Beyond the Project: Towards Sustainable Gender and EDI Change in Mediterranean Research Institutions

Abstract

This article examines gender inequalities in scientific research in the Mediterranean, with a particular focus on STEM disciplines. It draws on qualitative data from the STEP (STEM and Equality, Diversity and Inclusion for Research Enhancement in Portugal) project, a European Commission-funded initiative aimed at embedding equality, diversity, and inclusion (EDI) principles across partner institutions in Portugal, Italy, France, and Spain. Using semi-structured interviews with five scientific leaders and an inductive thematic analysis, the study explores early-stage mechanisms in the institutionalisation of EDI policies and women’s empowerment trajectories from an intersectional perspective. The analysis identifies emergent patterns suggesting: (i) a gradual strengthening of EDI mainstreaming in contexts with initially limited awareness; (ii) the role of transnational collaboration in enhancing visibility, mentoring, and peer learning; and (iii) the potential of time-bounded initiatives to catalyse participant-observed shifts and institutional routines in formation. Rather than measuring longitudinal impact, the article traces how legitimation, routinisation, and network diffusion may enable EDI principles to extend beyond project lifespans and become embedded in governance structures. These mechanism-focused insights offer a transferable framework for future European cooperation initiatives and contribute to ongoing debates on sustainable gender and EDI policy implementation in Mediterranean research contexts.

1. Introduction

The gender gap in STEM research—an acronym for Science, Technology, Engineering, and Mathematics—remains a persistent challenge across Europe [1,2], particularly in Mediterranean contexts where structural (e.g., lack of transparent recruitment procedures), cultural (e.g., implicit bias in evaluation committees), and institutional barriers (e.g., absence of formal EDI policies) continue to shape academic pathways [3,4]. The Portuguese context is particularly relevant to this study. Although Portugal has made progress in aligning with European gender equality frameworks, important asymmetries remain in the translation of formal commitments into everyday academic and research practices, especially in STEM environments and in relation to leadership, recruitment, and career progression. In this sense, Portugal offers a useful vantage point for examining how EU-funded initiatives can operate not simply as policy add-ons, but as mechanisms for institutional learning, legitimation, and organisational change within a wider Mediterranean landscape. Empirical evidence supports this positioning [3,4,5]. Portugal is among the EU countries closest to gender parity in overall research participation, with women comprising almost 50% of active researchers [1,6]. However, significant vertical segregation persists, particularly in senior academic ranks and decision-making roles, where women remain markedly underrepresented. These asymmetries are especially pronounced in STEM fields [1], where progress in recruitment has not been matched by equivalent advances in leadership and career progression, reflecting structural and organisational barriers rather than deficits in participation or productivity. Despite the introduction of national strategies and legal frameworks addressing gender equality in higher education and research [7], studies consistently highlight a gap between policy adoption and everyday institutional practices, making Portugal a particularly instructive case for analysing how EU-funded initiatives can serve as catalysts for institutional learning, legitimacy, and sustainable organisational change.

The EU-funded STEP—STEM and Equality, Diversity and Inclusion for Research Enhancement in Portugal (2023–2025) project was developed in response to these challenges. Coordinated by the Instituto Politécnico de Bragança (Portugal), STEP involves four partner institutions across Portugal, Italy, France, and Spain. The project integrated EDI principles into research and teaching activities, with a particular focus on STEM fields, through mentoring programmes, staff exchanges, EDI committees, training modules, and communication initiatives. In fact, the STEP project aimed to strengthen the research excellence and international reputation of the Instituto Politécnico de Bragança (IPB) by integrating EDI within the STEM fields through a strategy of cross-fertilisation. To achieve this, the project pursued three integrated objectives: first, to enhance IPB’s scientific and technological capacity by pioneering innovative research programmes that bridge STEM and EDI, positioning the institution as a benchmark for inclusive excellence; second, to expand internationalisation and networking through strategic twinning mechanisms with leading European partners, fostering a globally oriented research community and promoting open science practices; and finally, to reinforce research management and administrative competencies by strengthening the institution’s ability to attract external funding and manage international cooperation projects, thereby ensuring long-term sustainability and competitiveness within the European Research Area.

