The Role of Higher Education Institutions in Shaping Sustainability and Digital Ethics in the Era of Industry 5.0: Universities as Incubators of Future Skills

Abstract

The transition toward human-centered innovation models, as reflected in Industry 5.0 frameworks, calls for the integration of sustainability and digital ethics into higher education. Despite the growing international discourse, little is known about how systematically these dimensions are embedded in curricula in Central and Eastern Europe. This study addresses this gap by analyzing the extent to which Polish higher education institutions (HEIs) incorporate elements of sustainable development and digital ethics into their educational programs. Drawing on survey data from 187 Polish HEIs, we employed Cramér’s V and chi-square tests to explore bivariate associations, multiple correspondence analysis (MCA) to examine patterns among categorical variables, and ordinal logistic regression to identify key predictors of curricular integration. The results reveal that institutions offering Industry 5.0-oriented specializations and maintaining regular cooperation with enterprises are significantly more likely to achieve full integration of sustainability and ethics, whereas many others remain at a stage of only partial adoption. These findings underscore the uneven progress of curricular reforms and highlight the importance of institutional capacity and external partnerships. This study contributes to theory by extending institutional and resource-based perspectives to curriculum innovation, and it contributes to practice by recommending targeted accreditation standards, cross-sector partnerships, and interdisciplinary modules (e.g., “Artificial Intelligence and Society,” “Sustainable Technology Futures”) as concrete mechanisms for embedding ethical and sustainable innovation competencies in higher education. Implications for policy, institutional practice, and future research are discussed.

1. Introduction

The ongoing digital and ecological transformation of society has brought higher education to a pivotal point. Faced with growing complexity, interconnectivity, and technological disruption, universities are increasingly expected to equip students with more than just technical know-how. They are called upon to develop competencies aligned with human-centric values, ethical reflection, and sustainable development—key tenets of the emergent paradigm of Industry 5.0 (also referred to as Economy 5.0 or Society 5.0) [1,2,3].

Industry 5.0 reflects a fundamental shift in how innovation is conceptualized and implemented. In contrast to the productivity-oriented logic of Industry 4.0, this new model promotes a vision of progress centered on social cohesion, environmental protection, resilience, and ethical integration of advanced technologies. Whether referred to in the context of Industry 5.0 or Society 5.0, the paradigm emphasizes the co-evolution of human and technological systems in ways that advance the common good [4,5].

Within this framework, higher education institutions (HEIs) are envisioned as incubators of future-ready skills—places where young people not only acquire technical competencies but also learn to navigate the normative and sustainability dimensions of digital transformation [6,7]. This is particularly relevant in relation to two key domains: sustainable development (as reflected in the United Nations’ Sustainable Development Goals—SDGs) and digital ethics, which concerns the responsible use of technologies like artificial intelligence (AI), internet of things (IoT), big data, and automation [8,9].

Despite the growing literature on Industry 5.0 and sustainability education, little is known about how systematically these elements are embedded in the curricula of higher education institutions in Central and Eastern Europe. Previous studies often focus on case-based examples or selected disciplines, leaving a gap in system-wide analyses [10,11]. Addressing this research gap, our study investigates Polish HEIs, asking to what extent they integrate sustainability and digital ethics into their programs of study and what institutional factors condition this integration.

The main objective of this article is to assess the presence and determinants of curricular elements related to sustainable development and digital ethics in Polish higher education institutions. Specifically, the study investigates which institutional characteristics influence the integration of these Industry 5.0-aligned themes into educational programs. In particular, we examine four institutional characteristics that, according to prior theoretical work, may shape curriculum design: (a) type of institution (e.g., university, technical, pedagogical, vocational), (b) cooperation with enterprises in digital and innovation-related projects, (c) presence of digital infrastructure or specialized units, and (d) existence of Industry 5.0–related specializations or course content.

The contributions of this article are twofold. Theoretically, we extend institutional and resource-based perspectives to the field of curriculum studies by showing how structural and strategic features of universities condition their ability to implement Industry 5.0 competencies. Practically, we provide recommendations for policymakers and university leaders, including accreditation standards, cross-sector partnerships, and interdisciplinary modules that operationalize sustainability and ethics in education. These contributions move beyond identifying gaps to offering concrete strategies for embedding Industry 5.0 values into higher education.

This article is structured as follows: it begins with a theoretical overview of sustainability and digital ethics in the context of Industry 5.0, followed by the formulation of research questions and hypotheses. The methodological section outlines the data collection and a three-stage analytical process. The results highlight how institutional characteristics relate to curriculum content. The discussion interprets these findings in light of international literature, offers recommendations for higher education, and reflects on the study’s limitations. The final section summarizes key contributions and broader implications for policy and practice.

2. Educating for the Future: Embedding Sustainability and Digital Ethics in the Era of Industry 5.0

2.1. Industry 5.0 and Sustainable Development: The Role of Higher Education in Promoting SDG-Oriented Competencies

The challenges posed by climate change, social inequality, resource depletion, and rapid technological advancement call for a new approach to education that prepares future professionals to act responsibly in a complex and interconnected world. Higher education institutions are increasingly expected not only to deliver technical expertise but also to develop students’ ethical reasoning, sustainability literacy, and digital responsibility. These demands align closely with the principles of Industry 5.0, a paradigm that emphasizes human-centricity, resilience, and sustainability in the next phase of industrial transformation [12,13].

Industry 5.0 expands upon the technological foundations of Industry 4.0 by integrating ethical and societal considerations into innovation processes. It envisions a future in which humans and machines collaborate to create inclusive, resilient, and sustainable societies [14]. Importantly, Industry 5.0 directly supports several Sustainable Development Goals, especially SDG 9 (Industry, Innovation, and Infrastructure), SDG 8 (Decent Work and Economic Growth), and SDG 13 (Climate Action) [15,16,17].

