with sustainability knowledge. ICT tools such as digital storytelling, concept mapping, simulation, gamification, and social media can help to merge text and context for embedding sustainability justice themes across curricula and teaching methodologies. Previous research shows that gender—as a social construction—is an important factor concerning environmental sustainability attitudes and behaviors [26]. Attitudes influence behavior and there has been a significant development in the field of academic discourse on sustainability issues in the last three decades [27]. It has been documented that gender has an impact on environmental attitudes, values, and beliefs showing that women are more sensitive to nature than men [28]. In terms of academic disciplines and school subjects, the level of environmental attitude was higher in disciplines such as physical sciences and biological sciences rather than in humanities [29]. Another attitudinal study [30] revealed significant differences between students in forestry and students in literature studies showing that students whose discipline was closely related to nature revealed higher levels of environmental attitudes. Diverse academic disciplines exhibit different levels of concern and express different beliefs on issues and attitudes related to environmental sustainability [31]. Therefore, the importance of gender, subjects of study, knowledge, and skills in the field of sustainability justice enabled by ICT, critical reflective practices related to sustainability, and transformative sustainability teaching and learning beliefs should not be underestimated or ignored. These findings could be useful to instructional designers, curriculum developers, and education policymakers.

All these issues were addressed through the ICTeEfS (ICT-enabled Education for Sustainability) project funded by the Erasmus+ CBHE European Commission-funded program carried out in Indonesia, Malaysia, and Vietnam from 2019 to 2023. These three countries are among the Asian countries whose teacher educators cannot address the embedding of sustainability justice in their teaching practices and study programs. In particular, faculties of education need academic teaching staff who can reconstruct their courses and teaching practices to accommodate the contextualization of ICT with EfS and vice versa. They also need to capitalize on the potential of ICT in advancing sustainability justice across all subject areas both in teacher education and school education curricula. Students attending undergraduate teacher education studies have certain experiences, attitudes, and behaviors that are acquired either before and/or during their studies. Uncovering these experiences, practices, attitudes, and beliefs can provide insights for the creation of a teaching and learning environment, and a curriculum conducive to promoting students’ active citizenship for sustainability justice.

Summary of the Literature Review

Our literature review presented in the introduction showed that education for sustainability enabled by ICT (ICTeEfS) is being implemented more and more into higher education curricula at various levels worldwide. As pointed out earlier, ICTeEfS is directly connected to the 17 Sustainable Development Goals (SDGs) and especially to quality education (SDG4) and lifelong learning. The quest for teacher educators and teachers, in general, to embed education for sustainability across all teaching subjects enabled by ICT highly concerns students, especially students educated to become teachers. To date, few researchers have investigated students’ knowledge for contextualizing ICT with education for sustainability, and other related beliefs and practices that are considered important factors in predicting students’ attitudes toward sustainability justice. While it is generally accepted that students’ personal theories in the form of practices, beliefs, dispositions, or attitudes towards a content area such as education for sustainability enabled by ICT are vitally important, limited research has been evidenced either in Southeast Asia or internationally. Therefore, to promote sustainability justice attitudes, it is of paramount importance to carefully examine the factors that affect these attitudes that reflect sustainable development pillars and their corresponding justice dimensions.

Consequently, this study focused on the role of ICTeEfS, critical reflection, and transformative teaching and learning beliefs in predicting students’ attitudes to seeking sustainability justice. In addition to this, the study focused on the impact of gender, subject of study, ICT knowledge, as well as educational knowledge for sustainability on attitudes towards sustainability justice. Considering the findings of previous research and the theoretical framework, the following hypotheses are proposed:

Hypothesis 1 (H1).

Men and women differ in their perceptions ofICTeEfS, critical reflection, sustainability justice, and transformative beliefs concerning teaching and learning.

Hypothesis 2 (H2).

Similarly, the students’ year of study, subject of study, ICT skills, and sustainability knowledge are expected to make a difference to their perceptions of the above constructs.

Hypothesis 3 (H3).

Knowledge of contextualizing ICT with sustainability issues is expected to positively affect attitudes towards sustainability justice.

Hypothesis 4 (H4).

Critical reflection and related practices positively affect sustainability justice attitudes.

Hypothesis 5 (H5).

Transformative teaching and learning beliefs positively affect sustainability justice attitudes.

