Data Collection Methods for Visually Impaired Populations: A Methodological Review of Accessible Research Practices

Data Collection Methods for Visually Impaired Populations: A Methodological Review of Accessible Research Practices

Full Article ---

Data Collection Methods for Visually Impaired Populations: A Methodological Review of Accessible Research Practices

Abstract

Visually impaired individuals remain systematically underrepresented in empirical research across the social, health, and behavioral sciences. This methodological review synthesizes evidence from 79 sources to map the landscape of accessible data collection practices, critically appraising instrument design, assistive technology mediation, conformance evaluation, and ethical safeguards. We identify five interdependent domains that must be addressed to achieve genuinely inclusive research: epistemological repositioning toward emancipatory frameworks, participant-centered instrument design, modality adaptation for qualitative and quantitative methods, technology mediation through both established and emerging platforms, and ethical–relational safeguards grounded in critical reflexivity. The review reveals persistent gaps between researcher priorities and the expressed needs of visually impaired communities, particularly in digital platform accessibility and the operationalization of participatory principles. We conclude with an actionable methodological roadmap for researchers and institutional review boards.

---

1. Introduction: The Dual Failure of Methodological Exclusion

Visually impaired individuals constitute a population whose systematic underrepresentation in empirical research represents what we term a dual failure: it simultaneously compromises the external validity and generalizability of findings while perpetuating epistemic injustice by silencing a population whose lived experiences are critical to inclusive knowledge production. The problem is not an absence of willingness to participate. When visually impaired people are directly consulted about research, they express a clear desire for greater inclusion and articulate specific priorities for both the content and methods of inquiry (Duckett, 2001). Research commissioned by visually impaired communities themselves has identified that studies should respect individual differences, be practical and action-oriented, and contribute to an empowering research agenda (Duckett, 2001).

Yet despite the proliferation of digital survey platforms and the maturation of assistive technologies, the academic literature has lacked a consolidated methodological review that synthesizes accessible research practices across paradigms. An early review of visual impairment research found a persistent scarcity of emancipatory and participatory work, with the authors concluding that the call for greater use of emancipatory methods within disability studies had not been adequately realized in the visual impairment field (Duckett, 2006). This gap is structural, not incidental: data collection instruments and protocols are rarely designed with the access needs of visually impaired people in mind, and when accessibility is addressed, it is typically through post-hoc accommodation rather than through the principles of universal design.

This review addresses that gap by systematically mapping five interdependent domains of accessible research practice: epistemological repositioning, participant-centered instrument design, modality adaptation, technology mediation, and ethical–relational safeguards.

---

2. Epistemological Foundations: From Extraction to Emancipation

Any methodological review of accessible research practices must begin with epistemology, for the choice of method is never neutral. The dominant research paradigms in social, health, and behavioral sciences have historically positioned visually impaired people as objects of inquiry rather than as co-producers of knowledge. This epistemological stance has practical consequences: it determines who sets research agendas, who designs instruments, who interprets findings, and who benefits from the knowledge produced.

2.1 Emancipatory and Participatory Paradigms

The emancipatory research tradition, grounded in the social model of disability, insists that research must be accountable to disabled people and must actively challenge the conditions that produce disablement. The Lifeworld methodology, for instance, has been evaluated as providing a supportive framework for researchers engaged in emancipatory inquiry, particularly by structuring engagement with the complexity of disabled people's lived experiences (Hodge, 2007). However, its application to visual impairment research remains underdeveloped. The "idea-circles" methodology developed by the Swedish Disability Federation offers a more concrete model: disabled participants, alongside invited researchers, generated 72 research drafts from their lived experience, though participants struggled with the transition from advocacy discourse to operationalized research questions (Erdtman, 2012). This tension—between the authenticity of lived-experience knowledge and the methodological conventions of academic research—is a recurring challenge that accessible research design must navigate.

Comparative insights from adjacent fields are instructive. Emancipatory autism studies have demonstrated how a "loosely formed epistemological community" of neurodiverse researchers can challenge conventional academic knowledge production, co-producing scholarship that troubles normative meanings of academic practice (Rosqvist, 2019). Similarly, service user/survivor-led research in mental health has drawn on standpoint epistemology to argue that knowledge produced by marginalized groups is not merely "alternative" but offers "strong objectivity" precisely because it is self-reflectively partial and socially grounded (Rose, 2017).

2.2 Decolonial Praxis in Disability Research

The decolonial turn in disability research insists that methodological accessibility cannot be disentangled from the geopolitics of knowledge production. Research in Southern contexts reveals specific challenges: identifying sample groups where disability categorization systems differ from Western frameworks, selecting methods appropriate to local communicative norms, and interrogating the underlying assumptions that imported research protocols carry (Singal, 2010). A model for decolonial praxis in community-based data collection, though developed with girls in the juvenile legal system rather than disabled populations, demonstrates that quantitative experimental designs can be anchored in anti-colonial scholarship by training researchers in critical compassion and by conceptualizing research as a site of resistance to hegemony (Singh, 2018).