Previous experiences of international partnerships in Portugal, such as the MIT–Portugal Program, have shown how time-limited initiatives can trigger processes of internationalization and institutional innovation [8].

STEP was designed with an innovative approach, particularly in making EDI mainstreaming its cornerstone. While addressing gender issues, it embraces the entire EDI theme, adopting an approach now referred to as gender+ [9]. At the same time, EDI issues are presented as a scientific discipline that achieves its best results through full interdisciplinary dialogue with STEM disciplines, overcoming obstacles often encountered in SSH/STEM dialogue [10,11].

1.1. Theoretical Framework

This study draws on two complementary theoretical perspectives to interpret the processes of institutional and cultural change observed in the STEP project.

Neo-institutionalism [12,13] emphasises how organisational practices are shaped not only by efficiency or rational choice, but also by legitimacy, norms, and environmental pressures. Concepts such as isomorphism (the tendency of organisations to adopt similar practices under uncertainty) and decoupling (the gap between formal policies and actual routines) provide analytical tools for understanding how EDI practices diffuse across institutions and whether they become substantively embedded or remain ceremonial [12]. These lenses are particularly relevant for Mediterranean research contexts, where EDI maturity varies and where external legitimation—for instance, through EU funding—can accelerate institutional alignment [12,13].

The analysis of institutional change from a gender perspective enriches this perspective by showing that organizations are never gender-neutral and that reforms can produce both progress and implicit forms of resistance.

Actor-Network-Theory (ANT) [14] shifts attention from organisations as stable entities to the dynamic assemblies of human and non-human actors (researchers, tools, documents, platforms, funding schemes) that produce change through distributed agency. Concepts such as translation (how local adaptations of global ideas occur) and mediator (an actor that transforms the network it enters) help to capture how STEP connected heterogeneous elements—STEM researchers, EDI specialists, Gender Balance Reports, mentoring programmes, European policy instruments [15,16,17] —into a functioning transnational network. ANT is particularly suited to explain how even small-scale, time-limited projects can generate lasting effects by reconfiguring relationships and stabilising new practices through inscriptions (e.g., guidelines, monitoring tools, shared protocols).

By combining neo-institutionalism and ANT, this study addresses both the macro-level dynamics of legitimacy and institutional alignment and the micro- and meso-level processes of network-building and distributed agency [15]. Together, they provide a coherent analytical framework for investigating how time-limited EU projects can trigger sustainable EDI change in Mediterranean research institutions.

1.2. Research Questions and Study Aim

This article investigates the mechanisms through which time-limited EU-funded projects foster emergent patterns of institutional and cultural adaptation in EDI policies within Mediterranean research institutions. Specifically, it addresses the following research questions:

• How do time-limited EU-funded projects (such as STEP) trigger institutional and cultural change in EDI policies within Mediterranean research institutions?
• What mechanisms—legitimation, routinisation, and network diffusion—can be identified as drivers of sustainable EDI mainstreaming?
• What tensions, contradictions, or unintended effects emerge in the process of institutionalising EDI practices across diverse national and organisational contexts?

Drawing on preliminary qualitative data from the STEP project, the study analyses how legitimation, routinisation, and network diffusion operate across four partner organisations in Portugal, Italy, France, and Spain. Rather than assessing the overall impact or success of the STEP project, the analysis focuses on the processes through which EDI practices become embedded—or fail to do so—in institutional routines, governance structures, and collective awareness. Based on semi-structured interviews with five scientific leaders involved in STEP, the study identifies recurring dynamics of institutional change and offers an empirically grounded understanding of how even small-scale, time-bound initiatives catalyse participant-observed shifts and institutional routines in formation in contexts where EDI maturity is still developing.

2. Methodology

2.1. Research Design

This study adopts a qualitative, inductive research design [18,19], aimed at exploring how time-limited EU-funded projects trigger institutional and cultural change in EDI policies. Rather than testing pre-defined hypotheses, the research seeks to identify recurrent mechanisms—legitimation, routinisation, and network diffusion—through the analysis of participants’ accounts. This design is particularly suited to investigating under-theorised processes in context-specific settings, such as Mediterranean research institutions [20,21].