2. Method

2.1. Participants

The study targeted undergraduate students from seven universities in Indonesia (374), Malaysia (426) and Vietnam (697) whose courses were chosen for revision to embed sustainability justice issues elicited through SDGs. Thus, it was a purposive sample targeted to a specific corpus of students. This decision served the purpose of monitoring student changes during the implementation of the revised courses. In total, 1497 students (65% females and 35% males) responded to the research instrument across all seven partner universities. The authorities responsible for carrying out the survey gave their permission and the students were informed of the anonymity of their responses. Some 60% of the respondents were from the 1st year of study and the remaining 40% from the final year. In terms of the subject area, 87% were related to the fields of education and the remaining 13% to sciences. It has also been revealed that 83% had declared good knowledge of ICT, 2% poor, and 15% minimal.

2.2. Measurements

The survey instrument based on a five-point Likert scale was divided into two sections: the first covered demographic questions such as gender, year of study, subject of study, knowledge of ICT and education for sustainability (EfS), as well as knowledge of EfS, and the second included the following four scales that measured attitudes to the items below. As pointed out in the introduction, the items in these scales were previously piloted in some studies [21,22] and this is the reason why they constitute distinct structures instead of one construct for factorial measurement. The five-point Likert scale used either the level of agreement with a format from “strongly disagree”, “disagree”, “neither agree nor disagree”, “agree” and “strongly agree” or the level of an action “not at all”, “to a slight extent”, “to a moderate extent”, “to a large extent” and to “a great extent”, giving the score of 1 to 5 for both rating formats.

(a)

Knowledge of merging ICT with EfS (ICTeEfS), 9 items;

(b)

Critical reflective practice (CRIREFLECT), 11 items;

(c)

Transformative teaching and learning beliefs (TRANSTEB), 12 items and

(d)

Sustainability justice attitudes (SUSTJUST), 17 items

2.3. Methods of Data Analyses

The data analysis was based on an examination of the internal structure and the reliability of the instrument using principal component analysis through the IBM SPSS^® Exploratory Factor Analysis (EFA) and Cronbach’s alpha (α) reliability test. Descriptive statistics (mean and standard deviation) along with testing mean differences of the four scales against gender, year of study, subject of study, and knowledge of education for sustainable development were calculated using a t-test and one-way analysis of variance. A stepwise regression method was used to identify the strength of the hypothesized predictors on the sustainability justice attitudes. Before running the t-test, one-way ANOVA, and multiple regression analysis, an examination of the statistical conditions to be met such as normality, homogeneity, and multicollinearity was carried out. The P-P plot of normality and Levene’s test of homogeneity of variance were used. In terms of data distribution, the points were followed approximately along the diagonal line, meaning that the data were close to normal distribution. In any case, the sample was very high and any deviations from normality would have a minimal effect. The Levene statistic for testing homogeneity as shown in the results section revealed that the significance value was greater than 0.05 in all tested variables except the ICTeEfS was slightly up to 0.05 level, meaning that the requirement of homogeneity of variance had been met, and the ANOVA test could be considered to be robust.

Since the one-way ANOVA test does not tell us in which of the various pairs of means the difference is statistically significant, the Tukey post hoc test was used to find possible statistically significant differences. Multicollinearity or collinearity was tested by the variance inflation factor (VIF), which shows the strength of correlation between the predictor variables in a regression model as well as the tolerance which shows the strength of the beta coefficients’ impact. Generally, a VIF above 4 (some argue above 10) or tolerance below 0.25 shows that multicollinearity might be present, and thus further investigation is required [32,33]. As pointed out in the results section both the VIF values for the predictor variables and tolerance indicate that there is no problem with multicollinearity in the regression model.

3. Results

3.1. Factorial Analyses

In this section, the results of the exploratory factor analyses (EFA) are reported, addressing the four constructed scales measuring ICTeEfS, critical reflective practice, beliefs in transformative teaching, and attitudes toward sustainability justice.

3.1.1. Exploratory Factorial Analysis of the ICTeEfS Scale

The Kaiser–Meyer–Olkin (KMO) value was 0.941, a measure that shows a very good sample sufficiency and adequacy. Bartlett’s test of sphericity was relevant (χ² = 8407.47; df: 36; Sig. p < 0.01), showing that the ICTeEfS-composed items were significantly correlated and suitable for further statistical analyses. The exploratory factor analysis resulted in a one-factor solution applying a direct varimax rotation method and eigenvalues > 1.0 extracting a considerable amount of variance (64.63%). The one-factor solution seemed to be very well-defined and interpretable in terms of theoretical reliability and construct relevance with an eigenvalue of 5.82. All nine items included in the ICTeEFS scale were retained for their substantial loadings (>0.39), as can be seen in

Table 1. Items’ saturation for the ICTeEfS composite variable.