However, scholars have also drawn attention to the risks of this paradigm. Ozdemir and Hekim [18] highlight that Industry 5.0 discourses may risk “greenwashing” if sustainability is not embedded structurally. Alazab and Alhyari [19] warn that educational reforms linked to SDGs often reproduce inequalities, leaving behind marginalized groups. Maddikunta et al. [20] emphasize tensions between human-centrism and efficiency, pointing to contradictions in how Industry 5.0 is framed. Including such critical perspectives ensures that Industry 5.0 is not presented as a uniformly positive development but as a contested field that requires careful interrogation.

Higher education plays a strategic role in equipping graduates with the competencies needed to implement and scale these transformations. For example, programs that integrate AI, robotics, and sustainable innovation can prepare students to work in sectors such as precision agriculture [21], healthcare 5.0 [22], and smart energy systems [23]. These sectors not only represent future labor market opportunities but also directly contribute to global goals.

The integration of Industry 5.0 technologies with sustainability principles is already showing tangible results in diverse domains. In healthcare, for instance, technologies such as exoskeletons, AI-driven diagnostics, and IoT-based monitoring are improving service quality while reducing costs [24,25,26]. In urban development, smart city applications are leveraging Industry 5.0 tools to optimize traffic, energy usage, and disaster response, thus advancing SDG 11 (Sustainable Cities and Communities) [27,28].

From an educational perspective, introducing students to the ethical, social, and ecological implications of such technologies becomes imperative. The SDGs offer a ready-made framework through which to operationalize this knowledge, ensuring that curricula foster systems thinking, intergenerational responsibility, and global citizenship [29,30].

Furthermore, digital tools such as augmented and virtual reality (AR/VR), digital twins, and AI-powered simulations enable experiential learning that makes sustainability tangible and actionable [31,32]. These tools are especially powerful when used in interdisciplinary courses that combine engineering, ethics, economics, and policy—hallmarks of a truly holistic approach to sustainability education.

The literature increasingly emphasizes that Industry 5.0 technologies can support not just economic development but also social inclusion and environmental protection. For example, applications of AI in agriculture are improving food security and reducing waste [33,34], while machine learning algorithms are being deployed to detect water pollution and manage infrastructure sustainably [35,36]. Similarly, blockchain technologies promote transparency and accountability in public services, contributing to SDG 16 (Peace, Justice, and Strong Institutions) [37].

In this context, HEIs are increasingly viewed as “SDG incubators” [38], responsible not only for educating but also for modeling sustainable and ethical behavior. This requires integrating sustainability goals into institutional strategies, research priorities, and teaching practices [39]. Courses focused on responsible innovation, ethics of AI, and sustainable design thinking are particularly relevant in equipping students for leadership in Industry 5.0 contexts [40,41].

To ensure effectiveness, such initiatives must be aligned with the challenges of the local and global context, including labor market needs and cultural norms. As argued by Liu et al. [42], co-bots and smart automation are transforming the workplace, and universities must prepare graduates for human–machine collaboration while critically engaging with the implications of job displacement and digital inequality.

To better understand why some institutions are more capable of integrating sustainability, we apply theoretical frameworks. Institutional Theory [43] explains how accreditation standards, societal expectations, and isomorphic pressures shape university practices. The Resource-Based View [44] highlights that HEIs with greater resources, stronger partnerships, and specialized curricula are better positioned to embed transversal competencies. Finally, Education for Sustainable Development (ESD) provides a normative framework, emphasizing that sustainability should be integrated holistically across teaching, research, and engagement [10,11]. These frameworks, taken together, provide the conceptual basis for our study.

In sum, embedding SDG-related themes within higher education programs—particularly in the context of Industry 5.0—is not merely a curricular enhancement. It is a necessary transformation that responds to pressing global challenges. It offers students the tools to become not only skilled professionals but also responsible agents of change.

2.2. Digital Ethics and Higher Education: Preparing Students for Responsible Technology Use

As technological innovations such as AI, big data, the IoT, and immersive environments (e.g., VR/AR) become increasingly integrated into all aspects of life, the ethical implications of their use grow in significance. Within the framework of Industry 5.0, which emphasizes human-centric and value-driven innovation [45,46], digital ethics emerges as a key domain for educational intervention. Higher education institutions thus face a dual challenge: to equip students with the technical skills required in the digital economy and to foster ethical awareness, critical reflection, and responsible digital citizenship.

Digital ethics encompasses issues such as algorithmic bias, data privacy, surveillance, digital inclusion, environmental impacts of computing, and the social consequences of automation and AI-driven decision-making [47,48]. These topics directly affect public trust and democratic governance, particularly as advanced technologies increasingly shape our behaviors, choices, and relationships. The responsible development and deployment of digital tools require a solid ethical foundation grounded in transparency, accountability, human dignity, and social justice [49,50].

HEIs are increasingly integrating these themes into curricula, often within courses on digital ethics, AI ethics, or technology and society. However, to make a tangible impact, ethical literacy must become a transversal component of study programs across disciplines [51]. For example, engineering students may explore the ethical consequences of automation and machine decision-making [52]; education majors may consider the biases in adaptive learning systems [53]; and public health students may critically assess the ethics of data use in surveillance or predictive modeling [54,55]. Recent evidence confirms that embedding ethics across programs is more effective than offering standalone modules: Barth and Rieckmann [56] demonstrated the success of “Embedded EthiCS” in computer science education, while Sandri et al. [57] showed that online ethics training improves science, technology, engineering, and mathematics (STEM) students’ awareness. These examples suggest that ethics should be mainstreamed institution-wide.

Recent research highlights the importance of developing “digital responsibility” as a future-proof competence [51]. This includes not only understanding the risks associated with digital technologies but also cultivating habits of critical inquiry, ethical reasoning, and anticipatory thinking. The ethical challenges of Industry 5.0—ranging from AI governance to cybersecurity and ecological sustainability—require interdisciplinary solutions [58,59]. Yet several barriers persist: Lendvai and Gosztonyi [47] point out that AI governance frameworks remain underdeveloped; Atenas et al. [51] stress that universities lack strong institutional incentives to promote interdisciplinary teaching of ethics; and ethical literacy differs widely across fields. These critiques indicate that the integration of digital ethics is as much an institutional challenge as a curricular one.