2.2. Sampling and Participants

A purposive sampling strategy [22] was adopted to select individuals with direct scientific leadership roles within the STEP project. Inclusion criteria were: (i) formal responsibility for coordinating or implementing EDI-related activities; (ii) direct involvement in institutional decision-making or operational rollout of EDI measures; and (iii) availability for an in-depth interview and a subsequent written follow-up. The final sample consisted of five participants from four research and higher education institutions in Portugal, Italy, France, and Spain.

While the sample size (N = 5) is small, it is consistent with qualitative research designs that prioritise information-rich cases and analytic depth rather than statistical representativeness. Previous methodological studies have shown that thematic saturation can be achieved with a limited number of interviews, particularly in relatively homogeneous samples [23]. Focusing on scientific leaders as a clearly defined and homogeneous role category enabled the identification of recurrent institutional mechanisms across cases, rather than individual-level variability. To minimise the risk of re-identification, participant characteristics are reported only in aggregated form; information on career stage, professional experience, or gender was not systematically collected, in line with a data-minimisation and confidentiality-by-design approach.

2.3. Data Collection

Between June and September 2025, semi-structured interviews [19] were conducted by personnel specifically trained in qualitative interviewing. A ten-point interview guide was developed based on the STEP project’s objectives and European policy frameworks [2]. The guide was shared in advance with participants, together with an information document, allowing time for reflection and more substantive responses.

Interviews lasted between ninety minutes and two hours, conducted either in person or online. For ethical and logistical reasons—specifically, to ensure participant comfort and to facilitate open discussion in a small, recognisable community—no audio recordings were made. This was a deliberate ethical choice, not a logistical constraint. To ensure data reliability despite the absence of recordings, a two-step triangulation procedure was adopted: detailed interview notes were taken by trained interviewers, and each participant subsequently provided written answers to the same questions, allowing cross-verification and reducing memory bias or content loss [24].

2.4. Ethical Approval

The study received prior ethical approval from the Ethics Manager of the STEP Project, who was formally appointed by the Project Steering Committee during the kick-off meeting. Informed consent was obtained from all participants. The data collected did not include sensitive information and was subsequently anonymised through the assignment of pseudonyms.

2.5. Thematic Framework and Analysis

The semi-structured interview guide was organised around ten thematic areas derived from the STEP project’s objectives and the relevant literature on gender equality and EDI: institutional change, empowerment, intersectionality, international collaboration, cultural change, sustainability, critical issues, impact assessment, agency, and external dissemination. These ten areas served solely as a conversational framework to ensure coverage of key topics and comparability across interviews. They were not used as a pre-defined coding scheme in the subsequent analysis. Their function was limited to guiding the interview process, not to imposing analytical categories on the data. This approach is consistent with exploratory, mechanism-focused qualitative research [22].

Data analysis was conducted using inductive reflexive thematic analysis following the six-phase procedure outlined by Braun and Clarke [18,25]: familiarisation, initial coding, searching for themes, reviewing themes, defining themes, and writing. Crucially, the ten thematic areas from the interview guide did not serve as a starting point for coding. Instead, codes and patterns emerged organically from the data through line-by-line reading of interview notes and written follow-ups. The authors maintained sustained attention to intersectional dynamics throughout the coding process.

Although the interview guide contained ten thematic areas, the inductive analysis did not treat them as fixed categories. The resulting themes were not required to map onto these areas; rather, the analysis identified patterns across them. After the inductive coding was complete, the authors examined whether the resulting themes could be meaningfully clustered under broader analytical dimensions. This post-analytic interpretative step yielded four higher-order dimensions—institutionalisation, empowerment and inclusion, transnational networking, and sustainability—which are used in the Results section to organise the presentation of findings. These dimensions reflect an interpretative synthesis of the inductively generated themes, not a reduction of the original ten areas, and were not imposed a priori on the data. The coding process was discussed and jointly validated by all authors, strengthening the intersubjective reliability of the categories developed. Although the sample size is small, the qualitative depth of the interviews, the triangulation of sources (notes + written follow-ups), and the procedural rigour of the analysis lend methodological robustness and interpretive solidity to the findings [25].