2.3 Research Priority-Setting by Visually Impaired Communities

A foundational principle of accessible research is that the questions being asked must matter to the communities being studied. When visually impaired and deafblind individuals in the Netherlands were directly engaged in a four-phase participatory priority-setting exercise involving 89 consumers in focus groups and 784 survey respondents, they identified clear hierarchies: in medical research, direct investigation of disease causes and regenerative medicine were prioritized over quality-of-life studies; in the sociopsychological domain, technology improvement and navigation/accessibility of public space emerged as paramount (Scholvinck, 2017). Critically, research priorities showed uniformity across consumer groups, suggesting that collaborative, community-driven research agendas are feasible and need not fragment along diagnostic lines (Scholvinck, 2017).

---

3. Participant-Centered Instrument Design

3.1 Typographical and Visual Accessibility

The legibility of research instruments—whether digital surveys, paper questionnaires, or consent forms—is a foundational yet frequently neglected dimension of accessibility. Research on typographical components of legibility establishes that font sizes of 12–14 point, high text-background contrast, optimized inter-letter spacing, and adequate line spacing (leading) significantly improve reading speed, comprehension, and accuracy (Suroya, 2024). For screen-based instruments, sans-serif fonts such as Georgia, Verdana, and Lucida Sans demonstrate superior legibility, while warm background colors may enhance readability for some users (Suroya, 2024). Critically, both excessively narrow and excessively wide letter spacing can be detrimental, indicating that legibility optimization requires deliberate testing rather than assumption (Suroya, 2024).

Eye-tracking research with low-vision readers provides a more granular understanding. Commercial eye trackers equipped with accessible calibration interfaces can gather gaze data of comparable quality from both low-vision and sighted individuals, revealing distinct gaze patterns associated with low-vision reading that can inform instrument design (Wang, 2023). These patterns—including longer fixation durations and distinctive saccadic movements when using screen magnifiers—suggest that instruments designed for sighted populations may impose cognitive burdens that compromise response quality (Wang, 2023).

3.2 Digital Survey Accessibility and Emoji

Digital surveys introduce additional complexity. Emoji, increasingly used in survey instruments as visual anchors for Likert scales or as communication cues, pose specific challenges for visually impaired users. A mixed-methods study with 58 visually impaired participants found that screen readers vocalize emoji inconsistently, that emoji descriptors often fail to match visual representations, and that the sheer volume of emoji options creates search and selection difficulties (Tigwell, 2020). The study recommends that emoji should never be used for conveying critical information, that the number of emoji per survey item should be limited, and that emoji placement should be carefully considered to minimize screen-reader disruption (Tigwell, 2020).

3.3 Persona-Based Design

A holistic methodology for developing personas with total visual impairment proposes integrating traditional user-experience techniques with neuroscientific approaches to capture both conscious and subconscious dimensions of web interaction (Marin-Alvarez, 2024). This multidimensional approach, adaptable to other disabilities, recognizes that accessibility needs extend beyond surface-level interface adjustments to encompass emotional responses, cognitive load, and anticipatory anxiety that may shape how visually impaired users engage with research instruments (Marin-Alvarez, 2024).

---

4. Qualitative Methods: Adaptation Across Modalities

4.1 Interviews and Focus Groups

The World Health Organization's International Classification of Functioning, Disability, and Health (ICF) has demonstrated utility as a framework for structuring interview schedules with visually impaired populations. In a national survey of 960 visually impaired individuals in Britain, the ICF framework enabled participants to articulate both individually-based barriers (severity of impairment, recent sight loss) and socially-based barriers (inadequate public transport, inaccessible technology) in a participant-centered manner while preserving statistical generalizability (Douglas, 2012). A parallel application of the ICF to computer-use interviews revealed that older visually impaired individuals were significantly less likely to adopt ICT, with both individually-based and socially-based barriers contributing to non-use (Douglas, 2006).

Focus group adaptations for visually impaired children require particular methodological care. Research investigating cognitive-behavioral therapy adaptations with 16 visually impaired children aged 9–13 in South Africa found that focus groups effectively elicited themes around difficulties, coping strategies, and conceptual understanding of feelings, thoughts, and behaviors—but that standard CBT materials required modification to function in non-visual modalities (Visagie, 2017).

