Approach

Mixed methods, defensibly designed.

Every Evident study pairs quantitative breadth with qualitative depth. We measure what is measurable and surface what numbers alone cannot — the behaviours, norms and frictions that shape outcomes.

Quantitative

Structured surveys at scale

Statistically powered samples with stratification by district, gender and socio-economic strata
Difference-in-Differences, Propensity Score Matching, RDD where designs allow
Disaggregated regression analysis with margins of error and significance reporting
Indicator dashboards built in R, Stata and Python for client handover

Qualitative

Depth where it matters

Focus Group Discussions with primary, influencer and gatekeeper cohorts
Key Informant Interviews across the implementation chain — frontline to district
Structured observation studies — classrooms, health facilities, BC outlets, ration shops
Thematic coding in NVivo / Dedoose with triangulation against survey findings

Field technology

Mobile-first data collection — auditable from response to report.

CAPI on SurveyCTO / KoboToolbox / ODK

Offline-first forms with skip logic, constraints, and built-in validations to reduce data-entry error at source.

GPS-tagged household interviews

Every interview geo-fenced to the assigned PSU; route maps reviewed daily by field supervisors.

Audio audits & call-back back-checks

Random audio-recorded sections and 15% telephonic back-checks on completed interviews.

Data quality & ethics

Defensible from sampling frame to final report.

Informed consent

Plain-language consent in the respondent's preferred language; opt-out preserved at every stage.

Back-checks & spot audits

15% back-checks plus supervisor spot visits to randomly-selected interviews each day.

Data privacy & storage

PII separated at ingestion; encrypted-at-rest on Indian servers; access logs retained for 24 months.

Independent ethics review

Sensitive studies (adolescents, GBV, health) cleared by external IRB before fieldwork commences.