Items—I Have Learned to	Saturation
1. Use interactive methods enabled by ICT to discuss local sustainability issues.	0.67
2. Use ICT to engage future learners in studying local and global issues.	0.81
3. Use ICT in ways that could strengthen future learners’ outdoor classroom learning activities.	0.84
4. Use ICT in ways that could promote learners’ knowledge of local environmental problems.	0.85
5. Use various learning resources and tools related to education for sustainable development with the support of ICT.	0.67
6. Use learners’ real-life experiences to develop their digital/ICT knowledge and skills.	0.76
7. Use ICT as a means to integrate future learners’ experiences in addressing sustainability.	0.79
8. Adjust educational content in ways that make it relevant to my future learners’ lives outside the school environment.	0.79
9. Apply ICT in ways that could enhance learners’ active engagement in solving real-life problems.	0.80

. The Cronbach’s alpha reliability test of the scale revealed a value of 0.93 which shows a very good index and very good internal consistency. The ICTeEfS rotation matrix is depicted in Table 1. The final model obtained through EFA has a good fit and contains nine items with factor loadings ranging between 0.67 and 0.85.

3.1.2. Exploratory Factorial Analysis of CRIREFLECT

The Kaiser–Meyer–Olkin (KMO) value was 0.93, indicating very good sample sufficiency and adequacy. Bartlett’s test of sphericity was (χ² = 6911,95; df: 55; Sig. p < 0.01), indicating that the items were significantly correlated and that the results were suitable for further analyses. A single factor was obtained, accounting for a considerable amount of variance (50.39%). The EFA yielded a one-factor solution with a direct varimax rotation method and eigenvalues > 1.0. The obtained one-factor structure was very well defined and interpretable with theoretical reliability and construct relevance and an eigenvalue of 5.54. All eleven items were retained for their substantial loadings (>0.39) on a single factor and the single factor structure was interpreted as satisfactory factor loading and theoretical relevance of all items to the factor. The Cronbach’s alpha of the scale was calculated and the value of 0.90 was obtained, indicating very good reliability and very good internal consistency. The rotation matrix is reported in

Table 2. Items’ saturation for critical reflective practice.

Items	Saturation
1. I like to think over what I have already done and find alternative ways of doing things.	0.59
2. I often reflect on what I do and explore ways of doing things better.	0.71
3. I often reflect on my actions to improve myself.	0.73
4. I often step back and reflect on what I am thinking and doing.	0.73
5. I frequently think about whether I am doing wrong or right.	0.58
6. Reflection has helped me to change my way of thinking and behaving.	0.68
7. Reflecting on what I was thinking and how I was behaving has challenged my values and actions.	0.66
8. Critical reflection has motivated me to change.	0.74
9. Due to reflection, I discovered faults in what I believed to be right which prompted me to change.	0.68
10. Reflecting on previous thoughts and beliefs has motivated me not to take things for granted.	0.66
11. Reflecting critically on other people’s actions and behaviors has increased my sense of co-responsibility.	0.64

. The final model obtained through EFA has a good fit and contains eleven items. The factor loadings of the items ranged between 0.58 and 0.74.

3.1.3. Exploratory Factorial Analysis of the TRANSTEB Scale

The Kaiser–Meyer–Olkin (KMO) value was 0.907, indicating very good sample sufficiency and adequacy. Bartlett’s test of sphericity was (χ² = 6720.899; df: 66; Sig. p < 0.01), indicating that the items were significantly correlated and that the results were suitable for further analyses. The EFA was forced to a one-factor solution with a direct varimax rotation method accounting for a considerable amount of variance (43.85%). The one-factor structure was very well defined and interpretable with theoretical reliability and construct relevance with an eigenvalue of 5.262. Twelve items were retained for their substantial loadings (>0.39) on a single factor, as shown in Table 1, while four items were discarded. A single-factor structure was interpreted as satisfactory factor loading and theoretical relevance of the items to the factor. The Cronbach’s alpha of the scale was 0.87 indicating good reliability and internal consistency. The rotated matrix of the scale is presented in

Table 3. Items’ saturation for transformative teaching and learning beliefs.

