Skip to content

guide

Public participation questionnaire: sample questions for Kenya

A questionnaire shows what people who never attend a forum think, and whether those who took part felt heard. Sample questions, and how to use them.

Takwimu editorial team, Editorial team5 min read

A public participation questionnaire does two jobs that a forum can't. It reaches people who never attend, and it tests whether those who did attend felt heard. Attendance at a forum is self-selected, so its views may be far from typical. A survey with a proper sample shows what the ward or county as a whole thinks, which supports the Supreme Court's call for participation with both quantity and quality.

When should you use a questionnaire?

Use one to learn what people want before you draft a plan, to test views on a proposal, or to check afterwards whether people knew about the exercise. It adds to the forum, and doesn't replace it, because a survey asks your questions, while a forum lets people raise their own.

Which questions should you ask?

Keep it short, and use plain words. These sample questions work for most exercises.

On awareness:

  • Before today, had you heard about the plan to CHANGE?
  • How did you hear about it?
  • Did you know that residents could give their views?

On the issue:

  • Which of these matters most to your household: LIST OF OPTIONS?
  • How much do you support or oppose the proposal? Answer on a five-point scale.
  • What is the main reason for your answer? Let people use their own words.

On access:

  • Did you take part in a forum or send in your views? If not, what stopped you?
  • Which way would you prefer to give views in future?

On feedback:

  • Were you told what happened to the views people gave?
  • Do you believe the county considered residents' views when it decided?

On the respondent:

  • Ask for sex, age group, ward, main occupation and whether the person lives with a disability.

How should you word them?

Ask one thing at a time. Avoid leading words, such as "Do you agree that the excellent new plan...?" Offer "don't know", so nobody has to guess. Translate each question, and test the translation with real people, as checking a translated questionnaire explains.

How many people should you ask?

It depends on the smallest area you want to report on. A simple random sample of about 400 gives a margin of error of about five percentage points at 95% confidence. If you need a figure for each of several wards, each ward needs its own sample. A small sample can still help, but state its size, and don't present it as the whole county.

How do you choose who to ask?

Draw a sample designed to represent the area. Interviews in the market on a Tuesday only tell you about people in the market on a Tuesday. Good surveys pick places and households by a set method, then use weights to reflect the population. If you use an open web link, say so plainly, and don't describe the results as representative.

How should you collect the answers?

Use trained interviewers for households, a web link for those online, and SMS or USSD for very short polls. Each channel reaches different people, so record which channel each answer came from. Tell people who is asking, why, and how the data will be used, and get their consent. The study belongs to the body running it, which answers for the data.

How should you use the results?

Report them beside what you heard at forums, and be clear about the difference. A forum tells you what people who came said. A survey estimates what the population thinks. Where the two disagree, say so, because that is often the most useful finding.

Publish the questions, the sample size, the dates and the margin of error. Readers can then judge the results for themselves.

What should you avoid?

Avoid long forms, because people stop answering. Avoid asking about matters you can't change, which raises false hopes. Avoid asking people to name those who broke rules, since it can put them at risk.

How does it support your record?

A survey can show that people knew about an exercise, and that they felt heard. The same two questions, asked each year, give a county a simple measure of whether its participation is improving. Pair the results with who took part, and you can show reach as well as attendance.

How do you test it first?

Try it on about 20 people, like those in your sample but outside it. Watch where they hesitate. Ask what they understood by each question, in their own words, because the answer often surprises you. Time the interview, and cut anything that isn't needed. A survey over 15 minutes tends to lose people, and to make interviewers rush.

Test the translation as well. A word such as "participation" or "consulted" may have no easy match, so choose words people actually use. Then freeze the questions, so that every interview asks the same thing. If you change a question later, you can't compare the answers.

Where TI fits

Takwimu Intelligence (TI) comes with consultation surveys in English and Kiswahili that your team can edit, with weighting and margins of error shown, so the data stands up.

To see it on your own consultation, book a walk-through.

Takwimu Intelligence (TI) is a public participation system for Kenya.

Sources

  1. 1

    British American Tobacco Kenya, PLC v Cabinet Secretary for the Ministry of Health & 2 others, Petition No. 5 of 2017, Supreme Court, judgment of 26 November 2019 · Supreme Court of Kenya63 pages. Paragraphs 89, 95 and 96 (the guiding principles for public participation) read. Paragraph 95 quotes paragraph 75 of the High Court's judgment in Robert N. Gakuru & others v Governor Kiambu County & 3 others (Petition No. 532 of 2013, consolidated) [2014] eKLR, which is how Gakuru is cited here: Kenya Law's copy refuses scripted requests.

More from the library