3. Results

3.1. Main Thematic Dimensions Emerging from the Interviews

The analysis of interview notes and written follow-ups yielded four higher-order analytical dimensions: (i) institutionalisation of EDI practices, (ii) empowerment and intersectional inclusion, (iii) transnational networking, and (iv) sustainability and post-project continuity. The subsections below present participant accounts organised around these dimensions, highlighting emergent patterns and institutional routines in formation. Table 1 synthesises the key findings and their theoretical anchoring.

Table 1. Synthesis of the main results.

Area	Main Results	Examples/Evidence	Theoretical Linkage
Institutional change	Systematic integration of EDI into research, teaching, management	EDI+ working groups, awards, communication routines	Mimetic isomorphism [13]—adoption of legitimate practices under uncertainty
Empowerment of women	Greater visibility, mentoring, career support	More women co-authors, increased female enrolment, fee reduction	Distributed agency [14]—change from alignment of heterogeneous actors
Intersectional approach	Inclusion of ethnicity, disability, SES	Differentiated pathways, personalised mentoring	Spillover effect—inclusive norms generate isomorphic pressures [13]
International collaboration	Exchange of best practices, shared guidelines	Accessible documents, staff exchange, interdisciplinary projects	Translation [26]—local adaptation while maintaining network coherence
Cultural change	Normalisation of inclusive language and practices	Greater EDI awareness, integration into GEPs and curricula	Routinisation [12]—from ceremonial to taken-for-granted practice
Sustainability	Consolidation beyond project duration	Regular training, gender balance reports, active interdisciplinary groups	Institutionalisation through routinisation [12]—reduced dependence on champions
Critical issues	Limits of scale, difficulty in profound reforms	Outdated/generic indicators, focus on already-sensitised groups	Decoupling [12]—symbolic adoption without behavioural integration
External impact	Influence beyond partners	Public presentations, collaborations with NGOs and external networks	Network effects/Mediator [14]—reconfiguring relationships beyond original consortium

3.1.1. Institutional Change and Policy Adoption

Participants described a gradual shift in how EDI was perceived within their institutions, moving from a peripheral concern to a more recognised element in project design, teaching, and academic events. As one interviewee noted: “At first, EDI seemed like a peripheral issue, but now it is a recognised criterion in all the projects we present.” Respondents reported the introduction of EDI committees at the research centre level, dedicated awards, blended training modules, mentoring programmes, and more structured communication routines, including newsletters, institutional websites, and social media dissemination. While cultural resistance and limited EDI literacy were frequently mentioned, participants observed that alliances with institutional leadership and the external legitimacy conferred by EU funding helped navigate these barriers. Resistance from senior staff and limited institutional resources nevertheless remained significant obstacles.

These accounts illustrate how mimetic isomorphism can operate in contexts with initially low EDI maturity, with transnational networks providing models perceived as legitimate and actionable [13].

3.1.2. Empowerment of Women

Interviewees highlighted increased opportunities for visibility through conferences, seminars, and international mentoring activities, which they associated with greater female participation in STEM events, co-authorship, and collaborative research teams. Empowerment was primarily observed at the network level, where early-career female researchers described gaining confidence and recognition through structured peer support: “For us young female researchers, mentoring has been fundamental, giving us confidence and visibility.” Participants also referred to improved participation in research groups and greater recognition of women’s contributions within collaborative activities. While intersectional factors such as career stage and socioeconomic context were acknowledged, respondents noted that systematic institutional responses to these dimensions remain in early stages of development.

The dynamics described by participants reflect forms of distributed agency, in which empowerment emerges through the alignment of mentoring structures, peer networks, institutional support, and transnational collaboration [14].