Digital environments also provide tools for teaching ethics more effectively. For instance, simulations and virtual scenarios can immerse students in dilemmas that demand ethical decision-making, allowing them to experience the complexities of real-world challenges in safe but realistic conditions [60,61]. Such approaches foster empathy, reflection, and systems thinking—skills essential for ethical leadership in digital societies. However, successful implementation requires adequate infrastructure, trained faculty, and supportive policies—factors that differ significantly across HEIs, reflecting both external pressures (accreditation, societal expectations) and internal resources (digital infrastructure, partnerships, specialization).

Moreover, the design of digital tools themselves must reflect ethical values. This includes developing inclusive AI that mitigates gender and racial biases, as well as promoting accessibility for people with disabilities [62]. HEIs can play a proactive role not only by educating future designers and developers but also by setting institutional standards for digital responsibility in research and administration [63].

Finally, students must be equipped to critically assess the broader societal implications of technological transformation. For instance, the use of surveillance drones in public health and policing [64], or the repurposing of commercial AI for military use [65], raises questions about power, consent, and the erosion of civil liberties. The university is an ideal site to explore such tensions and to promote normative frameworks that balance innovation with democratic values. From a theoretical standpoint, such integration can be interpreted through Institutional Theory [43], which emphasizes external pressures shaping HEIs’ behavior, and the Resource-Based View [44] which highlights the role of organizational capacities such as infrastructure, specialization, and cooperation with enterprises. Together, these perspectives explain why some institutions succeed in embedding digital ethics systematically while others remain limited to partial adoption.

In conclusion, embedding digital ethics into higher education is essential to ensure that the benefits of Industry 5.0 are distributed fairly and responsibly. By cultivating ethical awareness, HEIs contribute not only to workforce readiness but also to the formation of critically engaged citizens capable of navigating and shaping the digital future.

3. Research Questions and Hypotheses

Building on the theoretical premises of Industry 5.0—particularly its shared emphasis on human-centered innovation, sustainability, and digital ethics—this study seeks to explore how these priorities are reflected in the current educational practices of Polish higher education institutions. Universities play a critical role in developing future-oriented skills and values, and prior research has shown that institutional characteristics such as governance model, specialization, and external cooperation can significantly shape the way new curricula are developed [10,66]. Accordingly, the following research questions were formulated:

• RQ1: To what extent do Polish higher education institutions include elements of sustainable development and digital ethics in their curricula?
• RQ2: Are there statistically significant associations between institutional characteristics and the inclusion of these elements?
• RQ3: Which factors significantly increase the likelihood that an institution incorporates sustainability and ethical considerations into its teaching programs?

On the basis of the literature review, we propose the following hypotheses:

H1: 

Institutions that declare cooperation with business in the area of digital technologies are more likely to include sustainability and digital ethics in their curricula.

This expectation is grounded in Institutional Theory, which highlights the influence of external stakeholders and isomorphic pressures on universities [43]. Collaboration with enterprises often brings both technological resources and external legitimacy, encouraging the incorporation of digital ethics and sustainability topics [66].

H2: 

Institutions that already incorporate content or specializations related to Industry 5.0 are more likely to introduce ethical and sustainability-related topics.

Prior studies indicate that universities offering specialized Industry 5.0 curricula tend to mainstream related themes such as sustainability and ethics [6,8]. This also aligns with the Resource-Based View, where unique curricular content and expertise enhance the institution’s capacity to address emerging societal needs.

H3: 

Public institutions are more inclined to embed sustainability and ethics compared to private institutions.

Evidence suggests that public universities are more strongly influenced by national policy frameworks and accreditation requirements related to ESD [10]. By contrast, private institutions often prioritize market-driven curricula, which may limit the integration of ethical and sustainability themes [51].

H4: 

The institutional profile of a university significantly affects the inclusion of sustainability and digital ethics elements in curricula.

Differences in disciplinary orientation and organizational mission are known to shape curriculum design [67]. Technical universities may emphasize digital competencies but show uneven attention to ethics, while comprehensive universities may be more likely to integrate sustainability across faculties. Therefore, the institutional profile is expected to moderate curricular adoption.

These hypotheses are examined using a sequence of methods, including descriptive statistics, chi-square tests with Cramér’s V, multiple correspondence analysis, and an ordinal logistic regression model. The aim is to identify both the patterns and predictors of curriculum development aligned with the human-centered values of Industry 5.0.

4. Material and Methods

4.1. Data Collection

Data were collected in November 2024 using the Computer-Assisted Web Interviewing (CAWI) method. The target population comprised all 354 higher education institutions in Poland listed in the official registry of the Ministry of Science and Higher Education for the 2023/2024 academic year [68]. Of these, 187 institutions completed the survey, representing a response rate of 52.8%.

The sampling strategy was designed as a census of the population of Polish HEIs rather than a sample. Invitations were sent via official institutional e-mail addresses obtained from publicly available directories, and each institution was asked to designate a qualified respondent (e.g., vice-rector for education, head of a curriculum committee, or digitalization officer). Participation was voluntary and anonymous, and no financial incentives were offered.

The online questionnaire was developed by the research team based on a review of international literature on Industry 5.0, Education for Sustainable Development, and digital ethics (e.g., [6,8,11]), and on instruments such as the SULITEST framework. The final tool contained 18 closed-ended questions in total: nominal and ordinal questions. The survey was distributed in Polish, as this is the official language of higher education in Poland.

Prior to full deployment, the questionnaire underwent pilot testing with 10 academic staff members representing different types of institutions. Respondents provided feedback on clarity, relevance, and wording of items. As a result of the pilot, several modifications were introduced: one question was clarified and redundant wording in two items was shortened. Informed consent was obtained electronically prior to beginning the survey, and participants could withdraw at any stage without providing a reason.

