Media, Newsrooms & Opinion Polling
Methodology Disclosure Sheet
A one-page, plain-language statement of how a poll was run, ready to publish with the story.
- Typical length:
- 4 pages
- Format:
- Live dashboard and PDF
- Channels:
- SMS, USSD, WhatsApp, Web panel
Sample size
1,500
Adults 18 and over
Margin of error
2.5 pts
At 95% confidence
Feature-phone share
51%
SMS and USSD interviews
The evidence
- WhatsApp31%
- Web panel18%
- SMS24%
- USSD27%
What this tells you
- 1
Half the sample is on feature phones
SMS (24%) and USSD (27%) together account for 51% of interviews, keeping rural and older adults in the result.
- 2
Weighting changes little
Raw and weighted toplines differ by under 2 points, which suggests quotas did most of the work.
- 3
Subgroup error is larger
Groups of 200 to 400 respondents carry margins of 5 to 7 points, and the sheet says so explicitly.
Decisions this supports
- Whether the poll meets the newsroom's publication standard
- What caveat wording to print under charts
- Whether subgroup claims are safe to make
Method and confidence
Surveyed: 1,500 adults, 31% WhatsApp, 18% web panel, 24% SMS and 27% USSD, fielded over 60 hours, weighted to census on age, gender, region and urban or rural, margin of error 2.5 points at 95%. No observed social data used.
Closer labels every figure as surveyed or observed, with sample size and margin of error where they apply. All numbers on this page are invented to show the format and do not describe real Kenyan opinion, markets or organisations.