Evidano home
Commentary on News

Qualitative Analysis of Admissions: Design Thinking

Evidano6 min read

Admissions teams face new complexity: GenAI, remote interviews, and the loss of standardized tests. A two-hour design thinking workshop at UNC (January 2025) generated approximately 76 ideas and clustered them into three themes: interview format, GenAI integration and compliance, and broadened admissions criteria (Davidson et al., published June 23, 2026). Read the original study at PLOS ONE. This post shows how to convert DT artifacts into rigorous qualitative analysis of admissions using AI tools to go from sticky notes to reproducible insight.

Key Takeaways

This post shows how to convert design-thinking (DT) workshop artifacts into reproducible qualitative analysis of admissions and how Evidano supports that workflow.

  • A purposive group (n = 15) at UNC ran a 3‑phase DT workshop in January 2025 that generated approximately 76 raw ideas, 43 advanced ideas, and 4 rapid prototypes (Davidson et al., published June 23, 2026).
  • Use inductive coding with double-coding and percent-agreement checks to ensure trustworthiness, the UNC study reports 90.5% inter-coder agreement.
  • A reproducible workflow includes digitizing artifacts, normalizing metadata (role, vote counts, station), AI-assisted seed extraction, and exporting intercoder metrics and auditable quotes.

Fast take + source

Fast take: A purposive group (n = 15) at UNC used a three-phase design thinking workshop in January 2025 to surface candidate-evaluation ideas amid GenAI and test sunsetting, and the paper reporting the workshop was published June 23, 2026.

Source: PLOS ONE

Findings snapshot

MetricValueSource / Note
Workshop dateJanuary 2025In-person DT session at UNC Eshelman School of Pharmacy
Participantsn = 15 (94% survey response)Faculty, admissions staff, reviewers, student ambassadors
Ideas generated≈76 total; 43 selectedVoting used; top votes per idea ranged 0–13
Survey highlights (means ± SD)Creative problem solving 4.7±0.5; Idea generation 4.8±0.4; Quality of ideas 4.9±0.4Post-workshop Likert items (1–5)
Inter-coder agreement90.5%Inductive coding; two coders

What happened: workshop & analysis (plain English)

The workshop ran a structured design thinking sequence (inspiration, ideation, prototyping) and the analysis converted multi-format artifacts into coded themes.

The team refined three launching points: fair technology use, reliable candidate comparison post-PCAT, and strategic use of technology by evaluators.

Data sources included brainstorming sheets, group charts, reflection worksheets, facilitator notes, and a brief Likert plus open-text survey.

  • Analysis: materials were transcribed and coded via reflexive thematic analysis, with percent agreement 90.5%.
  • Outputs were generative (ideas and prototypes) and were not implemented; the paper describes methods and themes rather than implemented outcomes.
  • Key tensions surfaced: equity versus on-the-spot assessments, remote convenience versus integrity, and whether and where GenAI should be permitted.

So what for researchers and admissions teams: apply qualitative analysis of admissions

For UX & qualitative researchers

UX and qualitative researchers should digitize and standardize DT artifacts to create an analyzable corpus.

DT workshops produce dense, multi-format artifacts including handwritten notes, charts, and votes, so turn these into transcripts, images, and CSVs.

Use inductive coding with double-coding and percent-agreement checks, the UNC study reports 90.5% agreement, to ensure trustworthiness and reproducibility.

For admissions directors

Admissions directors should capture reproducible evidence for later evaluation and pilot testing.

Prioritize reproducibility by capturing raw prompts, vote counts, and constraints (cost, geography, equity) so later evaluation ties back to prototype rationale.

Track candidate-facing policy ideas, for example real-time writing stations or GenAI critique tasks, as discrete interventions to pilot and measure.

Ethics & limitations (short)

Ethics and limitations require proactive assessment before operational pilots.

This study is exploratory and non-diagnostic; any operational pilots should include equity impact assessment and ADA and legal review before deployment.

Do more, faster with Evidano, mapping DT workshop outputs to an AI-enabled qualitative workflow

Problem: Messy, multimodal DT artifacts → Solution: Ingest & unify

Evidano is an AI-powered qualitative data analysis platform that ingests multimodal DT artifacts, supports built-in transcription and PII redaction, and unifies noisy text into a single analyzable corpus.

Upload photos of chart paper, handwritten notes, and typed worksheets and use built-in transcription with a custom dictionary to convert noisy text into a single corpus for analysis.