3.1.3. International and Collaborative Dimension

Comparative discussions across partner countries revealed varying levels of EDI policy maturity. Participants noted that transnational dialogue fostered mutual learning, with smaller institutions describing how European networks provided external legitimacy and reduced feelings of isolation: “Comparing ourselves with partners from other countries allowed us to understand that we were not alone in facing certain resistance.” Another respondent added: “European networks gave us legitimacy: being small does not mean being irrelevant.” Interviewees also emphasised that collaboration between institutions with stronger expertise in equal opportunities and those with STEM-oriented competences generated methodological innovations, staff exchanges, and common operational tools.

These dynamics suggest how transnational connections functioned as mediators, aligning diverse institutional contexts around shared EDI practices and enabling the circulation and local adaptation of guidelines, tools, and procedures across sites [14,26].

3.1.4. Cultural and Structural Change

Participants reported a growing openness to inclusive language and practices, describing them as gradually becoming more routine in meetings, committees, and project activities: “Inclusive language has now become normal in our meetings, and this was by no means a given.” Respondents also referred to increased collective awareness around EDI issues and to modest but meaningful structural adjustments, including gender monitoring systems, equality plans, biennial surveys, interdisciplinary working groups, and newly formalised institutional roles dedicated to EDI.

These developments were described as routines in formation rather than fully consolidated reforms. The gradual normalisation of inclusive language and accessible design practices reflects the transition from ceremonial adoption toward more embedded organisational routines, in line with neo-institutional perspectives on routinisation and institutional embedding [12].

3.1.5. Critical Reflections and Limitations

Interviewees consistently noted that project duration and funding constraints limited the scope of deeper institutional reforms. As one participant observed: “Two or three years is too short a time to really change an institution; continuity is needed.” Respondents highlighted that resistance to change persisted particularly in more hierarchical institutional settings and that participation by male researchers remained comparatively limited in several activities. Participants suggested that future initiatives would benefit from longer project cycles, stronger incentives for leadership engagement, micro-grants supporting EDI pilot actions, and the integration of EDI-related indicators into research evaluation systems.

These tensions reflect broader contradictions already identified in the literature between gender equality strategies and organisational inertia, particularly in academic systems characterised by hierarchical and competitive structures [15].

3.1.6. Impact Assessment and Measurement

When discussing evaluation, participants emphasised the importance of developing more precise and differentiated indicators capable of monitoring participation across gender and intersectional categories. Several interviewees referred to participation in events, women’s involvement in project proposals and publications, and the introduction of periodic surveys as relevant monitoring tools. Respondents also highlighted the need to disaggregate indicators across students, academic staff, and technical-administrative personnel.

Unexpectedly, participants reported that EDI-related activities received explicit recognition during external institutional excellence evaluations conducted by international experts. One interviewee noted: “For the first time, our centre has also been positively evaluated for its EDI policies: this was an important recognition.” Such accounts suggest that external validation can reinforce internal legitimacy and contribute to strengthening institutional credibility, even in the absence of formal longitudinal impact metrics.

3.1.7. Individual and Collective Agency

Participants described the development of both informal and formal networks extending beyond the original consortium, now involving more than 200 researchers from over 20 countries through summer schools, conferences, and Blended Intensive Programmes (BIPs). These connections were reported to foster intergenerational dialogue and strengthen collaborative relationships between students, early-career researchers, and senior academics. As one participant stated: “With the BIPs and summer schools, we have built relationships that go beyond the project.” Another respondent added: “The international network created by STEP has made us feel part of something bigger”.

These narratives point to forms of distributed agency emerging through sustained peer interaction, transnational collaboration, and the circulation of EDI practices across institutional and disciplinary boundaries [14].

3.1.8. Sustainability and Mainstreaming

Respondents described several practices introduced during STEP as progressively becoming embedded in daily institutional routines, including regular training activities, mentoring schemes, systematic communication practices, inclusive language protocols, and periodic surveys. Interviewees particularly emphasised the stabilisation of EDI committees, interdisciplinary groups, and research awards linked to EDI+ initiatives. As one participant stated: “EDI committees are now part of the routine, there is no going back.” Another added: “If European funds run out, at least we have stable procedures to continue”.

