AI-enabled qualitative analysis of digital gender exclusion

Digital gender exclusion is a complex social problem that qualitative methods expose through interviews, focus groups, and co-design; the primary keyword is digital gender exclusion qualitative analysis. The PLoS One study protocol by Shahil-Feroz et al. (2026) investigates why marginalized women in Azam Basti, Karachi are offline and how co-designed solutions can promote inclusion. This post refracts that protocol through the lens of AI-enabled qualitative research, showing how AI transcription, translation, thematic coding, and cross-segment analysis can accelerate rigorous synthesis for researchers and program designers.

Key Takeaways

According to PLoS One (Shahil-Feroz et al., 2026), gendered social norms and household gatekeepers are central drivers of why marginalized women in Azam Basti, Karachi lack meaningful digital access, and the study tests a community-based participatory approach to co-design solutions.

The PLoS One protocol plans a purposive qualitative sample of approximately 36–45 participants across marginalized women, family gatekeepers, and community leaders, with recruitment scheduled for August–November 2026.
The PLoS One authors cite the GSMA Mobile Gender Gap Report 2024 showing women in Pakistan were 38% less likely than men to own a mobile phone or access mobile internet, and Pakistan ranked 145 of 146 in the 2024 Global Gender Gap Index, according to the study.
The project begins February 2026 and expects data collection completed by April 2027 and results available beginning in late 2027, making rapid, transparent qualitative analysis essential for timely policy input.
The protocol explicitly targets sociocultural and structural barriers beyond device access, arguing that “deeply entrenched sociocultural barriers” and family control require co-designed, household-level strategies (Shahil-Feroz et al., 2026).

What happened and how the study will work

Answer: The PLoS One protocol (Shahil-Feroz et al., 2026) lays out a three-phase community-based participatory research (CBPR) study in Azam Basti, Karachi that uses focus groups, in-depth interviews, and co-design workshops to explore digital gender exclusion.

According to Shahil-Feroz et al. in PLoS One (2026), Phase 1 forms a Community Advisory Committee (CAC) with two community leaders, two marginalized women, and three family gatekeepers to co-lead the work and validate instruments.

According to Shahil-Feroz et al. (2026), Phase 2 will run two focus group discussions per participant group (4–5 participants per group), plus 4–5 individual interviews per group, for an estimated total sample of 36–45 participants, using photo-elicitation for women and role-play for gatekeepers.

According to Shahil-Feroz et al. (2026), Phase 3 runs three co-design workshops with 8–10 participants each (25–30 total) to co-create 2–3 context-specific strategies, and the team will use reflexive thematic analysis with two independent coders and NVivo for organization.

According to Shahil-Feroz et al. (2026), study materials include images generated with an AI tool: "All images are generated through Open AI ChatGPT image generation tool", and transcripts will be produced in Urdu then translated into English for analysis.

Method constraints: The PLoS One authors note the single-community design may limit generalizability and that social desirability bias can shape responses; the team plans reflexive memos, CAC validation, and ethical safeguards including informed consent and multiple ethics approvals.

Findings snapshot

Date	Metric	Value	Implication
2024	GSMA Mobile Gender Gap metric cited in the protocol	Women in Pakistan 38% less likely than men to own a mobile phone or access mobile internet	Quantifies the access gap the qualitative study aims to explain at the household and cultural level
2024	Global Gender Gap Index (World Economic Forum) ranking cited	Pakistan ranked 145 of 146 countries in 2024	Contextualizes deep structural gender inequalities shaping technology access
Feb 2026	Project start (planning)	Two-year project begins Feb 2026; prep Feb–Jul 2026	Sets timelines that require efficient transcription and analysis pipelines
Aug–Nov 2026	Participant recruitment window	Recruitment expected Aug–Nov 2026 for ~36–45 participants	Highlights need for rapid, multilingual transcription and coding
Apr 2027	Expected completion of data collection	Data collection completed by Apr 2027; results expected late 2027	Signals a ~5 month analysis and dissemination window where AI tools speed synthesis

Implications for qualitative researchers and program designers

Answer: The PLoS One protocol implies that rigorous, participatory qualitative research must combine culturally sensitive methods with timely analysis to inform policy and programs.

