Informed by a critical paradigm and PAR approach, we adopted Bronfenbrenner and Morris’ Bioecological Model (Figure 1) (Bronfenbrenner & Morris 2006; Taylor & Ali 2017) to guide and structure this study. The ongoing and systemic discrimination of SWDs is something that must be understood across many levels, including SWDs’ perception of self (e.g. self-concept; i.e. Individual level), immediate connections (e.g. home supports; i.e. Microsystem level), (lack of) societal and social supports (e.g. societal pressures, peer relationships, mentors and bullying, stigma; i.e. Mesosystem level) and across the pervasive ableist culture (i.e. Macrosystem level). In applying the Bronfenbrenner model to our research, we found that the collectivist nature of Ethiopian culture necessitates placing parents’ friends and extended family within the Microsystem level, rather than the Exosystem level. In Ethiopia, parenting is not left to parents only, but it is a collective task given to extended family and friends. They all engage in child-rearing business with almost with equal engagement in nutrition, safety, security, early learning and care giving (Abesha 2012; Aemro 2015; Tafere & Chuta 2016; Yeshanew 2024). The experiences of SRs must also be set in the context of cumulative discrimination and ableism effects over their lifetime (Chronosystem levels).

In this study, we adopted a collaborative, in situ, and exploratory approach of PAR research principles and methods. By avoiding prescriptive methods for data collection and analysis, we were able to guard against an ableist orientation that could easily become pervasive, condescending and disempowering.

In Phase I of this study, two modules were collaboratively developed and implemented with the six SRs and research team, each on average lasting 4 h. These modules provided foundational knowledge, the opportunity to discuss the value of a collaborative approach, normalise power equality and establish trust for all members of our group.

Student researchers conducted interviews and focus groups with other SRs and SWDs. Initially, SRs were invited to write reflections on their lived experiences, which informed the development of interview questions in conjunction with insights from Phase I. Subsequently, the SRs conducted a total of 18 interviews, including four interviews with fellow SRs, 12 interviews with non-SR SWDs and two focus groups (seven scholars and seven non-scholars) co-facilitated by SRs and faculty researchers with non-SR SWDs who were both Mastercard Foundation Scholars Programme scholars and non-scholars.

Our approach encouraged the active data collection participation of SRs and faculty, aiming to bolster the transparency of our findings through continuous reflexivity (Dodgson 2019) and memoing (Razaghi, Abdolrahimi & Salsali 2015; eds. Reason & Bradbury 2008) by all stakeholders, including SRs, SWDs and faculty. Student researchers from the UoG were engaged without imposing any exclusion criteria related to the duration of SWDs’ studies or the nature of their disability. We used a sample stratification to ensure the inclusion of a diverse sample of SWDs on the research team. Additionally, a sampling grid was developed, considering various factors such as disability type, gender, academic year and discipline to guide the recruitment of SRs as indicated in Table 1.

The recruitment of participants was carried out using multiple methods, including SWDs’ telegram groups, notice board postings and snowball sampling techniques or students’ social networks. The research team, including the SRs, provided prospective participants with detailed information about the study and consent forms in accessible formats, such as audio and written English and Amharic, and braille to ensure a thorough understanding among the SRs. Although English is the medium of instruction at the university, the data collection was conducted in Amharic, which is the first language of the participants. Thus, interpreters for Amharic and sign language were provided in interviews and focus groups to accommodate the linguistic diversity of the participants. Moreover, we made all necessary accommodations, such as Internet data to us Job Access with Speech (JAWS) software and assistants, to facilitate participation, implementing these adjustments wherever feasible.

During the development and implementation of modules, qualitative data were collected using audio and video recordings through Zoom® meetings between researchers and SRs. Additionally, comprehensive meeting notes were taken and subsequently validated by all researchers, providing further context and insights into the meeting. These interactions were grounded in the principles of dialogue, fostering informal and open conversations to build effective working relationships among the research team and SRs (Wosk 2019; Yankelovich 2001). Upon completion of the modules, the recorded meetings and feedback from SRs during the modules were analysed and discussed between the team.

Drawing on the principles of Constructivist Grounded Theory (Charmaz 2006; Denzin & Lincoln 2018), we engaged in an iterative process of transcription, translation (Gawlewicz 2019), thematic analysis (Braun & Clarke 2021) and axial coding (Corbin & Strauss 2015) throughout both phases of this study. All interviews and focus groups were transcribed in Amharic and translated into English by professionals from UoG. All of the Amharic transcribed and translated data were checked and validated by SRs.