Items	Saturation
1. I believe that learners need skills that go far beyond what they have learned at school.	0.58
2. I believe that learners love learning but are oppressed with sterile knowledge.	0.47
3. I believe that learners have a voice and should be heard.	0.73
4. I believe that learners can make a difference.	0.74
5. I believe that I can get more from my future learners than I can teach them.	0.63
6. I will do my best to make my future learners more independent in their learning.	0.73
7. I believe that the teacher should fight injustice even at the risk of losing his/her job.	0.60
8. I believe in the advisory role of the teacher.	0.71
9. I believe that the teacher should bring the school to society and society to school.	0.77
10. I believe that the teacher should talk to his/her learners about the causes of an unsustainable world.	0.72
11. I believe that the teacher should aim to teach the transformation of self and society.	0.75
12. I believe that the teacher should see the curriculum as a fixed and prescribed package of knowledge.	0.41

. The final model obtained through EFA has a good fit and contains twelve items. The factor loadings of the items ranged between 0.41 and 0.77.

3.1.4. Exploratory Factorial Analysis of the SUSTJUST Scale

The KMO value was 0.91, indicating very good sample sufficiency and adequacy. Bartlett’s test of sphericity was (χ² = 6720.90; df: 66; Sig. p < 0.01), indicating that the items were significantly correlated and that the results were suitable for further analyses. The EFA was forced at a one-factor solution with a direct varimax rotation method accounting for the following amount of variance: 33.91%. The one-factor structure was well defined and interpretable with theoretical reliability and construct relevance with an eigenvalue of 5.76. Seventeen items were retained for their substantial loadings (>0.39) on a single factor, as shown in Table 1, while three items were discarded. A single-factor structure was interpreted as satisfactory factor loading and theoretical relevance of the items to the factor. The Cronbach’s alpha of the scale was calculated and the value of 0.87 was obtained indicating good reliability and internal consistency. The rotated matrix of the scale is reported in

Table 4. Items’ saturation for the sustainability justice attitudes.

Items	Saturation
1. When I see injustice and do nothing about it, I feel guilty.	0.47
2. Helping those less fortunate than me will change my life.	0.56
3. I appreciate the presence of refugees and immigrants in my city.	0.40
4. All groups should feel the need and the right to express their differences in public.	0.53
5. Resources should be distributed so that everyone can live a decent life.	0.61
6. Members of different groups should be treated equally	0.67
7. An alternative to the minimum wage could be a salary that sustains a decent standard of living.	0.53
8. All people must have the opportunity to participate in decisions that can affect them.	0.63
9. Economic activity should be within the bounds of morality.	0.60
10. It is my life’s work to struggle to build a just and sustainable world.	0.64
11. I can make a difference for a better future.	0.59
12. I feel that we have to claim the common good through our actions on an individual and collective level.	0.69
13. The marginalized people must have their voice in public affairs.	0.66
14. The interests of today’s generations should not be paramount over those who will be born in 50 years’ time.	0.39
15. We must respect the cultural heritage of others as well as our own.	0.67
16. The rights to the goods of nature belong to everyone.	0.58
17. Economic prosperity is not an indicator of the wealth we hold, but whether this wealth is distributed fairly.	0.59

. The final model obtained through EFA has a good fit model and contains seventeen items. The factor loadings of the items ranged between 0.39 and 0.69 (Table 4).

3.2. Mean Differences (t-Test and One-Way ANOVAs)

Achieving gender equity is critical to education for sustainability and sustainable development in general. As shown in

Table 5. Mean differences by gender. Significant differences are shown in bold.

Composite Variables	Gender	N	Mean	SD	Levene’s Test		t-Test for Means Equality
					F	Sig.	t	Diff	Sig.
ICTeEfS	Male	502	2.99	0.83	0.154	0.695	−0.251	−0.01	0.802
	Female	882	3.00	0.87
CRIREFLECT	Male	499	3.65	0.56	3.02	0.083	0.120	0.01	0.905
	Female	936	3.64	0.80
SUSTJUST	Male	507	3.85	0.51	2.36	0.125	−3.56	−0.10	0.000
	Female	934	3.95	0.47
TRANSTB	Male	511	3.90	0.56	0.018	0.893	−2.58	−0.08	0.010
	Female	958	3.98	0.55

Note: ICTeEfS stands for ICT-enabled education for sustainability, CRIREFLECT (critical reflection), SUSTJUST (sustainability justice), and TRANSTB (transformative teaching and learning beliefs).

, there was a statistically significant gender difference with male students showing lower scores (mean = 3.85, SD = 0.51) than female students (mean = 3.95, SD = 0.47) in their sustainability justice attitudes (t(1441) = −3.56., p = 0.000). Similarly, female students (mean = 3.98, SD = 0.55) scored higher than male students (mean = 3.90, SD = 0.56) in their beliefs about transformative teaching and learning to address education for sustainability (t(1469) = −2.58, p = 0.010). However, there was no statistically significant difference in terms of gender concerning knowledge in enabling education for sustainability through ICT (p = 0.80) and in terms of critical reflective practices (p = 0.90).

