Alisha Giri Research Portfolio

Research Portfolio

Alisha Giri, PhD, MPH

I study how people make sense of, trust, and act on uncertain information from new health technologies, and where the risks and downstream harms fall.

Alisha Giri

Mixed-methods researcher trained in medical anthropology and public health, with 8 years of applied research in academic and nonprofit settings.

alishagiri.comORCIDEmail
Current · Stanford Center for Biomedical EthicsGenomics: risks & downstream harmsSocial & behavioral genomics · 7 stakeholder groups Current · Stanford Center for Biomedical EthicsEthics of rare disease drug development & the value of a diagnosisInterviews · cognitive interviews · survey design Project 3 · CWRUNutrition incentive program evaluationImplementation research · program change Project 4 · DissertationSuccess & livelihood outcomes in rural NepalNSF-funded · ethnography · network analysis Project 5 · PAMFTeen Lifestyle BalanceHealth coaching · usability testing
Data sources & methods

Quantitative

Data sourcesSurveys & questionnaires · Secondary data (BRFSS, NHANES, HCUP, CMS, ICPSR) · Network data
Techniques & toolsDescriptive & inferential statistics · Social network analysis · R, SPSS, Qualtrics, REDCap, ArcGIS

Qualitative

Data sourcesInterviews (semi-structured, in-depth, longitudinal) · Focus groups · Participant observation · Ethnography · Usability tests · Process audits
Techniques & toolsThematic analysis · Grounded theory · CFIR · NVivo, Dedoose, MaxQDA · LLM-assisted synthesis
7stakeholder groups
10+in-depth interviews per group
2-passprimary & secondary coding
1publication, Behavior Genetics 2025
Risks & benefits of genomic screening for behavioral traits Social & behavioralgenomics researchers Academic journaleditors Mediajournalists Direct-to-consumerindustry leaders Parents who tested orare considering testingfor a behavioral trait Parents of childrenwith ADHD or dyslexiainterested in screening K–12educators
Seven stakeholder groups, at least ten in-depth interviews each: the people who produce, publish, report, sell, teach around, and buy genomic screening.
My role (joined mid-study)
Recruit & interviewRecruited the parent and educator groups; conducted interviews across groups
Build the codebook with the PIPrimary and secondary coding of the interview corpus
Write & analyzeDrafted the manuscripts; support the ongoing analysis
Editors as mediators, journalists as translatorsEditors saw themselves shepherding research between authors and reviewers; journalists as translators, and sometimes interrogators, of science.
Everyone agrees it's risky; nobody owns the riskBoth groups called the field ethically sensitive and prone to misinterpretation, said systematic guidelines were lacking, and deferred responsibility to others.
The gap is a framework, not awarenessThe paper calls for explicit publishing and reporting frameworks for the responsible communication of social and behavioral genomics.

Martschenko, D.O., & Giri, A. (2025). A qualitative study of the roles and responsibilities of academic and journalistic publishing in social and behavioral genomics. Behavior Genetics. DOI · Findings from the other five stakeholder groups are in analysis and not shown here.

A pattern across my Stanford interviews

When clinicians don't have time to explain, patients, caregivers, and consumers turn to AI to interpret their genetic results and work out what they mean.

People without clinical expertise described using AI tools to fill the communication gap: to make sense of screening findings, understand implications, and get answers they weren't getting elsewhere. That gap, and what fills it, is the design problem for any AI health product.
200participant target across the ecosystem; final interviews under way
36parents, up to 3 sequential interviews each
10cognitive interviews on the draft survey
4iterations of the instrument
3collaborators: PI, genetic counselor, statistician
Ethics of rare disease drug development Patients &caregivers Patient advocacygroup leaders (majority) Biotech & pharmaprofessionals Federal regulatorsNIH · FDA IRBmembers
Interviews follow a therapy from bench to bedside: the people who live with the disease, advocate for it, develop and fund the drugs, and regulate them.
Building the caregiver instrument
Qualitative groundworkParent interviews + related studies Existing scalesFlanagan QoL + health/QoL instruments Draft instrumentcaregiver QoL + value 10 cognitive interviews4 iterations Pilot (now)Healthy volunteers + caregivers Validate with statisticianPopulation-specific instrument Interviews inform items · cognitive interviews test them · statistics validate them
Feeling heard is designableParents' experience turned on providers' availability and responsiveness, validation of their concerns, and continuity across providers.
People respond to silence in opposite waysSome parents escalated (persistence, advocacy); others withdrew and resigned. Same system, divergent strategies.
Breakdowns weren't universalWhich means the parent–provider interaction is an intervention point, not a fact of life.