Problem: Inconsistent coding across coders → Solution: AI-assisted codebook + intercoder metrics

Solution: Seed your codebook and run AI-assisted coding to standardize application and measure agreement.

Import your codebook or let Evidano generate a seed codebook from samples, run AI-assisted coding, and export percent-agreement and per-code frequencies to reproduce the study’s 90.5% check quickly.

Problem: Hard to compare segments (roles, vote counts, prototypes) → Solution: Cross-segment analysis & visualizations

Solution: Use cross-segment analytics to reveal stakeholder differences at a glance.

Produce thematic frequency tables, co-occurrence networks, and hierarchical code to subcode visualizations so differences (faculty versus students) are visible quickly.

Problem: Need rapid, defensible reporting → Solution: Clickable quotes, exportable reports

Solution: Generate auditable, exportable reports that link quotes to original artifacts for member-checking and IRB-ready documentation.

Generate shareable summaries and export CSVs of quotes linked to original artifacts for member-checking and documentation; Evidano does not train third-party models and stores data encrypted.

Problem: Want to follow up with participants → Solution: AI avatar interviews

Solution: Run scalable follow-ups without hiring additional interviewers.

Run autonomous avatar interviews to collect participant feedback or run pilot probes on prototype stations at scale.

Checklist: 7-step workflow to reproduce the UNC DT analysis with AI

Use this reproducible checklist to go from sticky notes to triangulated, auditable insights.

Step-by-step to go from sticky notes to reproducible insights:

  • 1) Digitize all artifacts (photos, scans, typed worksheets).
  • 2) Ingest into Evidano; apply transcription with custom dictionary and PII redaction.
  • 3) Normalize metadata (role, vote count, station, constraint tags: cost, geography, equity).
  • 4) Run seed thematic extraction; review and refine codebook with domain experts.
  • 5) Apply AI-assisted coding across the corpus; export inter-coder agreement and adjust rules.
  • 6) Produce cross-segment analyses, co-occurrence maps, and top quotes per theme.
  • 7) Share a two-page member-check summary and schedule pilot tests for one to two highest-priority prototypes.

FAQ: qualitative analysis of admissions

Q: Can AI detect GenAI-written application text?

A: Detection is noisy and not definitive.

The UNC workshop suggested real-time tasks (for example on-stage writing) to reduce GenAI use, and for archival materials combine stylometrics with manual review and policy disclosures.

Q: How do I compare vote-weighted ideas across groups?

A: Capture vote counts as structured fields and run weighted-priority analyses.

Capture vote counts as structured fields and run frequency plus weighted-priority analyses to show which ideas had traction across stakeholder groups.

Q: Is it safe to run admissions data through AI tools?

A: Use platforms with end-to-end encryption and clear model-training policies.

Choose platforms with end-to-end encryption and explicit policies on model training; Evidano stores data encrypted and does not use customer data to train third-party models.

Wrapping up & next steps (CTA)

Wrapping up: Turn DT workshop artifacts into auditable, cross-segment qualitative analysis to prioritize pilots and defend policy decisions.

Start a pilot: upload one DT session (photos and worksheets) and produce a reproducible codebook, inter-coder metrics, and a two-page summary in under a day with Evidano homepage.

Read the original study: PLOS ONE (Davidson et al., published June 23, 2026).

Ready to convert workshop outputs into action? Try Evidano for free.

Topics

  • qualitative analysis of admissions
  • design thinking admissions
  • GenAI in admissions
  • intercoder agreement qualitative
  • Evidano qualitative platform

Keep reading

Browse all articles

Company

HomeContactBlogsGuides

About

FeaturesThematic Analysis SoftwareAI Thematic AnalysisAI Qualitative CodingQuote TraceabilityMultilingual AnalysisCompare QDA ToolsPlansCase StudiesData SecurityHow We Trained Our AI ModelsAffiliate ProgramHow to Cite & Journal AI PoliciesCompare Evidano with Other SolutionsCertified TrainersHuman vs AIEnvironmental Impact

Preview Evidano

View sample AI docx/pdf analysisView sample AI xlsx/csv analysisView sample AI audio transcriptionTry a demo AI survey / interviewPreview AI ChatbotAI Detection

Account

LoginRegister for Free Plan (free forever, no expiry)Book demo

Newsletter

© Evidano. All rights reserved.

Privacy PolicyTerms of Service