In terms of the student study year (

Table 6. Mean differences by study year. Significant differences are shown in bold.

Composite Variables	Study Year	N	Mean	SD	Levene’s Test		t-Test for Means Equality
					F	Sig.	t	Diff	Sig.
ICTeEfS	1st	851	2.88	0.88	7.77	0.005	−6.4	−0.30	0.000
	4th	532	3.18	0.78
CRIREFLECT	1st	844	3.60	0.59	0.207	0.650	3.0	−0.10	0.002
	4th	590	3.70	0.58
SUSTJUST	1st	858	3.89	0.50	0.410	0.522	−1.8	−0.05	0.067
	4th	588	3.94	0.54
TRANSTB	1st	891	3.92	0.54	0.832	0.362	−2.7	−0.08	0.006
	4th	586	4.00	0.56

), there were statistically significant differences in all tested scales except sustainability justice attitudes. As was expected, first-year students (mean = 2.88, SD = 0.88) scored less than fourth-year students (mean = 3.18, SD = 0.78) concerning their knowledge in ICT-enabling education for sustainability (t(1383) = −6.40, p = 0.000). Similar trends were found for their critical reflective practices (t(1434) = −3.0, p = 0.002) and in their beliefs for transformative teaching and learning (t(1477) = −2.70, p = 0.006). However, there was no statistically significant difference in terms of the study concerning students’ attitudes toward sustainability justice (p = 0.067, higher than p = 0.050).

The findings presented in

Table 7. One-way ANOVA results by subjects of study. Significant differences shown in bold.

Composite Variables by Subject of Study		Sum of Squares	Df	Mean Square	F	Sig.
ICTeEfS	Between groups	20.648	6	3.441	4.74	0.000
	Within groups	999.103	1377	0.726
	Total	1019.751	1383
CRIREFLECT	Between groups	4.116	6	0.686	1.95	0.069
	Within groups	501.584	1428	0.351
	Total	505.700	1434
SUSTJUST	Between groups	4.047	6	0.675	2.78	0.011
	Within groups	348.471	1440	0.242
	Total	352.518	1446
TRANSTEB	Between groups	1.859	6	0.310	1.02	0.410
	Within groups	443.739	1462	0.304
	Total	445.598	1468

show that there were statistically significant differences between the subject areas, including (1) education, (2) applied sciences, (3) humanities, (4) technology, (5) biology, (6) computer sciences, and (7) engineering, as demonstrated by one-way ANOVA (F(6,1377) = 4.74, p = 0.000) for knowledge enabling the contextualization of ICT with education for sustainability and sustainability justice attitudes (F(6,1440) = 2.78, p = 0.011). A Tukey HSD post hoc test showed that the significant difference was located between groups 4 (technological subjects) and 5 (biology), with a mean difference of 0.61, at p = 0.02, and between groups 5 and 7 (civil engineering) with a mean difference of 0.42, at p = 0.01. The statistically significant difference for sustainability justice attitudes in terms of subject areas was located between the groups 3 (humanities) and 4, with a mean difference −0.44, at p = 0.018. There was no statistically significant difference between the subjects for critical reflective practices (p = 0.067) and transformative teaching and learning beliefs (p = 0.41).

In terms of the students’ knowledge and skills in ICT measured on a 5-point scale from poor to excellent, the results shown in

Table 8. One-way ANOVA results by ICT skills. In bold significant differences.

Composite Variables by ICT Skills		Sum of Squares	df	Mean Square	F	Sig.
ICTeEfS	Between groups	81.030	4	20.257	29.76	0.000
	Within groups	938.722	1379	0.681
	Total	1019.752	1383
CRIREFLECT	Between groups	28.830	4	7.208	21.61	0.000
	Within groups	476.870	1430	0.336
	Total	505.700	1434
SUSTJUST	Between groups	2.260	4	0.565	2.33	0.054
	Within groups	350.258	1442	0.243
	Total	352.518	1446
TRANSTEB	Between groups	3.011	4	0.753	2.49	0.042
	Within groups	442.588	1464	0.302
	Total	445.599	1468

revealed statistically significant differences as demonstrated by one-way ANOVA (F(4,1379) = 20.25, p =0.000) for contextualizing ICT with EfS; for critical reflective practices (F(4,1430) = 21.61, p = 0.000) and transformative teaching and learning beliefs (F(4,1464) = 2.49, p = 0.042). Lastly, the Tukey HSD multiple comparison test revealed that students’ beliefs in transformative teaching and learning in terms of their ICT skills were statistically different in group 1 (poor knowledge) and sequentially in group 2 (minimal) to group 3 (sufficient), group 4 (above average) and 5 (excellent) at p > 0.05 with mean differences ranging from −0.292 to −0.313. It is worth pointing out that the distribution of frequencies in terms of knowledge of ICT ranged from 2% (poor) to 15% (minimal), 46% (sufficient), 31% (above average), and 6% (excellent).