4.2. Data Analysis

All analyses were conducted using R software (version 4.4.3) [69]. Several R packages were employed to implement specific analytical techniques: vcd [70] and rcompanion for Cramér’s V, FactoMineR [71] and factoextra [72] for multiple correspondence analysis, ordinal [73] for ordinal logistic regression modeling, gofcat [74] for testing the proportional odds assumption, and rcompanion [75] for pseudo R² statistics. Visualization of predicted probabilities was carried out using ggeffects [76] and ggplot2 [77].

The analytic procedure was structured in three consecutive stages, each corresponding to a different methodological approach and addressing a distinct research purpose.

4.2.1. Bivariate Associations Using Cramér’s V and Chi-Square Tests

As a preliminary step, chi-square tests of independence were conducted to assess the presence of significant relationships between categorical variables. In parallel, Cramér’s V coefficients were calculated to quantify the strength of association. Cramér’s V ranges from 0 (no association) to 1 (perfect association). Although no universal cutoff exists, values above 0.70 are sometimes interpreted as indicative of strong association and potential multicollinearity among categorical variables [78]. To maintain analytical robustness, variables exceeding this threshold should be evaluated carefully prior to inclusion in multivariate models.

Significance was assessed at conventional levels (p < 0.05, p < 0.01, p < 0.001), and the results are presented in the form of a heatmap, combining statistical significance with visual intensity of association. This combination of chi-square and Cramér’s V was chosen because both the predictors and outcomes in our study are categorical. These measures provide a straightforward and well-established way to screen for statistically significant dependencies and to gauge their strength before proceeding to more complex multivariate modeling [79].

4.2.2. Multiple Correspondence Analysis

Following the initial screening, multiple correspondence analysis (MCA) was applied to explore the underlying structure of categorical data and to detect latent patterns across institutional variables. MCA is a data reduction technique designed to visualize complex associations between multiple categorical variables in a low-dimensional space [80,81]. It is particularly well-suited for datasets in which both predictors and outcomes are categorical and unordered or ordinal in nature.

The output of the MCA was used both as a tool of interpretation and a means to refine hypotheses prior to regression modeling. Categories that appeared close to one another in the MCA plot were interpreted as conceptually and empirically similar, suggesting a higher likelihood of association in predictive analysis.

The rationale for employing MCA was twofold: first, it allows the identification of clusters and proximities among categories that might not be apparent through pairwise tests; second, it provides an exploratory foundation to refine hypotheses and guide model specification in the subsequent regression analysis. Previous educational and organizational studies recommend MCA in contexts where categorical predictors are numerous and potentially interrelated [82].

4.2.3. Ordinal Logistic Regression

The final stage of analysis involved estimating an ordinal logistic regression model. The dependent variable (elements) captured the degree to which higher education institutions integrated sustainability and digital ethics elements into their curricula. This variable had three ordered categories: 1 = not at all, 2 = partially, 3 = fully. The variable capturing the extent of curricular integration of sustainability and digital ethics was based on the self-assessment provided by institutional representatives. As such, it reflects the subjective perceptions of universities rather than externally verified indicators.

All predictors were recoded into binary (dummy) variables, and appropriate reference categories were defined based on institutional typology and prior theoretical considerations (e.g., traditional universities as baseline for the “institution” variable, public institutions for “type”).

The ordinal regression model assumes the proportional odds assumption (also known as the parallel lines assumption), which posits that the relationship between each pair of outcome groups is the same. This was tested using the Brant test, where non-significant p-values (p > 0.05) across predictors indicate that the assumption holds [83].

Ordinal logistic regression was selected because the dependent variable is inherently ordinal, with a natural ordering of categories but without equal spacing. This model appropriately accounts for the ordered structure of the outcome while avoiding the assumptions of linear regression. It also provides interpretable odds ratios, which are widely used in social science and education research to convey effect sizes [84].

Model performance was evaluated using likelihood-based measures: log-likelihood, Akaike Information Criterion (AIC), and three pseudo R² statistics—McFadden, Cox & Snell, and Nagelkerke. Log-likelihood and AIC are primarily meaningful when compared across alternative model specifications with different sets of predictors and are therefore reported here for completeness. Pseudo R² values, in turn, do not measure explained variance in the same way as in linear regression, but rather indicate the improvement in model likelihood relative to a null (intercept-only) model. Importantly, values above 0.1 are generally considered strong in the context of logistic models [85].

The general form of the ordinal logistic regression model is as follows [79,86]:

$$\mathit{log}\left({\displaystyle \frac{P\left(Y\le j\right)}{P\left(Y>j\right)}}\right)={\theta }_j-{\beta }_1{X}_1-{\beta }_2{X}_2-\dots -{\beta }_k{X}_k,$$

(1)

where $Y$ is the ordinal outcome variable, $j$ indicates the cumulative logit for each cutpoint between outcome levels (e.g., 1 vs. 2 & 3, and 1 & 2 vs. 3), ${\theta }_j$ is the threshold (intercept) for the $j$th cumulative logit, and ${\beta }_1,{\beta }_2,\dots ,{\beta }_k$ are regression coefficients for predictors $X_1,X_2,\dots ,X_k$.

Interpretation of coefficients is typically expressed in terms of odds ratios (ORs), where OR > 1 indicates increased odds of higher category membership on the outcome variable. Conversely, OR < 1 indicates decreased odds of being in a higher category of the outcome variable, suggesting a negative association between the predictor and the likelihood of progressing to higher levels of the dependent variable. It should be noted that these interpretations are always made relative to the reference category of each predictor (e.g., “no specialization,” “public institution”), against which the alternative category is compared.

5. Results

5.1. Characteristics of the Study Sample

The study sample consisted of 187 higher education institutions in Poland.

Table 1. Sample characteristics (N = 187).