Participants nevertheless acknowledged that full institutionalisation remains contingent on stable resourcing and ongoing governance support. The formalisation of procedures and roles was perceived as reducing dependence on individual champions, although resistance and uneven engagement persisted across institutional contexts. Several respondents also referred to the development of new project proposals intended to ensure continuity beyond the original funding cycle.

These developments are consistent with neo-institutional interpretations of institutionalisation processes, whereby practices initially introduced as innovations become progressively stabilised through formal structures and organisational routines [12].

3.1.9. Intersectionality and Inclusivity

Interviewees reported that the project encouraged a more explicit consideration of intersectional factors such as disability, socioeconomic status, and ethnicity. As one respondent explained: “We have learned that it is not enough to look at gender alone: economic status, disability, and other factors also make a difference.” Participants described experimenting with differentiated pathways for students and staff, personalised mentoring activities, and practices aimed at avoiding one-size-fits-all approaches.

Several interviewees also observed that environments characterised by stronger gender balance appeared more receptive to addressing other forms of inequality: “Where there is gender balance, discrimination on other fronts also decreases.” These accounts suggest possible spillover effects from EDI mainstreaming processes, although systematic data collection on intersectional inequalities remains limited [16,17].

3.1.10. External Impact and Dissemination

Beyond partner institutions, participants described increasing engagement with national EDI networks, NGOs, public bodies, and policy forums. Dissemination activities were reported to amplify the project’s visibility and to strengthen external collaborations. As one participant noted: “We now collaborate with external NGOs and networks: the project has gone beyond academic boundaries.” Another respondent observed: “Presenting STEP at national conferences has given us visibility that we would not have had on our own.”

Interviewees highlighted that conferences, digital dissemination, shared guidelines, and collaborative initiatives with external organisations contributed to extending the project’s influence beyond the formal consortium. These patterns indicate how even relatively small-scale EU-funded initiatives may generate network effects and forms of institutional visibility that persist beyond the immediate project timeframe [14].

4. Discussion

Rather than measuring longitudinal impact, this study traces the early-stage mechanisms through which time-limited EU-funded initiatives can seed institutional and cultural adaptation in EDI policies. By combining neo-institutional and actor-network perspectives, the analysis identifies three interrelated processes—legitimation, routinisation, and network diffusion—that help explain how project-based activities may transition from temporary interventions to embedded institutional routines. These mechanisms do not imply linear or guaranteed change; rather, they illuminate the conditions under which EDI practices gain traction in Mediterranean research contexts where policy maturity remains uneven.

The empirical accounts suggest that EU funding and transnational collaboration functioned as external sources of legitimacy, accelerating the adoption of EDI structures such as dedicated committees, monitoring protocols, and inclusive communication practices. From a neo-institutional standpoint, this aligns with mimetic isomorphism: organisations operating under uncertainty often draw on externally validated models to enhance their institutional credibility [12,13]. However, participants also highlighted a temporal gap between formal adoption and substantive integration. In several cases, EDI visibility increased more rapidly than its embedding into core academic routines (e.g., recruitment, curriculum design, research evaluation), pointing to a form of partial decoupling [12]. This tension underscores that ceremonial alignment does not automatically translate into behavioural change, and that sustained embedding requires continuous leadership endorsement, resource allocation, and alignment with national policy frameworks.

The analysis further indicates that project-based initiatives tend to consolidate engagement among already receptive actors, highlighting the challenge of extending EDI mainstreaming beyond early adopters. Routinisation emerged unevenly across contexts, shaped by prior institutional maturity, hierarchical structures, and the availability of dedicated staff. These patterns suggest that sustainability is less a project outcome than a contingent process, dependent on whether EDI practices are progressively detached from individual champions and integrated into standard governance procedures. To track this transition, future research could adopt a lightweight follow-up design measuring (i) continuity of activities post-funding, (ii) integration into core institutional processes, and (iii) allocation of stable budget lines. Such indicators would help distinguish symbolic adoption from durable organisational adaptation.