Finally, in terms of the students’ knowledge and skills in education for sustainability,

Table 9. One-way ANOVA results from knowledge of education for sustainability. Significant differences are shown in bold.

Compositive Variables by Knowledge of Education for Sustainability		Sum of Squares	df	Mean Square	F	Sig.
ICTeEfS	Between groups	104.405	4	26.101	39.32	0.000
	Within groups	915.347	1379	0.664
	Total	1019.752	1383
CRIREFLECT	Between groups	25.468	4	6.367	18.95	0.000
	Within groups	480.231	1430	0.336
	Total	505.700	1434
SUSTJUST	Between groups	1.313	4	0.328	13.5	0.250
	Within groups	351.205	1442	0.244
	Total	352.518	1446
TRANSTEB	Between groups	1.088	4	0.272	0.896	0.466
	Within groups	444.511	1464	0.304
	Total	445.599	1468

reveals statistically significant differences as demonstrated by one-way ANOVA (F(4,1379) = 39.32, p = 0.000) for knowledge of ICT in EfS and (F(4,1430) = 18.95, p = 0.000) for critical reflective practices. The Tukey HSD multiple comparison test revealed that the students’ knowledge and skills on education for sustainability were statistically different in group 1 and sequentially to group 5 at p > 0.05 with mean differences ranging from −0.34 to −1.12, showing that the higher the knowledge and skills, the higher the competence in contextualizing ICT with EfS and practicing critical reflection.

3.3. Stepwise Regression Analysis

The stepwise regression method was applied to identify the strength of the relationship between the sustainability justice attitudes considered as the dependent variable and the eight predictors, namely gender, year of study, subject of study, ICT skills, EfS knowledge, knowledge on contextualizing ICT with EfS, critical reflection and beliefs in transformative teaching. A final five-factor prediction model (

Table 10. Stepwise regression analysis.

Model	Predictors	R2adj	Std. Error	Change Statistics				Collinearity Statistics
				R2change	F Change	Df1/2	Sig.	Tol	VIF
1	TRANSTEB	0.388	0.372	0.388	810.19	1/1280	0.000	0.864	1.16
2	CRIREFLECTION	0.447	0.354	0.060	138.40	1/1279	0.000	0.753	1.32
3	GENDER	0.449	0.353	0.002	4.03	1/1278	0.045	0.991	1.00
4	STUSUBJECT	0.451	0.353	0.002	4.09	1/1277	0.043	0.981	1.02
5	ICTeEfS	0.454	0.352	0.002	4.76	1/1276	0.029	0.832	1.20

1. Predictors (constant-TRANSTEB), 2. predictors (const-TRANSTEB, CRIREFLECT), 3. predictors (const-TRANSTEB, CRIREFLECT, GENDER), 4. predictors (const-TRANSTEB, CRIREFLECT, GENDER, STUSUBJECT), 5. predictors (const-TRANSTEB, CRIREFLECT, GENDER, STUSUBJECT, ICTeEfS).

) emerged from the regression analysis, excluding year of study, ICT skills, and knowledge of education for sustainability that did not meet the probability threshold (p > 0.050). The coefficient of determination achieved was R²_adj = 45.4 with F (8, 1273) = 132.36, p = 0.000, explaining 45.4% of the variability in the sustainability justice score. More specifically, the stepwise regression analysis revealed that students’ beliefs in transformative teaching and learning alone explained 38.8% of their attitudes to sustainability justice (R²_change = 0.388, F_change (1,1280) = 810.19. at p = 0.000), followed by critical reflective practices accounting for 6% (R²_change = 0.060, F_change (1,1279) = 138.40 at p = 0.000). The remaining three predictors, gender, subject area and knowledge on contextualizing ICT with education for sustainability accounted for a minimal but statistically significant percentage. The individual scores are reflected by the beta values presented in

Table 11. Regression statistics.