Variable	Description	Category	%
Institution	Type of higher education institution by profile	University	19.8
		Technical university	33.2
		Economic university	9.1
		Pedagogical university	12.3
		Medical university	10.7
		Art university	5.9
		Vocational university	9.1
Type	Ownership status of institution	Public	52.4
		Private	47.6
Unit	Existence of a unit dedicated to digital technologies	Yes	51.9
		No	48.1
Cooperation	Degree of cooperation with enterprises in digital technologies	Regularly	31.0
		Occasionally	43.9
		No	25.1
Adm_or_sci	Use of AI tools in university administration or scientific research	Yes	72.2
		No	27.8
Content	Inclusion of Industry 5.0-related content in curricula	Yes	65.2
		No	34.8
Specializations	Existence of Industry 5.0-related specializations	Yes	65.2
		No	34.8
Elements	Extent of integration of sustainability and digital ethics into curricula	No	26.2
		Partially	62.6
		Fully	11.2

presents the distribution of the institutions by selected characteristics.

The most frequently represented institutional profiles were technical universities (33.2%), followed by comprehensive universities (19.8%) and pedagogical universities (12.3%). Public institutions accounted for slightly more than half of the sample (52.4%), while 47.6% were private universities.

Just over half of the institutions (51.9%) reported having a dedicated unit focused on research and development in digital technologies. Regarding collaboration with businesses in the implementation of digital technologies, 31.0% of institutions indicated regular cooperation, 43.9% engaged occasionally, and 25.1% reported no such cooperation.

A significant proportion (72.2%) of institutions declared the use of AI technologies in either administrative processes or scientific research. Furthermore, the integration of Industry 5.0–related content into educational programs was noted in 65.2% of institutions, and an equal percentage offered specializations related to this area.

Finally, with respect to sustainable development and digital ethics, 62.6% of institutions reported partial inclusion of such elements in curricula, while only 11.2% reported full integration. Approximately one-quarter of the institutions (26.2%) indicated no inclusion of these elements.

This distribution directly reflects the survey items asking institutions to indicate their type, ownership, and whether they include Industry 5.0 content, specializations, and AI use in their activities.

5.2. Bivariate Associations: Cramér’s V and Chi-Square Analysis

The first stage of the analysis focused on assessing the pairwise relationships between categorical variables using chi-square tests of independence and calculating Cramér’s V coefficients to estimate the strength of association. The results are presented in the heatmap in Figure 1, where statistically significant associations are marked with asterisks and the intensity of color indicates the strength of association.

The strongest correlations were observed between institutional profile and type of university (V = 0.60***), institutional profile and presence of a dedicated digital technologies unit (V = 0.64***), type of university and inclusion of specializations related to Industry 5.0 (V = 0.63***), presence of specializations and inclusion of Industry 5.0 content in curricula (V = 0.58***), and presence of a dedicated digital technologies unit and cooperation with enterprises (V = 0.55***).

These relationships indicate that some variables are conceptually and statistically interrelated. However, most associations did not exceed a Cramér’s V threshold of 0.60–0.70, which is sometimes suggested as a warning level for potential multicollinearity among categorical variables. Therefore, the inclusion of these variables in subsequent multivariate analyses was justified.

Moreover, the dependent variable in the later regression model—elements, representing the degree of integration of sustainability and digital ethics into educational programs—demonstrated moderate and significant associations with variables such as content (V = 0.44***), specializations (V = 0.42***), and cooperation with enterprises (V = 0.28***). These findings further support the analytical relevance of those predictors in explaining variance in the outcome variable.

Given the observed statistical significance and the moderate strength of associations, all variables were retained for subsequent multiple correspondence analysis, which served to explore latent patterns in the data. The results of MCA, in turn, informed the specification of an ordinal logistic regression model, in which elements was treated as the dependent variable with three ordered categories: no inclusion, partial inclusion, and full inclusion of sustainability and ethics in curricula.

5.3. Latent Patterns: Multiple Correspondence Analysis

To explore the relationships among categorical variables, a multiple correspondence analysis was conducted. The resulting two-dimensional solution is visualized in Figure 2. The plot provides a geometric representation of the associations between categories of the variables included in the analysis. The MCA plot therefore provides a visual synthesis of multiple questionnaire items, grouping institutions with similar response patterns across the survey variables.

The first dimension (horizontal axis) appears to differentiate institutions based on the extent to which they incorporate elements of Industry 5.0 and sustainability-related practices into their educational activities. On the left side of the axis, we observe the clustering of categories such as technical, public, elements_fully, specializations_yes, content_yes, and cooperation_regularly, all indicating stronger institutional engagement with digital and sustainable transformation. These profiles are also associated with the presence of dedicated digital units (unit_yes) and the use of AI technologies (adm_or_sci_yes).

Conversely, on the right side of Dimension 1, profiles such as pedagogical, medical, private, elements_no, content_no, and specializations_no suggest lower levels of engagement with the pillars of Industry 5.0, and they may reflect institutions with more traditional or narrowly focused educational profiles.

Dimension 2 (vertical axis) introduces a subtler separation. For example, vocational and art institutions occupy the upper area of the plot, possibly indicating different strategic orientations or specialization patterns. Meanwhile, economic universities, positioned in the lower part of the plot, are more distinct from the rest, suggesting a unique institutional identity or curricular approach.

The spatial proximity of response categories suggests conceptual similarity and empirical association. For example, elements_partially, cooperation_ocassionally, and content_yes are located near each other, reflecting institutions with intermediate engagement in digital and ethical transformation.

Overall, the MCA confirms and visualizes meaningful associations identified earlier in the Cramér’s V analysis. The geometric distances between categories reflect underlying patterns of institutional behavior in the context of digital transformation, sustainability, and ethics in higher education.

The spatial distribution of categories in the MCA plot suggested meaningful associations between institutional characteristics and the inclusion of sustainable and digital transformation elements. These results guided the selection of predictors for the next stage of analysis—ordinal logistic regression.

5.4. Ordinal Logistic Regression Results

To identify the institutional and program-related factors associated with the integration of sustainability and digital ethics into educational programs, an ordinal logistic regression model was estimated. The outcome variable was elements, representing the level of inclusion of sustainability and digital ethics (1 = not at all, 2 = partially, 3 = fully). All categorical predictors were recoded into binary dummy variables, with meaningful reference categories. For example, the reference group for institution was traditional universities; for type, public institutions; and for content-related variables, the absence of specific features.

