Methodology
How our research works
An Ask A Day is a daily question-and-answer app for people in Ghana. Every day, players answer short questions and occasional longer surveys in exchange for points they can use in sweepstakes and our rewards shop. Their answers, combined across many people, become the insights we publish and share with partners.
Who answers
Our panel is people who use the An Ask A Day app in Ghana, currently on Android. Anyone aged 13 and over can join; players under 18 need a guardian’s approval to collect some rewards. When they join, players tell us their age, gender, region and, optionally, their town, plus their education, work and relationship status. This lets us break results down by group and send questions to the right people.
Players confirm their phone number with a text message before they can collect rewards, so each account belongs to a real Ghanaian mobile number.
How we ask
- One answer per person. Each player can answer a question once.
- A featured question every day, plus shorter side questions and longer surveys.
- Repeated questions to track change. Some questions are asked again with exactly the same wording, so we can show how opinions move over time.
- Rewards for taking part, not for the answer. Points are the same whatever someone chooses, so there’s no “right” answer to aim for.
How we check answers
Research is only useful if answers are honest. Every account has an answer quality score, based on:
- Consistency: whether answers match the player’s profile and their earlier answers to the same question.
- Attention: whether players pass checks such as “please choose Somewhat disagree”.
- Effort: whether they rush through surveys, give the same answer down a whole grid, or write very short or copied replies.
- Duplicate accounts: whether one phone is used by more than one account.
Results can leave out low-quality accounts, and our staff review an account before acting on its score.
How we make results represent Ghana
People who use apps don’t mirror the whole country: our players are younger and more likely to live in Greater Accra than Ghanaians overall. So, where we say a result is weighted, we adjust it to match Ghana’s population aged 13 and over in the 2021 Population and Housing Census (Ghana Statistical Service), by region, sex and age group. We use raking, a standard method in survey research. No single answer can count for more than five times an average one, so a few people can’t swing a result.
Alongside every weighted result we give the effective sample: roughly how many people the weighted result is worth. When weighting has to work hard, the effective sample is smaller, and we say so.
What our results can and can’t tell you
- They represent people in Ghana who use a smartphone app, adjusted to the population by region, sex and age. They don’t fully represent people without a smartphone or internet access.
- Weighting corrects for region, sex and age. It doesn’t correct for things like education or urban versus rural living, so we show those breakdowns rather than weight by them.
- Profile details are self-reported. Our consistency checks catch many errors, but not all.
- Small groups give less certain results. We show how many people answered so you can judge.