4.2 Co-Production and Participatory Methods

Co-productive qualitative methods, in which visually impaired participants are positioned as active collaborators rather than passive informants, have demonstrated particular value in healthcare research. A study involving seven older people with vision impairment in researcher-facilitated workshops with healthcare students generated four themes—Trauma and Loss, Vulnerability and Feeling Invisible, Being Disabled, and Feeling Safe—that would likely have been inaccessible through conventional interview methods alone (Wilson, 2024). The study's adherence to the COREQ checklist and involvement of the Sheffield Royal Society of Blind in planning demonstrates that co-production can be methodologically transparent and ethically rigorous (Wilson, 2024).

Participatory video and diagramming methods offer alternative qualitative approaches, particularly in Global South contexts. These methods can create spaces for disabled individuals to express experiences and challenge oppression, making lives more visible than through conventional interviews. However, scholars caution that participatory methods are not immune to issues of representation and power dynamics, particularly when Western, non-disabled researchers are involved (Bezzina, 2022). This "seductive nature" of participatory research—the tendency to misapply it as a panacea for extractive research dynamics—requires persistent critical vigilance (Oliveira, 2020).

---

5. Technology Mediation: Platforms and Assistive Tools

5.1 Social Media Platforms as Data Collection Venues

Digital platforms increasingly serve as sites for survey distribution and ethnographic data collection. Comparative analysis of Facebook, Instagram, and Twitter accessibility for visually impaired students in Saudi Arabia reveals statistically significant differences: Facebook achieved the highest effectiveness (4.2/5) and usability (4.0/5) ratings, followed by Instagram (4.1 and 3.9), with Twitter receiving the lowest scores (3.8 and 3.7) (Khasawneh, 2023). ANOVA confirmed significant differences between Facebook and Twitter (p = 0.031) and between Facebook and Instagram (p = 0.027), suggesting that platform selection has measurable consequences for data quality and participant experience (Khasawneh, 2023).

The flourishing of blind content creators on TikTok—documented through interviews with 45 "BlindTokers"—demonstrates that visually impaired individuals actively leverage platform affordances for identity expression and community building, engaging in what has been theorized as "flourishing labor" (Lyu, 2024). This finding suggests that digital ethnographic methods may be productively conducted within platforms where visually impaired communities already maintain robust presence, provided researchers approach these spaces with cultural competence and ethical sensitivity.

5.2 Virtual and Augmented Reality

Virtual reality holds promise as both a data collection environment and a tool for accessibility research. A user-centered VR platform for library orientation, incorporating spatial audio cues, adjustable text, customizable lighting and contrast, achieved a System Usability Scale score of 67.83 among visually impaired students, with qualitative feedback emphasizing the value of spatial audio while recommending improvements in contrast personalization and audio speed control (Yensathit, 2025). VR has also been deployed as a testbed for electronic travel aids, simulating age-related macular degeneration, diabetic retinopathy, and glaucoma in healthy participants to evaluate haptic feedback devices in controlled, safe environments (Ricci, 2023).

A comprehensive review of inclusive immersion efforts in VR, AR, and the metaverse identifies key design principles directly applicable to accessible data collection platforms: output and input redundancy (providing information through auditory and haptic channels simultaneously), integration of assistive technologies, customizability, enhanced assistance, and inclusive design from inception (Dudley, 2023).

5.3 AI Assistants and Screen Reader Compatibility

The assistive technology landscape is being reshaped by artificial intelligence. A systematic review of 646 studies on assistive AI for blind and low-vision individuals, triangulated with interviews with 24 BLV participants, found only a "weak positive correlation" between researcher focus and participant-perceived importance of tasks (Gamage, 2023). Participants expressed preference for conversational agent interfaces and head-mounted devices—preferences that diverge from the smartphone-centric orientation of much assistive technology research (Gamage, 2023). AI assistants designed through co-design methodologies with blind users are being developed to enhance accessibility of productivity applications such as word processors and spreadsheets, directly relevant to the platforms on which researchers build and administer instruments (Perera, 2024).

However, the systematic review of AIoT (Artificial Intelligence of Things) in assistive technology reveals methodological limitations that should temper enthusiasm: the field relies heavily on simulated data, employs small samples, and lacks field validation, while technical challenges of device interoperability remain unresolved (Vusi, 2025).

Qualitative data analysis software—the "back end" of qualitative research—presents its own accessibility barriers. Visually impaired researchers face substantial challenges in utilizing commercial Q-DAS platforms, with problems rooted in design choices that assume visual modes of interaction for coding, memoing, and visualization of thematic networks (Emara, 2023).

---

6. Conformance Evaluation: Testing Accessibility

The evaluation of instrument and platform accessibility requires a dual approach. Automated accessibility checkers can identify technical conformance failures (e.g., missing alt text, incorrect heading hierarchies, insufficient color contrast), but they cannot assess the experiential quality of participation. User-centered testing—in which visually impaired participants complete instruments while providing concurrent or retrospective verbal protocols—captures the phenomenological dimension that automated tools miss.