Dialogues and reflective sessions with SRs and the research team during the module creation and implementation were instrumental in formulating the interview and focus group queries for the subsequent phase. Following the completion of the two modules, meeting recordings, notes and reflexivity documents were reviewed by the research team, allowing us to narrow and refine our research scope. During the second module, member checking and comprehensive discussions were used to iteratively develop and re-develop the protocols and observation tools. The evolving research questions, tools and methodologies were consistently framed within the Bioecological Model of Bronfenbrenner and Morris (2006).

Through constructivist and critical perspectives, we adopted a cyclical process of data analysis, coupled with the ongoing reflexivity among PAR members, which allowed for a multifaceted exploration of subjective realities at the intersection of various biases. Initally, an analysis manual tailored for the PAR team was developed by a research member with knowledge of PAR methodologies. Ethnographic (Rivoal & Salazar 2013) and informal data, including field notes, reflexivity journals, audio recordings, drawings and photo journals, were collectively contextualised and interpreted within our analysis approach. Our primary analytical method was thematic analysis (Braun & Clarke 2021), complemented by axial coding, inter-rater reliability and reflexivity approaches (Corbin & Strauss 2015; Denzin & Lincoln 2018; Glaser & Strauss 2017; Miles & Huberman 1994).

Six research team members engaged in a collaborative coding process over 5 days to complete preliminary coding and establish an initial codebook. This process involved rigorous discussions with the PAR team and SRs to refine the codebook, ensuring it accurately reflected the experiences and perspectives of SWDs. Each data transcript was coded in full by two research team members, leading to the formation of preliminary codes and categories. Then, the six research team members collaboratively developed the primary themes from the data, with constant input from SRs to validate the analysis and findings. Students with disabilities within our team were provided structured opportunities to critically interpret codes, categories and themes, ensuring the conclusions drawn were reflective of the participants’ realities. NVivo 14 software (Lumivero 2023) was utilised for coding, with inter-rater reliability achieved through the independent coding of interviews by pairs of researchers (Braun & Clarke 2006; Miles & Huberman 1994). Discrepancies in coding between researchers were resolved through discussion until a consensus was reached, ensuring a robust codebook was developed.

Using thematic analysis facilitated the comparison, contrast and synthesis of codes and/or categories across the intersected experiences of our participants, revealing overarching principles and forces conducive to addressing through significant funding avenues like the Mastercard Foundation. Themes derived from critical thematic analysis, grounded in Bronfenbrenner’s Theory, laid the groundwork for our initial findings. Finally, the data were then further analysed against the educational trajectories of SWDs (i.e. elementary and/or boarding, middle, high and post-secondary education phases). This chronologically situated analysis facilitated integrating and juxtaposing emergent themes and categories.

Ethical clearance to conduct this study was obtained from the University of Gondar Institutional Review Board (IRB) (No. V/P/RCS/05/2743/2021).

Findings from SRs and participants have more or less similar lived experiences in the trajectory of primary school to higher education. It is, therefore, possible to say that the findings from these two participants were found to be convergent. Participants painted a compelling and concerning picture of natural and built inaccessibility of schools. Without mobility aids, students were required to navigate hilly and muddy landscapes by ambulating with their hands. As one participant reported:

The lack of accessibility within the natural environment also impacted school attendance and learning and perpetuated stigmatising attitudes. A participant noted how a lack of accessibility was humiliating, physically harming and made it challenging to attend school:

On other occasions, students did not have peer (or other) support and were often left in perilous positions. The physical and educational inaccessibility challenges of the natural environment were exacerbated in communities where:

Built structures, including classroom learning environments, were also challenging for participants, especially once they matriculated into university. There were different experiences in elementary school; some students found the single-level buildings were more accessible as they did not have:

Inaccessibility was specifically challenging for those with visual and auditory impairments as recorders, electronic readers, braille, hearing-aid devices and sign language interpretation were often unavailable. The lack of tools required was compensated by the support of friends, as indicated by one participant who indicated:

Moreover, participants needed to cultivate friendships just so they could access the information needed for their studies. One participant describes this process in detail:

Although strategic approaches such as this are laudable and demonstrate creativity, they are also disempowering and make studying more of a challenge. One participant succinctly sums up the challenge of tool accessibility in elementary and high school by noting:

This sentiment, with some minor exceptions, was ubiquitously described by all participants. Frustratingly, even when tools were available, stigmatisation and/or ignorance resulted in the lack of availability of these tools:

As stated by a participant, part of the problem may have been that teachers lack knowledge. In one participant’s experience:

As a result, when students matriculated to university, where more tools were available, many of the students were unable to productively use these tools and as such were required to learn technological skills while completing schoolwork.