Dependent Variable—SUSTJUST MODEL	Unstandardized Coefficients		Stand. Coeff.	t	Sig.
	B	Std. Error	Beta
1 (Constant)	1.258	0.095		13.21	0.000
Gender	0.043	0.021	0.043	2.059	0.040
Year of Study	−0.031	0.021	−0.032	−1.49	0.135
Subject of study	0.018	0.008	0.049	2.355	0.019
ICT knowledge and skills	−0.008	0.014	−0.015	−0.602	0.547
Sustainability knowledge and skills	0.005	0.014	0.009	0.376	0.707
ICT-enabled Edu. for Sustainability	0.031	0.013	0.055	2.318	0.021
Critical reflective practice	0.196	0.019	0.244	10.089	0.000
Transformative teaching beliefs	0.458	0.020	0.524	23.333	0.000
Model	R	R²	Adj. R²	Std. Er.	D-W
1 Model Summary Statistics	0.674	0.454	0.451	0.353	1.95
ANOVA	Sum of Squares	Df	Mean Square	F	Sig.
Regression Residual Total	131.644 158.256 289.900	8 1273	16.45 0.124	132.367	0.000

1. Dependent variable: sustainability justice attitudes, 2. Predictors: (constant), transformative teaching beliefs; education for sustainability knowledge; the subject of study; gender; year of study; ICT-enabled education for sustainability knowledge; critical reflective practices and ICT knowledge and skills.

. In general, Beta weight is the average amount by which the dependent variable increases when the independent variable increases one standard deviation and other independent variables are held constant.

4. Discussion

The present study aimed to use the method of exploratory factor analysis to investigate the factorial structure of the four key scales of the study, namely: ICT-enabled education for sustainability knowledge; critical reflective practices; transformative teaching and learning beliefs, and sustainability justice attitudes in a Southeast Asian purposive sample of undergraduate students. The EFA seemed to mainly favor a one-factor solution across all four scales, which was shown to explain over 70% of the variance. In cases of the one-factor solution, all of the measurement items fit onto a single theoretical construct and one dimension/scale. The reliability was further measured through the intercorrelations of scale items, and item–total-correlations, using Cronbach’s α, which is the most important reliability index supplemented by the Bartlett test of sphericity along with the KMO test (Kaiser–Meyer–Olkin measure of sampling adequacy) that examined sample sufficiency. The items in the four scales are consistent with each other and measure the same property and the four scales developed in the present study could be used by researchers in the field of ICTeEfS as well as in other relevant research areas. It would be interesting to use the scales in different cultural contexts and further examine their validity and theoretical value.

According to the findings, the research results show that in most cases, the female students had a more favorable attitude to education for sustainability enabled by ICT (ICTeEfS) compared to the males, proving the first hypothesis. In general, these results are consistently substantiated by previous research [34,35]. As seen in the literature review, women are seen to be more pro-environmental than men. This is explained by previous research which shows that women are motivated by eudaimonic happiness compared to men who are more motivated by hedonic happiness [36]. Although an ancient concept originating from Aristotle [37], eudaimonia has reemerged in modern psychology because despite capitalism with its consumer-based economies having led to increasing material wealth, it has not necessarily yielded the expected rises in happiness or well-being [38,39,40].

In general, females hold more pro-environmental attitudes and behaviors compared to males who seem to disassociate themselves from issues related to sustainable development [41], despite some opposite results [28]. It is worth pointing out that gender-based differences are largely driven by societal norms and expectations and have nothing to do with genetics [42]. There is a need for further studies to identify the motivational factors related to gender differences on sustainability issues that could inform policy and practice. The quest for sustainability justice must not be driven by gender or socio-cultural background.

Regarding the second hypothesis dealing with the year of study, the research results show, as expected, the fourth-year students scored higher than the first-year students in terms of critical reflective practices (t(1434) = −3.0, p = 0.002) and in their beliefs about transformative teaching and learning to address education for sustainability. However, no statistically significant difference was found in terms of the year of study concerning students’ attitudes toward sustainability justice. The third hypothesis examined possible differences in terms of the student’s subject of study. The results revealed that students belonging to environment-related subjects such as biology, were more favorable in their sustainability justice attitudes. These results have been substantiated by previous research [30]. However, there was no statistically significant difference between the subjects for critical reflective practices (p = 0.067) and transformative teaching beliefs (p = 0.41). Attitudes towards the environment usually drive decisions about environmental sustainability. It is evident from previous research that, during the last decade, the number of academic disciplines that have been reoriented to address sustainability issues has been steadily increasing. In general, there is a trend that higher education must respond to the quest for sustainability providing students with the necessary knowledge and skills to cope with the uncertainties and complexities of an unsustainable world that they will meet after their graduation [43,44,45].