Practical implication 1: According to Shahil-Feroz et al. (2026), engaging family gatekeepers and community leaders shifts interventions from individual-level literacy programs to household and community norms change, which should change how researchers design sampling and facilitation.

Practical implication 2: According to Shahil-Feroz et al. (2026), photo-elicitation and role-play generate mixed-media data (images, enactments, audio) that require robust transcription, translation, and multimodal coding workflows to capture meaning across languages and registers.

Practical implication 3: According to Shahil-Feroz et al. (2026), co-design workshops produce iterative solution prototypes; researchers and NGOs should plan for rapid, transparent synthesis and stakeholder-ready outputs such as policy briefs and arts-based materials.

How Evidano helps: accelerating credible qualitative synthesis

Evidano definition

Evidano is an AI-powered qualitative data analysis platform that helps researchers analyze interviews, open-ended surveys, and documents.

Problem: Multilingual transcription and translation. Solution: Evidence from the PLoS One protocol shows transcripts will be in Urdu and translated to English; Evidano offers automated speech-to-text with custom dictionaries and translation to speed accurate bilingual workflows and reduce backlogs.

Problem: Volume of heterogeneous data (FGDs, role-play recordings, photo-elicitation). Solution: Evidano organizes audio, transcripts, and images into searchable corpora and applies thematic coding plus cross-segment frequency analysis to compare marginalized women, gatekeepers, and leaders.

Problem: Need for rapid team coding and community validation. Solution: Evidano supports collaborative codebooks, hierarchical codes→subcodes visualizations, and AI-assisted thematic summaries to accelerate reflexive thematic analysis while preserving audit trails for rigor.

Problem: Tight timelines for policy-ready outputs. Solution: Evidano’s exportable dashboards and AI chat over your documents enable researchers to produce stakeholder briefs and visualizations faster; see the product features.

Problem: Manual transcription bottleneck. Solution: Evidano’s speech-to-text pipeline with PII redaction and custom dictionaries fits the protocol’s need for secure, accurate Urdu→English transcription and speeds analysis.

FAQ: digital gender exclusion qualitative analysis

How can qualitative methods identify why marginalized women are offline?

Answer: Qualitative methods reveal beliefs, power relations, and household decision rules that explain why women lack access.

According to Shahil-Feroz et al. in PLoS One (2026), focus groups with photo-elicitation and role-play with gatekeepers uncover norms and safety concerns that surveys miss, enabling actionable co-designed solutions.

What sample size and designs are appropriate for this question?

Answer: Purposeful, small-sample CBPR designs that reach theoretical saturation are appropriate for in-depth understanding.

The PLoS One protocol plans approximately 36–45 participants across FGDs and interviews and notes saturation will guide the final counts, which matches qualitative best practice for exploratory, community-rooted work (Shahil-Feroz et al., 2026).

How should teams manage Urdu audio, translations, and coding at scale?

Answer: Use secure, automated transcription and bilingual review workflows plus AI-assisted coding to preserve nuance and speed analysis.

The PLoS One team will transcribe in Urdu and translate to English for reflexive thematic analysis; platforms that offer custom dictionaries, PII redaction, and bilingual verification reduce error and support reproducibility.

Can AI tools be used ethically in participatory research with marginalized groups?

Answer: Yes, if teams apply strong ethics, informed consent, and data governance safeguards.

The PLoS One protocol obtained ethics approvals and emphasizes consent and confidentiality; similarly, researchers should use encrypted platforms, obtain explicit consent for recording and AI-assisted processing, and include community advisory review for outputs.

Conclusion & Next Steps

The PLoS One protocol by Shahil-Feroz et al. (2026) makes clear that sociocultural norms and household gatekeepers drive much of the digital gender exclusion observed in Pakistan, and that participatory, co-designed interventions are required to change those norms.

For qualitative researchers and implementers working on digital inclusion, combining rigorous CBPR methods with AI-enabled transcription, translation, and thematic analysis shortens the path from fieldwork to policy-ready recommendations.

If you are planning a similar study and want secure, bilingual transcription, shared codebooks, and AI-assisted thematic synthesis, explore how Evidano integrates transcription, translation, and collaborative analysis in one platform: Try Evidano for free.

Topics

digital gender exclusion qualitative analysis
gender digital exclusion Pakistan
qualitative methods digital inclusion
AI qualitative analysis

Keep reading

Browse all articles