A key challenge in elementary and high school was:

The heroic, strategic and heart-breaking activities participants used to cope with poverty included selling gum, lottery tickets and soft drinks, begging, eating food discarded from restaurants, shining shoes, sleeping in cemeteries and continuing habitation with abusive family members. One participant summed up the compounding effects of financial constraint on their quality of life and quality of education as follows:

During these trying times, many young students:

Families and communities believe that church education is the only ideal place for SWDs. In some cases:

Parents widely held the belief that:

It should be stressed here, in addition to facing stigma and abuse within their families, community members also ostracised both the families and the participants themselves. One participant stated:

This participant recalls his father saying:

On the rare occasions that family members did attempt to support their children, the community would react negatively. Alongside the belief that SWD were unable to learn and engage in society, members of the community and family also believed that disability resulted from sin. One participant recalled that after a childhood scuffle, a peer’s father insinuated that his disability was because of his dangerous character, saying:

Teachers in elementary and high school, as described later, were often key to supporting participants’ transition into positively reinforcing cycles of self-concept development. However, there were also many occasions when teachers stigmatised SWDs through overt abuse and low expectations. A participant related a principal who refused to allow them to sit at the front of the classroom, but instead said that:

One participant recalled a time where the teacher’s abuse took the form of enabling the stigmatisation and ostracisation of others. Teachers also stigmatised students through low expectations and accusing:

One participant noted that in this regard:

Although potentially less detrimental at the university level because of participants’ profound strength and their accurate self-concept as capable students, SR provided a salient example of how this continued stigmatisation is both unfair and frustrating:

In contrast, educational support and the lack of stigma within the boarding school system were clearly identified. Although these schools provided less opportunity to develop relationships with students without disabilities, they were fundamental to student skill development and the development of an accurately positive self-concept as a capable person and student. One participant nicely summarised this sentiment:

As such, there was limited discussion of stigma generally within the Ethiopian culture, and as such, there was little evidence that participants’ direct interactions with macrosystem stigma (e.g. news and media) directly impacted their educational journey. However, participants acknowledged the importance of systemic bias at the macrosystem level:

Male and female participants indicated that sexism and sexual abuse were an additional challenge for females. One participant was:

A male participant also noted that:

This is a particular problem once they leave their family home as:

The outcome of stigma and abuse, especially prior to and during early education, was that participants developed a self-concept of being inferior and unable to interact with peers, unable to learn and ashamed of their disability. The self-concept erosion starts from a negative reaction of parents, relatives and neighbours to a CWD, which leads to rejection and negative emotional and psychological feelings. Self-stigmatisation was a problem as they:

Collaborators had a general feeling of lacking, and as one participant noted:

The isolation and hopelessness that participants felt were effectively summarised by one participant who said:

Threshold facilitators are individuals who provide supportive guidance and play key roles in helping SWDs build resilience throughout their journey, enabling them to reach their current stage. Participants reported a broad range of threshold facilitators that allowed them to start or continue their studies. Most members of their community and/or family (with notable exceptions) offered minimal protection against pervasive social stigma and abuse.

On two occasions, family members supported participants by rejecting familial pressures and enrolling them in primary education. In one case, the participant’s aunt travelled with them to the capital city of Addis Ababa and enrolled her in a specialised boarding school. On another occasion, the participant noted that they:

This participant was forced to flee persistent community and familial abuse, concealment and pressure to attend church school.

All participants indicated the importance of revenging the abuse they received throughout their lives by demonstrating their effectiveness. The juxtaposition of the words positive and revenge connotates the motivational benefits of participants demonstrating their success. The term was coined by SR, who stated:

Positive revenge was demonstrated in many ways. One SR explained the joy he felt in providing financial support to a step-father who had abused him in their childhood by stating:

One participant, whose father was berated by the community for sending him to (public) education, relates the joy of demonstrating:

Finally, participants demonstrated the clear connection between positive revenge and the development of self-concept by indicating:

Matriculating at university was seen by participants as an important milestone, which motivated their progress through lower classes. The motivational importance of university was summarised by one participant who said:

Similarly, another participant demonstrated the importance of university, especially at the end of high school, by indicating:

University matriculation also played an important role in positive revenge motivation. One participant summarised this sentiment by noting:

The focus group discussion was made entirely to get an answer for the research question 2. That is why findings from the analysis of the focus group discussion were not mixed with the in-depth interview analysis presented in the preceding sections. The focus groups that included participants with and without Mastercard Foundation funding highlighted key learning challenges and facilitators at the UoG.