Actor-Network Theory sharpens this interpretation by shifting the focus from institutions as static entities to the dynamic assemblies of human and non-human actors that enable change. Within this framework, STEP operated as a mediator rather than a simple intermediary: it translated EU-level EDI priorities into locally actionable formats by aligning researchers, gender balance reports, training modules, accessibility guidelines, and transnational partnerships. These artefacts functioned as inscriptions that stabilised commitments and rendered EDI practices portable across sites [14,26]. The project’s interdisciplinary orientation (STEM + SSH) further facilitated this translation process, demonstrating how cross-boundary dialogue can generate shared vocabularies and collaborative routines. The emergence of a transnational network involving over 200 researchers illustrates how distributed agency, once activated, can extend beyond formal project boundaries, though its long-term consolidation remains subject to institutional and funding contingencies.

The findings also resonate with broader policy agendas. The participant-observed shifts documented in this study map onto several Sustainable Development Goals (SDGs), offering a transferable framework for linking institutional EDI practices to wider sustainability trajectories. Mainstreaming inclusive pedagogies and monitoring mechanisms aligns with SDG 4 (Quality Education) and SDG 5 (Gender Equality), while the explicit attention to intersectional dimensions (disability, socioeconomic status, ethnicity) connects to SDG 10 (Reduced Inequalities). The formalisation of EDI roles and fairer career pathways relates to SDG 8 (Decent Work and Economic Growth), and transnational knowledge exchange reflects SDG 17 (Partnerships for the Goals) (Table 2). This dual orientation—organisational retention and societal relevance—strengthens the transferability of the identified mechanisms to other Mediterranean and widening contexts., as mapped onto SDG priorities (see Table 2) and further distilled into actionable institutional and policy insights (Box 1).

Table 2. Linking STEP results to SDGs as a sustainability lens beyond the project lifecycle.

Core Result Evidenced in the Study	SDG	Relevance for Sustainability
Mainstreaming of EDI into research and teaching routines	SDG4 (Quality Education)	Improves inclusive learning environments & equitable access/opportunities
Women’s empowerment pathways in STEM (mentoring, visibility, participation)	SDG 5 (Gender Equality)	Addresses structural gender gaps in careers and decision-making
Explicit intersectional approach (disability, SES, ethnicity)	SDG10 (Reduced Inequalities)	Targets multiple and compounding forms of inequality within institutions
Stabilisation of fairer organisational practices and monitoring	SDG8 (Decent Work and Economic Growth)	Supports inclusive working conditions and organisational well-being
Transnational collaboration and shared tools/guidelines	SDG17 (Partnerships for the Goals)	Reinforces policy learning and diffusion of practices across contexts

Taken together, the analysis suggests that the durability of EDI-related change depends less on project duration than on the progressive alignment of legitimacy, routine, and network connectivity. While the small qualitative sample precludes claims of generalisable impact, the mechanism-focused design offers analytic transferability: it identifies conditions under which time-bound initiatives may seed longer-term adaptation, and highlights the institutional work required to sustain it. These insights carry practical implications for the design of future European cooperation initiatives, particularly in contexts where EDI infrastructure is still emerging.

Box 1. Transferable insights for policy and institutional practice.

• Institutional anchoring matters: EDI actions are more likely to endure when formalised in governance structures—such as dedicated committees or offices, and explicit mandates.
• Routines over one-off events: Sustainability increases when training, monitoring, and communication become periodic routines rather than time-bound activities.
• Budgeting as a sustainability test: The allocation of stable resources, including dedicated staff time and recurrent budget lines, is a strong predictor of post-project retention.
• Measure embedding, not just participation: The focus should shift from counting participants to assessing integration into core institutional processes—HR, curricula, and evaluation procedures.
• Network legitimacy as a catalyst: Transnational peer learning helps overcome local resistance and accelerates institutional change.
• Intersectionality requires data: Disaggregated indicators are necessary to move beyond one-size-fits-all approaches and address intersecting inequalities.
• Leadership endorsement reduces inertia: Visible and sustained support from senior governance actors is particularly critical in hierarchical institutional contexts.