The ICF framework, originally developed for health classification, provides a robust evaluative vocabulary by directing attention to participation restrictions and environmental barriers rather than locating deficits within individuals (Douglas, 2006). When applied to technology-mediated data collection, the ICF directs evaluators to examine not merely whether a platform is technically screen-reader-compatible, but whether visually impaired participants can engage with it on terms of equality—including factors such as time burden, cognitive load, and the dignity of the interaction.

The Network 1000 project, a three-year panel survey of 1000 visually impaired individuals in Britain, exemplifies large-scale implementation of inclusive evaluation principles. Its methodology is characterized by visually impaired individuals' active involvement in all research stages, including generating data and identifying themes that shaped the main survey instrument, with a two-stage sample design operationalized to ensure democratic participation (Corcoran, 2004).

---

7. Ethical–Relational Safeguards

7.1 Shared Identity and Researcher Positionality

The question of who conducts research with visually impaired populations is not merely logistical but epistemological. Autoethnographic evidence from researchers who share the identity of visual impairment with their participants suggests that shared identity significantly enhances both recruitment rates and the quality of rapport during interviews (Emara, 2025). This "insider-outsider" position, however, requires critical reflexivity: the disabled researcher must navigate the tension between experiential affinity and scholarly distance, recognizing that shared impairment does not automatically confer shared perspective across lines of class, race, gender, and geographic location (Beckwith, 2022).

7.2 Informed Consent and Accessible Communication

Informed consent procedures must themselves be accessible. The research reviewed here indicates that multimedia dissemination—including audio, Braille, and large-print formats—is essential, and that consent should be understood as an ongoing process rather than a single signed event (Duckett, 2001). In co-productive research, consent extends to the collaborative interpretation of findings; participants should have meaningful opportunity to review, challenge, and contribute to how their data are represented.

7.3 Institutional Review Board Preparedness

Institutional review boards frequently lack specific expertise in evaluating the accessibility of proposed data collection protocols. An actionable ethical–technical framework for IRBs should include: (a) requirement for documented accessibility testing of instruments prior to deployment; (b) evidence of visually impaired community consultation in instrument design; (c) provision for multiple modalities of participation (phone, in-person with assistive technology, screen-reader-optimized digital); and (d) protocols for identifying and addressing accessibility failures that emerge during data collection, rather than treating accessibility as a static pre-condition.

---

8. Synthesis: A Methodological Roadmap

The evidence synthesized in this review converges on a set of actionable principles:

First, accessible data collection begins with epistemological humility. Researchers must cede the presumption that they know what questions matter. Community-driven priority-setting, whether through idea-circles (Erdtman, 2012), participatory ranking exercises (Scholvinck, 2017), or co-productive workshop designs (Wilson, 2024), should precede instrument development.

Second, instrument design must attend to the granular details of typography, contrast, spacing, and digital interface accessibility (Suroya, 2024), and must avoid assuming that visual elements—including emoji—are universally interpretable (Tigwell, 2020).

Third, modality pluralism should be a default, not an accommodation. Offering participants choice among phone, screen-reader-optimized digital, in-person, and tactile formats is a methodological strength, not a confound.

Fourth, technology selection requires evidence. Platform accessibility varies significantly (Khasawneh, 2023), and there exists a documented gap between what assistive technology researchers prioritize and what visually impaired people actually want (Gamage, 2023).

Fifth, ethical practice demands critical reflexivity about researcher positionality (Emara, 2025), the seductions of participation (Oliveira, 2020), and the structural conditions—including the geopolitics of knowledge production—under which research is conducted (Singal, 2010).

---

9. Conclusion

The systematic exclusion of visually impaired populations from empirical research is not an inevitability of impairment but a product of methodological choices that can and must be corrected. This review has demonstrated that the conceptual, technical, and ethical resources for accessible data collection already exist—distributed across disability studies, human–computer interaction, participatory methodology, and critical epistemology. What remains missing is their systematic integration into the routine practice of researchers and the evaluative standards of institutional review boards. Accessible research is not a niche methodological concern; it is an imperative of scientific validity and social justice. The roadmap offered here provides researchers, funders, and ethics committees with a framework for aligning data collection practices with the principles of universal design and equitable participation, thereby strengthening both the rigor and the moral legitimacy of the research enterprise.

---

Figure 1 depicts the five-layer methodological ecosystem synthesized from this review: epistemological foundations (emancipatory, decolonial, participatory), participant-centered instrument design (typography, emoji, personas), modality adaptation (interviews, focus groups, co-production, participatory video), technology mediation (social media platforms, VR/AR, AI assistants, screen readers, eye tracking), and ethical–relational safeguards (shared-identity dyads, accessible consent, IRB protocols). Bidirectional arrows indicate that evaluation findings must recursively inform design revisions, while community priorities must continuously reshape research agendas.