Participants in this study unanimously attributed their success to developing qualities of resilience, perseverance, and advocacy for themselves and others. When they viewed themselves as productive and inspiring members of society, they were able to overcome potentially devastating barriers. A key step to developing resilience was the support of threshold facilitators to begin and/or enable sustainable engagement in education (Ng et al. 2015). In turn, educational success was identified as an important factor in self-destigmatisation, and the development of self-concepts needed to thrive. There were many participant stories, and there is good statistical evidence (Jones, Seager & Yadete 2021), that SWDs who remain concealed, self-stigmatised, and ostracised by society live on the margins. Thus, our PAR participants advocate for early interventions that support SWDs to enrol in secular or regular education. It is important to fund threshold facilitators that can advocate for young SWDs, and demonstrate to communities, families and SWDs alike that PwDs can, have, and continue to be societal leader. University graduates with disabilities are excellent ambassadors (Kimball et al. 2016) and alongside minimal financial support, community engagement and individual mentorship threshold facilitators will be able to address stifling individual and microsystem (self) stigmatisation.

Although MCF-funded participants were grateful for generous financial support at the university, they noted that it was insufficient without reciprocal psychosocial support, which empowers SWDs and improves their accessibility to and belongingness in higher education (Akoto et al. 2022). Moreover, the students who now receive generous allotments first had to survive through periods of starvation and degrading strategies to acquire food and housing. Our PAR participants want to emphasise the importance of sufficient funding across all stages of education, including university. The financial support need only be sufficient to allow SWDs to concentrate on their studies instead of being required to spend time, energy and self-esteem acquiring the means of life.

Tools for physical and sensory accessibility are essential for SWDs. Like any student, SWDs have preferences for which tools to use based on ways of learning, personal skills and educational contexts. There have been many debates in the literature between academics, educators and policy makers regarding how to balance accessibility with socialisation in educational environments and how best to support SWDs. Unfortunately, as identified in our team’s 2022 systematic review, there were no studies that collaborated with SWDs to explore their lived experiences to identify what strategies they needed (Egan et al. 2022). Our results clearly demonstrate that SWDs, especially in upper elementary and high school, are more than able to determine the tools and environments that they need. In fact, collaborators in our work were able to cultivate relationships, improvise accessibility tools and strategies and rationalise unthinkable hardship to meet their goals. Although traumatic, the silver lining expressed by collaborators was self-knowledge, autonomy and self-advocacy. Funders have the opportunity to provide a minimum set of accessibility tools and training, listen to the true experts, the SWDs themselves and, within the limits available, provide tools (even communal tools) that are requested.

Students with disabilities in Ethiopia are required to precociously develop emotional, cognitive, social and psychological resilience and independence. To navigate daily life, they must quickly learn how to challenge societal ignorance rooted in fear and be resilient against prejudice and misconceptions about disability. They are compelled to exceed the performance of their peers to persist in education and overcome the barriers of stigmatisation and inaccessibility. Our participants were frustrated by previous research projects where data were collected about them, and conclusions and recommendations were drawn without them. This resonates with the motto of the 2004 International Day of Persons with Disabilities, ‘Nothing About Us Without Us’ Governmental and institutional conversations about how to serve SWDs rarely engage collaboratively with those they mean to serve. In short, those for whom society demands precocious maturation for survival are continually infantilised by entities that purport to be serving their interests. The community of sharing and mutual support that has blossomed at the UoG provides a clear and present example of the organic strength of SWDs who are given the means to control decisions and organise themselves.

From the overall findings, it can be inferred that although the study was conducted at UoG, the SWDs come from all regions of Ethiopia. This diversity reflects a broader representation of the country within wider continental contexts. Therefore, the findings of the present study can serve as baseline data for further research in other low- and middle-income countries across the continent and beyond. In these contexts, governments and other key stakeholders should strengthen their commitments and take deliberate action to minimise, and, where possible, eliminate, existing power, agency and structural inequalities. Such efforts should begin at the pre-kindergarten level and extend through university education, as well as across all areas of life. These initiatives should actively engage and empower SWDs through a shared vision and common goals to achieve accessible and inclusive education, ultimately fostering inclusive societies and sustainable development across the continent and beyond.