5.4.1. Model Fit and Assumptions

The model demonstrated good fit, with a log-likelihood of −135.13 and an Akaike Information Criterion of 300.25. While these measures are not interpretable in absolute terms, they are useful for comparing alternative specifications. Pseudo R² values indicated a relatively strong improvement in model fit compared to a null model: McFadden’s R² = 0.188, Cox & Snell = 0.284, and Nagelkerke = 0.342. Given that values above 0.1 are seldom observed in logistic regressions, these results point to a robust model performance.

The proportional odds assumption, a crucial requirement for ordinal logistic regression, was tested using the Brant test. The omnibus result was non-significant (χ² = 10.067, df = 13, p = 0.688), indicating that the assumption was satisfied across most predictors. Although one variable (technical university) approached significance (p = 0.038), its impact was marginal and did not substantially compromise model validity.

5.4.2. Key Predictors of Sustainability and Digital Ethics Integration

The results of ordinal logistic regression are presented in

Table 2. Ordinal Logistic Regression Predicting Sustainability and Digital Ethics Integration in Curricula.

Predictor	Estimate	Std. Error	OR	95% CI (OR)	p-Value
Institution: technical university	−0.079	0.493	0.92	[0.35, 2.44]	0.873
Institution: economic university	−0.141	0.685	0.87	[0.22, 3.33]	0.837
Institution: pedagogical university	−1.167	0.720	0.31	[0.07, 1.26]	0.105
Institution: medical	−0.312	0.641	0.73	[0.21, 2.57]	0.626
Institution: art	0.242	0.800	1.27	[0.27, 6.30]	0.762
Institution: vocational	0.416	0.721	1.52	[0.37, 6.27]	0.564
Type: private	0.875	0.531	2.40	[0.85, 6.86]	0.099 †
Unit: yes	−0.117	0.471	0.89	[0.35, 2.24]	0.804
Cooperation: regularly	1.230	0.586	3.42	[1.10, 11.02]	0.036 *
Cooperation: occasionally	0.303	0.460	1.35	[0.55, 3.34]	0.511
AI use (adm_or_sci)	0.491	0.403	1.63	[0.74, 3.62]	0.223
Content: yes	0.754	0.458	2.12	[0.87, 5.25]	0.100 †
Specializations: yes	1.513	0.560	4.54	[1.53, 13.88]	0.007 **

Note: OR = Odds Ratio. † p < 0.10, * p < 0.05, ** p < 0.01.

. Only a few predictors were statistically significant.

Specializations related to Industry 5.0 emerged as the strongest predictor of full integration of ethical and sustainability-related content (OR = 4.54; 95% CI: [1.53, 13.88], p < 0.01). Institutions offering such specializations were over four times more likely to implement sustainability and digital ethics components at higher levels. Regular cooperation with enterprises also significantly increased the odds of integration (OR = 3.42; 95% CI: [1.10, 11.02], p = 0.036). This finding aligns with the assumption that industry collaboration supports curricula innovation and relevance.

Two additional variables approached statistical significance at p = 0.1. Private institutions showed a non-significant positive association (OR = 2.40; 95% CI: [0.85, 6.86], p = 0.099), suggesting that private sector institutions may be more agile in adopting such content. Inclusion of Industry 5.0-related course content was marginally associated with higher odds of integration (OR = 2.12; 95% CI: [0.87, 5.25], p = 0.100).

Other variables, including types of institutions, administrative units, and digital administration/science applications, were not statistically significant. Notably, pedagogical universities showed a non-significant negative association (OR = 0.31; p = 0.105), implying a potential gap in integration that warrants further study.

5.4.3. Predicted Probabilities Based on Key Predictors

To further interpret the results of the ordinal logistic regression, predicted probabilities for the three levels of the dependent variable (elements: 1—no inclusion, 2—partial inclusion, 3—full inclusion) were calculated and visualized for four key predictors: specializations related to Industry 5.0, regular cooperation with enterprises, inclusion of Industry 5.0 content, and type of institution (public vs. private). These variables were selected due to their statistical significance or theoretical relevance in the previous analysis (Figure 3).

The predicted probability plots are shown in Figure 3 and allow for a clearer understanding of the direction and magnitude of effects:

• Specializations related to Industry 5.0

Institutions that offer dedicated specializations in areas related to Industry 5.0 demonstrate a markedly higher probability of fully including sustainability and digital ethics content in curricula. The probability of full inclusion (category 3) rises from about 5% for institutions without such specializations to nearly 15% for those with relevant specializations. Conversely, the likelihood of offering no such content (category 1) drops from around 50% to below 30%, while the probability of partial inclusion (category 2) increases significantly—from about 45% to over 55%.

• Regular cooperation with enterprises

Institutions engaged in regular cooperation with enterprises (compared to no cooperation) show a decrease in the probability of no inclusion (category 1) from about 20% to less than 10%, and a corresponding increase in the probability of full inclusion (category 3) from approximately 10% to over 20%. The probability of partial inclusion (category 2) remains high and relatively stable (around 70%), suggesting that collaboration with external stakeholders may foster deeper engagement with ethical and sustainability themes, especially among institutions already partially addressing these areas.

• Inclusion of Industry 5.0 content in curricula

Institutions that already include broader Industry 5.0 topics are also more likely to embed sustainability and digital ethics. The predicted probability of full inclusion (category 3) rises from approximately 10% to over 20%, and the likelihood of no inclusion (category 1) declines from 35% to below 20%. Meanwhile, the chance of partial inclusion increases modestly, reinforcing the pattern observed for other innovation-related variables.

• Type of institution (Public vs. Private)

Although this predictor was not statistically significant in the regression model, the visualizations reveal only modest and statistically inconclusive differences. The probability of no inclusion (category 1) decreases from approximately 25% in public institutions to around 15% in private ones. Partial inclusion (category 2) remains dominant across both types, with probabilities exceeding 70% in both cases. The likelihood of full inclusion (category 3) shows a slight increase in private institutions (rising from about 5% to 10%), although the wide confidence intervals suggest that these differences should be interpreted with caution. Overall, the type of institution appears to exert a limited influence on the integration of sustainability and digital ethics content.