Nguyen, M.M., Mahfoozi, S., Bonner, D., Martschenko, D.O., Giri, A., et al. (2025). "Nobody listened to us for years": Parents' experiences of provider communication in the diagnostic odyssey. Molecular Genetics and Metabolism. DOI · One manuscript under review; others in progress.

AI health coachTwo things this project trains: designing an instrument that captures what a health outcome is worth to the person living it, and studying the moment a system fails to listen.

100+Produce Perks sites statewide; 6 counties evaluated
3user groups: SNAP customers, register staff, store owners
+11%participating households
+19%fruit & vegetable redemptions
Recruit3 user groupsSNAP customers · grocery employees · store owners
CollectAudits, focus groups, interviewsIn-store audits · semi-structured focus groups · structured interviews
Pivot · COVID-19Remote interviewsIn-person focus groups → phone & Zoom; timeline kept
AnalyzeCFIRCodebook · thematic analysis · Dedoose, SPSS
DeliverBriefs & presentationsRecommendations to statewide partners
Key leverage points diagram: grocery store solutions, cashier burden, community solutions
Key leverage points from the evaluation: where the program stalled (cashier burden, no incentive to promote it) and the fixes proposed for stores and community partners.
The barrier was awareness, not priceMany eligible customers didn't know the program existed.
Cashiers were the missing channelNo incentive, no training, no easy way to promote it at checkout.
Fix the information, not the benefitPOS integration, in-bag promo cards, staff training, and trusted community partners (WIC, SNAP-Ed, food pantries).
ImpactProgram changes informed by the evaluation were associated with 11% more participating households and 19% more produce redeemed. Research briefs, stakeholder presentations, MPH capstone, manuscript under review.

AI health coachA well-designed intervention underperforming for a reason nobody had named; implementation research named it, mid-pandemic, without losing rigor.

Study participant holding a goat kid in rural Nepal Rice planting in a village field Field work in a flooded paddy Community members in a field during fieldwork
180participants · 60 in Phase 1, 120 in Phase 2
15villages
40m–2hsemi-structured interviews, in Nepali
120surveys with network & life-history data
3research assistants trained & managed
Pre-fieldworkPrepareLiterature review · key stakeholders & collaborators
Phase 1 · n = 60Local metrics of successParticipant observation · semi-structured interviews
Phase 2 · n = 120Networks & livelihoodsSurveys · network data · life histories
AnalysisSynthesizeThematic analysis (NVivo) · descriptive stats · SNA (R)
SEEDROUND 1ROUND 2ROUND 3R4 Migrant returneeNon-migrant
Network sampling: each seed named counterparts across four rounds, alternating returnees and non-migrants.
What does it mean to be successful in rural Nepal?
OccupationalGovernment jobs · Entrepreneurship · Army (Nepal, Indian, British)
CharacterHard-working · Honest · Resilient · Employs or teaches others · Healthy · Happy
MaterialCement housing · More land · Proximity to a town · Personal vehicle
Success is holisticFinancial stability plus community contribution, health, well-being, and the next generation.
Migration is a strategy, not an outcomeA bridge to a sustainable life at home for those with few local options.
Networks do the workAccess to capital, cultural continuity, and opportunity, locally and abroad.
Policy favors leavingOut-migration and remittances are rewarded; local development is not.
ImpactFindings identified an unmet need for a local market. With community stakeholders, I helped establish a weekly farmers market in the village while the study was still running.
Presenting the dissertation defense to the committee
Presenting the dissertation defense to my committee.

AI health coachPersonalization starts with learning what people are optimizing for, in their own words, before you instrument it.

Teen Lifestyle Balance program illustration

Goal: adapt an adult, evidence-based coaching program so it is clinically, socially, and developmentally appropriate for teenagers, and find the most effective way to deliver it.

Step oneFocus groups & interviewsTeenagers and their parents · coded in Dedoose & Excel
Step twoAdapt the modulesRewrite GLB sessions to be relevant and relatable for teens
Step threePilotTest the adapted sessions with pre-diabetic adolescents
Usability testingSystematically tested fitness- and food-tracking mobile apps and their web versions for day-to-day ease of use, since the program depends on self-monitoring.
Curriculum adaptationRewrote Group Lifestyle Balance modules using themes from the teen–parent sessions, the usability tests, and the literature.

AI health coachA health-coaching program with a self-monitoring loop, adapted for a population the original design ignored: what makes advice feel relevant, and how much logging people will tolerate.