Research is only as good as the trust behind it. This page states plainly how we gather, sample, and check the data that appears in our Election Data Explorer, briefs, and reports — and the rules our public-facing data follows regardless of who the client is. If a figure can't clear this bar, it doesn't get published.
1. Sourcing
Our public election figures are structured to mirror official Election Commission of India constituency returns — registered electors, turnout, and victory margin. Where we publish demonstration data (as in the current Explorer), it is clearly labelled as illustrative and synthetic. Production dashboards for clients are built against primary sources: official returns, government gazette notifications, and verified administrative records, cited at the figure level.
- Primary first. We prefer the original record over any secondary aggregation.
- One hop, documented. Where a secondary source is unavoidable, we record the chain back to the primary.
- Dated at ingest. Every dataset carries the date it was pulled and the release it reflects.
2. Sampling
Field research — surveys, booth-level sampling, and qualitative work — follows a documented sampling frame before any fieldwork begins. We define the population, the stratification (constituency, booth, demography), the target sample size for a stated confidence level, and the handling of non-response. We do not report a headline number without the sample behind it.
3. Verification
Before anything is published, it passes a three-stage check:
- Reconciliation. Totals are reconciled against the source of record; parts must sum to the whole.
- Second-analyst review. A second person re-derives key figures independently before release.
- Plausibility & outliers. Anomalies are investigated, not smoothed. An unexplained outlier is a reason to hold, not to publish.
4. Neutrality
Our public data tools are deliberately non-partisan. They show turnout, margins, and elector counts — never party names, colours, symbols, or claims — because the firm serves clients across the political spectrum. Editorial interpretation lives in bylined insights and reports, clearly separated from the raw figures in the charts. The charts state what happened; the writing argues what it means.
5. Confidentiality
Client-specific data is never published. Case studies and public figures are generalised or aggregated so that no confidential result, list, or model is exposed. What happens in the war room stays there; what we publish is either public-source or explicitly illustrative.
6. Updates & corrections
Datasets carry a visible last-updated stamp, and the underlying files are versioned. If we get something wrong, we correct it promptly and note the change rather than quietly editing. To flag an error or request a source, write to contact@yctcindia.com.
This methodology governs the Election Data Explorer, our reports, and the data cited in insights.