Together, these visualizations support the regression findings and underline the role of innovation-related institutional features—such as curricular modernization and external partnerships—in promoting ethical and sustainability awareness in higher education programs.

6. Discussion

6.1. Interpretation of Research Questions and Hypotheses

The empirical findings obtained through the combination of descriptive statistics, chi-square testing, multiple correspondence analysis, and ordinal logistic regression provide a multifaceted answer to the research questions and hypotheses formulated in this study.

RQ1 concerned the extent to which sustainability and ethics are embedded in Polish HEI curricula. The descriptive results indicate that, although the majority of institutions report partial inclusion of sustainability and digital ethics (62.6%), only a smaller portion (11.2%) declare full integration of these elements. A notable share of institutions (26.2%) do not include these components at all. This uneven distribution confirms that while Industry 5.0 discourses are gaining visibility in higher education, their curricular operationalization remains fragmented, echoing previous studies that caution against the risks of rhetorical rather than structural embedding of sustainability [3].

RQ2 examined associations between institutional characteristics and curricular inclusion. The chi-square tests combined with Cramér’s V coefficients revealed several statistically significant but low-to-moderate associations between the inclusion of sustainability/ethics content and variables such as university profile, cooperation with business, and curricular orientation toward Industry 5.0. These results reinforce the expectation, derived from Institutional Theory, that organizational environments and external partnerships shape curriculum reform [43].

RQ3 aimed to identify predictors of inclusion. The ordinal logistic regression model confirmed H1 and H2 but not H3 or H4. Institutions cooperating with business in the field of digital technologies showed significantly higher odds (OR = 3.42, p < 0.05) of integrating sustainability and ethics into their curricula, confirming H1. This demonstrates that cooperation with enterprises does not merely enhance technological competencies, but also creates opportunities for embedding human-centric values, aligning with research stressing the role of ecosystemic collaboration in advancing SDGs [10,66].

H2 was also confirmed: institutions offering content and specializations related to Industry 5.0 were significantly more likely to implement the target elements (OR = 2.12 and OR = 4.54, respectively; p < 0.1 and p < 0.01). This finding underscores the importance of specialization-based institutional capacity as highlighted by the Resource-Based View [44].

H3 was not supported: private institutions demonstrated higher odds of inclusion (OR = 2.40), but the effect did not reach conventional levels of statistical significance (p = 0.099). H4 was also not supported: institutional profile (e.g., technical, pedagogical) did not significantly affect integration. This suggests that sustainability and ethics can be embedded across institutional types, provided adequate strategic commitment and resources are available.

6.2. Practical Implications, Barriers, and Recommendations for Higher Education Institutions

The results of the study highlight several implications for higher education institutions, particularly in the context of curriculum development aligned with the human-centric values of Industry 5.0, which integrates both sustainability and digital ethics as foundational principles.

6.2.1. Institutional Readiness and Strategic Commitment

The strong predictive value of variables such as business cooperation and the presence of Industry 5.0-oriented specializations suggests that institutions with proactive innovation strategies are more likely to internalize the values of sustainable development and ethical awareness. This indicates that embedding these components into teaching is not merely a curricular decision, but often reflects wider institutional orientation and strategic partnerships.

HEIs that establish links with the digital economy—particularly through joint research, technology transfer, or co-designed learning modules—appear to perceive greater value in aligning their educational offerings with the principles of responsible innovation. This finding is in line with prior research emphasizing the importance of ecosystemic collaboration in advancing SDGs [87,88].

6.2.2. Barriers to Implementation

Nevertheless, the analysis also points to the limited and partial integration of sustainability and ethics in many institutions, particularly in those lacking specialization or cooperation capacity. This reflects barriers such as institutional inertia, insufficient faculty training, and resource constraints. These obstacles mirror international findings indicating that even in countries with supportive policy frameworks, higher education institutions often struggle to operationalize SDGs in curriculum due to path dependency and fragmentation between disciplines [6,10]. Prior studies highlight that barriers include the absence of incentives for interdisciplinary teaching [51], limited digital infrastructure [89], and competing institutional priorities such as marketization of higher education [11]. Together, these constraints demonstrate that curricular innovation requires not only pedagogical reform but also systemic support at policy and governance levels.

6.2.3. Recommendations and Future Directions

In light of the challenges and findings of this study, several recommendations can be articulated for different levels of the higher education ecosystem. At the policy level, national and regional authorities should consider introducing incentives or even requirements for the inclusion of sustainability and ethics learning outcomes within accreditation and program standards. Previous research demonstrates that supportive policy frameworks play a critical role in accelerating the mainstreaming of Education for Sustainable Development across institutions [67,90]. In parallel, curriculum innovation at the institutional level is needed, with universities encouraged to develop interdisciplinary modules that combine digital competencies with social responsibility, such as “Artificial Intelligence and Society” or “Sustainable Technology Futures”. Such transversal integration has been shown to be more effective than isolated elective courses, particularly when supported by faculty development and cross-disciplinary collaboration [56,57].

Another important direction involves strengthening cross-sector partnerships. Enhanced cooperation with enterprises, public institutions, and civil society organizations can provide students with exposure to real-world sustainability challenges and ethical dilemmas, thereby reinforcing the relevance of curricula and creating conditions for responsible innovation [91,92]. At the same time, robust mechanisms for monitoring and evaluation should be implemented to ensure that SDG-related education is not only present but also effective. International tools such as the SULITEST [93] and UNESCO’s Education for Sustainable Development indicators [94] offer valuable frameworks for tracking progress and benchmarking across institutions [10].

Ultimately, embedding the logic of Industry 5.0 into higher education goes beyond technological instruction. It requires cultivating critical thinking, civic values, and ethical reflexivity, preparing students not only to adapt to digital transformation but also to actively shape it in ways that serve both people and the planet. This holistic vision is consistent with global calls for universities to act as agents of sustainability and social responsibility [6,95].

