Polls and public opinion reports should explicitly label and treat open‑ended (qualitative) replies as illustrative unless the authors demonstrate representativeness; when publishing qualitative findings, include codebook, translator/vendor, interrater agreement, and a statement on which subset of respondents produced the text. This norm would reduce the common misreading of quoted open responses as population estimates, especially on polarizing issues.
— Creating a standard practice to flag and document nonrepresentative open responses would reduce misinterpretation of surveys on sensitive political topics and improve public trust in survey-based reporting.
Janakee Chavda
2026.09.15
100% relevant
Pew’s appendix notes that 1,585 open responses were translated by cApStAn, coded by three researchers with double‑coding and reconciliation, and explicitly presented as illustrative quotes rather than point estimates, showing the exact elements that a standard should capture.
← Back to all ideas