5. Conclusions

Within the limits of this qualitative study, STEP provides evidence that even short-term, modestly resourced European initiatives can promote cultural and institutional change, particularly when they operate through legitimation, routinisation, and network diffusion [27]. In particular, STEP has facilitated knowledge transfer from a transnational perspective, the consolidation of gender mainstreaming, and the adoption of EDI policies. These findings are consistent with previous studies [28,29] and extend them by specifying the mechanisms—legitimation, routinisation, network diffusion—through which such change occurs in Mediterranean contexts.

The added value of STEP lies in its ability to combine local actions with the European context, addressing gaps that remain evident in Mediterranean institutions regarding EDI policies [30]. The Portuguese coordination, based in the widening countries, has shown how realities considered “peripheral” can be particularly receptive and innovative, demonstrating that openness to change does not necessarily correlate with tradition or institutional seniority [30].

Another important element is the interdisciplinary approach adopted, which has enabled EDI policies to be linked to STEM disciplines. This perspective has proved fruitful not only for integrating gender policies, but also for producing new knowledge, in line with observations in the literature that emphasise the effectiveness of intersectional and transdisciplinary approaches in reducing gender gaps in scientific research [11,31].

Nevertheless, the STEP experience highlights the need to ensure structural conditions for the sustainability of the changes initiated. These include the consolidation of EDI offices and practices, systematic monitoring of indicators, involvement of institutional leaders, and allocation of stable resources. Future initiatives should therefore capitalise on STEP’s strengths while addressing its limitations through longer-term funding, more systematic integration of intersectional perspectives, and closer alignment with European and national policies [1,2].

6. Limitations and Transferability

This study is explicitly framed as an exploratory, mechanism-focused qualitative inquiry. Accordingly, several methodological and contextual boundaries should be acknowledged when interpreting the findings.

First, the sample is limited in size (N = 5) and restricted to individuals holding scientific leadership roles within the STEP consortium. While purposive sampling ensured access to information-rich cases directly responsible for EDI implementation and institutional decision-making, it also introduces a potential selection bias: leaders may be more inclined to highlight procedural advancements and underreport institutional resistance or disengagement. The analysis therefore captures participant-observed shifts and institutional routines in formation, rather than a comprehensive audit of organisational change across all staff categories. Future research could complement this perspective by incorporating interviews with early-career researchers, administrative staff, and individuals less directly involved in EDI governance, to map a fuller spectrum of institutional responses.

Second, the temporal scope of the study aligns with the project’s three-year duration, which is inherently shorter than the typical lifecycle of structural academic reform. As such, the findings document early-stage mechanisms of legitimation, routinisation, and network diffusion, rather than consolidated, longitudinal outcomes. This temporal constraint is intrinsic to time-bound EU initiatives, but it also clarifies the study’s aim: to trace how practices begin to stabilise, not to measure their long-term sustainability. The post-project indicator framework proposed in the Discussion offers a pragmatic pathway for follow-up research to assess whether emergent routines endure, decay, or adapt beyond external funding cycles.

Third, interviews were deliberately not audio-recorded to protect participant comfort and ensure candid dialogue within a small, highly visible professional community. While this ethical choice inevitably limited the capture of subtle discursive features (e.g., tone, pauses, or conversational dynamics), it was counterbalanced by a structured two-step triangulation procedure. Detailed contemporaneous notes were systematically cross-verified with written follow-up responses provided by each participant, minimising memory bias and reinforcing data reliability. Although this approach cannot replicate the granularity of verbatim transcripts, it prioritised ethical safety and procedural transparency in a context where anonymity and institutional sensitivity are paramount.

Taken together, these parameters preclude claims of statistical generalisation. The study instead offers analytic transferability: it identifies recurrent mechanisms through which time-limited initiatives can seed institutional and cultural adaptation in contexts with developing EDI maturity. By grounding these mechanisms in neo-institutional and actor-network perspectives, the analysis provides a transferable framework for researchers and policymakers operating in Mediterranean, widening, or similarly resource-constrained academic environments. The methodological constraints acknowledged here do not diminish the study’s explanatory value; rather, they delineate its scope and point toward targeted directions for comparative, longitudinal, and multi-actor follow-up research.