6.3. International Perspectives and Literature Comparisons

The findings of this study align with a growing body of international research highlighting the role of higher education institutions in embedding sustainability and digital ethics within curricular structures, particularly in the context of Industry 5.0 frameworks. We now situate our results in a broader global dialogue.

Recent studies underscore that universities worldwide are adapting curricula to cultivate human-centered competencies aligned with Industry 5.0. For instance, De Villiers [96] observed a global shift toward integrating technological literacy, ethics, and interdisciplinary learning within Industry 5.0-oriented programs. Similarly, Carayannis and Morawska [97] frame “University 5.0” as a vehicle for local-global social-innovation ecosystems combining technical, ethical, and sustainability competencies. In Poland, our findings reveal that institutions offering Industry 5.0-specific content and specializations are more likely to include ethical and sustainability components in curricula—consistent with this international trend. Moreover, parallels can be drawn with Shahidi Hamedani et al. [98], who identified that universities preparing students for Industry 5.0 emphasize digital skills and social responsibility.

Parallel literature on Industry 5.0 emphasizes the centrality of human values in technical education. Hashim et al. [99] argued that universities must develop new skills to balance economic efficiency with societal purpose, often through IoT or “human-centric innovation” programs. Our evidence—showing the importance of specializations and business cooperation—confirms that Polish HEIs engaging with these paradigms tend to exhibit stronger curricular alignment.

The concept of embedding ethics across technical disciplines finds support in work such as Grosz et al. [100], who implemented philosopher-led “Embedded EthiCS” modules within computer science curricula. This resonates with our recommendation that digital ethics be not treated as an “add-on”, but integrated into core programs—a recommendation reinforced by observed international best practices.

Global reviews, such as Peters et al. [101], highlight nascent but growing integration of sustainability in computing education. In Europe, education for climate action is gaining ground, as evidenced by curriculum reforms driven by Royal Institute of British Architects in the U.K. [102]. Our study’s identification of uneven sustainability content inclusion aligns with this global pattern of gradual progression, but notable gaps remain.

Digital ethics integration in Latin American universities was examined by Fernández-Miranda et al. [103], showing growing ethical awareness. Meanwhile, Nosratabadi et al. [104] found that countries in Northern Europe (e.g., Finland, the Netherlands, Denmark) benefit from high social sustainability in digital transformations. These findings resonate with our study’s observation that institutions in enabling environments (i.e., those with support for specializations and cooperation) display greater curricular integration.

Our study contributes to an international consensus: HEIs that invest in structural capacity—via specialized programs, external partnerships, and holistic curricular frameworks—are more effective in embedding sustainability and digital ethics, following a global trajectory influenced by Industry 5.0. At the same time, the prevalence of partial or absent integration highlights the importance of policy initiatives, academic community action, and curricular reform to bridge the gap between educational aspirations and institutional practice.

6.4. Limitations and Directions for Future Research

While this study provides new insights into how Polish higher education institutions address the principles of sustainable development and digital ethics in the context of Industry 5.0, several limitations should be acknowledged.

First, although the sample covered a relatively wide range of institutions (N = 187), the study relied on voluntary self-reporting, which may introduce selection bias. Institutions with greater engagement in sustainability or innovation-oriented reforms might have been more likely to participate. As a result, the findings may overestimate the true level of curricular integration of sustainability and ethics.

Second, the results apply specifically to the Polish higher education context. While this national lens is valuable for policy relevance, it limits the generalizability of findings to other countries with different policy frameworks, educational structures, or cultural approaches to innovation and sustainability.

The study used binary and ordinal-coded variables, based on self-assessed institutional declarations. While this structure allowed for consistent comparative analysis, the measures may not fully reflect the depth or quality of the curriculum content. For example, an institution marking “full integration” of sustainability may interpret this differently from another institution using the same label. Furthermore, the scope of the questionnaire limited exploration of student perspectives or teaching staff engagement, which could have added a richer, qualitative dimension.

Although multiple statistical techniques were used—Cramér’s V, chi-square tests, MCA, and ordinal logistic regression—the study is cross-sectional in nature, preventing causal inference. In particular, regression results capture associations but not the directionality or long-term evolution of curricular decisions. In addition, the ordinal logistic regression model assumes proportional odds, which although tested and mostly upheld in this study, may not fully capture complex nonlinear dynamics in institutional decision-making.

Finally, the analysis focused on sustainability and digital ethics as interpreted through the lens of Industry 5.0. Other relevant dimensions—such as student competencies, diversity and inclusion, or civic engagement—were beyond the scope of this study but may represent important future extensions, especially when aligned with broader SDG educational goals.

Future research could benefit from: mixed-methods approaches (e.g., combining surveys with interviews or case studies), longitudinal designs tracking changes over time, comparative studies across countries or within regions, deeper exploration of implementation barriers within institutions, student and faculty perspectives on the perceived relevance of sustainability and ethics content.

Despite these limitations, the present study offers a valuable starting point for ongoing monitoring and benchmarking of higher education’s role in advancing sustainable and ethically responsible innovation in the age of Industry 5.0.

7. Conclusions

This study provides a novel empirical perspective on the integration of sustainability and digital ethics elements into higher education curricula within the broader framework of Industry 5.0. Drawing on data from 187 Polish universities, the findings underscore the significant role that institutional characteristics—such as specialization offerings, cooperation with enterprises, and curricular modernization—play in fostering future-oriented competencies.

The results show that universities actively engaging with Industry 5.0-related innovations are more likely to incorporate sustainability and ethical considerations into their educational programs. Importantly, the study highlights that such engagement is not yet universal, and variability exists across types of institutions and levels of external collaboration.

These findings suggest that universities are not only knowledge providers but also incubators of future competencies, capable of shaping societal transformation through the integration of ethical and sustainable values into curricula. As such, higher education institutions should be viewed as strategic actors in building inclusive, resilient, and responsible digital societies.

Future research may explore longitudinal changes in curriculum content and extend the analysis to other national contexts. Meanwhile, policymakers and academic leaders are encouraged to support initiatives that strengthen universities’ capacity to serve as both catalysts and incubators of the competencies required in Industry 5.0.