Formal Rebuttal Statement
Statement on the “Pre-Election Survey” Publicised as a Neutral Assessment
We have reviewed the publicly circulated “pre-election survey” attributed to a joint initiative involving Projection BD, Jagoron Foundation, Narrative, and the International Institute of Law and Diplomacy (IILD), reportedly fielded from 21 November to 20 December 2025 with 22,174 registered voters across 64 districts / 295 constituencies.
While opinion research can play a constructive role in a democracy, this survey—as presented to the public— does not meet basic transparency and disclosure norms required for credible public polling. The release emphasizes headline vote-share figures (e.g., BNP 34.7%, Jamaat 33.6%, NCP 7.1%, Islamic Andolon 3.1%, Others 4.5%, Undecided 17%) without providing the methodological disclosures necessary for independent verification.
Our concerns are fourfold:
1. Insufficient methodological disclosure
The public material does not provide the minimum information needed to assess scientific validity—such as sampling design, selection procedures, mode of interviewing, weighting, response rates, interviewer controls, margin of error, and the full questionnaire.
2. Narrative framing that risks misleading the public
The way results are framed—particularly the “neck-and-neck” emphasis—can influence perceptions and voter behaviour. Without full transparency, such framing risks becoming a political communication instrument, not a neutral measurement.
3. No verifiable quality safeguards
Claims of very large sample sizes do not substitute for proof of representativeness and integrity. Large numbers are meaningless without documented sampling, weighting, and quality control.
4. Public-interest harm from opaque polling
In a pre-election environment, opaque surveys can erode trust, polarise society, and distort democratic choice.
Accordingly, we call on the organizers to publish—immediately and in full—standard methodological documentation consistent with international disclosure norms, including: sample frame, selection method, interview mode, weighting scheme, regional/urban-rural quotas (if any), response rate, margin of sampling error (and design effect), questionnaire wording and order, fieldwork protocols, and anonymized toplines/crosstabs.
Until these disclosures are made, the survey should not be treated as reliable evidence of public opinion and should not be amplified as authoritative forecasting.
Comparison against international polling standards (ESOMAR, AAPOR)
Below is a practical disclosure checklist derived from widely used international norms (ESOMAR/WAPOR transparency principles and AAPOR disclosure expectations). This is a standards-based comparison of what credible polls should disclose versus what is currently publicly available.
A. Transparency: what must be disclosed for public credibility
International expectation (ESOMAR / AAPOR):
- Funding/sponsorship and conflicts of interest
- Fieldwork dates
- Target population and sampling frame
- Sampling method
- Mode of interviewing and language(s)
- Sample size and geographic allocation
- Weighting and design effect
- Response and cooperation rates
- Margin of error or limits of inference
- Full questionnaire and order
- Quality control procedures
- Data access and documentation
What is publicly shown:
- Fieldwork dates and claimed sample size/coverage are stated
- Headline vote shares and undecided voters are stated
- Critical methodological disclosures are absent
Standards implication: Publishing toplines without these disclosures is insufficient for responsible public polling.
B. Representativeness & inference
Without disclosure of selection and weighting methods, national representativeness cannot be substantiated, turning polling into persuasion rather than measurement.
C. Questionnaire effects & neutrality
Absence of full question wording and undecided-handling rules prevents evaluation of bias or priming.
D. Quality control & data integrity
No information is provided on interviewer training, supervision, verification, or data-cleaning safeguards.
Minimum disclosure package required
- Sponsor/funder and governance
- Sampling frame and selection method
- Mode, language, and interviewer procedures
- Weighting methodology
- Response rates
- Margin of error or inference limits
- Full questionnaire
- Undecided handling rules
- Quality control documentation
- Toplines, crosstabs, and anonymised microdata
By Minhaz Samad Chowdhury
Independent Human Rights Defender and Policy Advocate
Focus Areas: State Violence • Media Freedom • Political Rights • Minority Protection • Democratic Accountability
No comments:
Post a Comment
Please validate CAPTCHA