It was also found that the subjects that are more education-oriented dominate in reorienting their study programs to address sustainability due to their pedagogical readiness, which is seen as a key driver in addressing EfS in multiple academic disciplines [46]. These results suggest that higher education institutions have to pay attention to embedding sustainability in multiple academic subjects emphasizing an experiential and multi-disciplinary curriculum development approach aligned with the four pillars of sustainable development and their corresponding justice dimensions. UNESCO [47] suggests several ways for EfS or ESD infusion or integration, including a thematic approach like climate change, air pollution, and deforestation, consistent with proper teaching methodologies and based on local realities. Infusing or embedding EfS into existing courses through proper revision enabled by ICT has been initiated by various regional initiatives and projects funded by the European Commission [48,49].

In terms of students’ knowledge of and skills in ICT, the results revealed statistically significant differences showing that the students’ knowledge and skills in ICT are positively related to critical reflective practices and competence to contextualize ICT with education for sustainability as well as for holding transformative teaching and learning beliefs. Similar trends were evidenced concerning students’ knowledge and skills in education for sustainability, meaning that as skills increase so does their knowledge of merging ICT with education for sustainability.

The regression analysis revealed that the highest explained variance of the students’ sustainability justice attitudes was attributed to their beliefs in transformative teaching and learning, in agreement with the stated hypothesis. Therefore, we advocate that professors and other academic instructors should give more explicit emphasis to transformative pedagogy. Such beliefs assume that students should be encouraged to go through their path of learning and be equipped with knowledge and skills that comply with societal needs. In addition, learners have to have a voice and should be heard, and learners can make a difference. Other beliefs stress that teachers should fight sustainability injustices even at the risk of losing benefits or rewards from work and that the teacher should redirect teaching to address societal needs. Moving beyond behavioristic and instrumental approaches to climate change education necessitates a paradigm shift in how the education stakeholders, including teachers, students, and education leaders think about their role in tackling the climate crisis. Other lessons learned through this study support the promotion of transformative sustainability justice enabled by ICT, especially in teacher education. Added to that, higher education institutions worldwide should reconstruct their study programs to address action-oriented teaching and curriculum practices, as well as community-based learning and engagement.

The current research has limitations due to the self-report instruments used to collect the data. Answers to such kind of tools could have a social desirability bias, that could impact the findings and should be considered while discussing the results. There is a need to stress the importance of experiential learning in future research designs, as a supplement to illuminate the quantitative research findings. Another limitation lies in the fact that several questionnaires were developed for this study. The tools’ structure was tested for the first time during this survey and, for two scales, some items were discarded. The newly constructed scales have to be tested in different contexts and with different participants. Another problem is connected to the exploratory nature of the research, which includes demographic information to maximize the data. Data such as gender can provide a sketch of the social profile of the students, which could be useful in an exploratory study but should be further discussed. However, the imbalance of gender (35% males and 65% females) with its consequent skewness suggests the interpretation of the gender results should be treated with caution. Finally, a critical point relates to the possible bias that characterizes the groups of participants, which were not necessarily balanced for gender, nationality, and subject of study.

5. Conclusions

Despite limitations, the current study provided a better understanding of the nature of sustainability justice, which appears to be a multidimensional concept composed of several constructs with specific reference to critical reflection and transformative teaching and learning beliefs. The current research involved university students from Asian higher education institutions and it would be interesting, in a future study, to explore whether recruiting students from different continents would produce different results. In addition, it would be nice to test the relevance of other quantitative scales assessing different constructs such as attitudes towards education for sustainable development (ESD) [50,51]. Even though increasing volumes of research focus on issues related to education for sustainability, few researchers have attempted to study the enabling role of ICT in teaching and learning for sustainability. Moreover, the current research further advances the justice dimensions of sustainability and its determinants, adding significantly to the education for sustainability research endeavors.

As demonstrated in the discussion, the results of this study and their interpretation have been highly echoed in previous research, along with ICT-enabled education for sustainability and critical reflective practices, which proved to be the critical explanatory factors that added to the prediction of the students’ sustainability justice attitudes. Through critical reflective practices mediated by transformative teaching and learning, the unstated assumptions that preserve unfair sustainability injustices may be changed. Critical reflection along with the contextualization of ICT with education for sustainability could inspire students and teachers to explore and experiment with current sustainability injustices, to evaluate our notions of who we are, what our identity is, what we believe, and how to join efforts for quality education in line with the UN 2030 Agenda for building a more sustainable and just society.