A survey can fail before the first interview begins if the collection method excludes the people the research is supposed to represent. In Nigeria, a telephone interview, a household visit and an online questionnaire may all ask identical questions yet reach markedly different groups. Selecting between CAPI, CATI and CAWI is therefore not just a technology choice. It is a decision about coverage, respondent experience, cost and the credibility of the conclusions.
The right method depends on who must be reached, how complex the questions are and what the data will be used to decide. Understanding the trade-offs is the first step towards a sound survey design.
What do CAPI, CATI and CAWI mean?
Computer-assisted personal interviewing (CAPI) is interviewer-administered, usually face to face, using a phone, tablet or other digital device to capture answers. Computer-assisted telephone interviewing (CATI) is interviewer-administered by phone, with answers entered into a digital questionnaire. Computer-assisted web interviewing (CAWI) is self-administered: respondents complete a questionnaire through a website or survey link.
All three can support automated skips, routing, question validation and digital data storage. Their fundamental difference is the interaction between respondent, interviewer and recruitment channel. That difference affects not only speed but also who participates and how they answer.
CAPI: strongest where physical reach matters
CAPI is useful for household studies, community assessments, retail audits and surveys that require physical observation. Interviewers can locate selected dwellings, explain unfamiliar concepts and sometimes verify conditions such as water sources, product availability or facility equipment. Devices can collect geographic coordinates when justified and permitted, while offline forms can support work where connectivity is intermittent.
These advantages come with operational costs. Teams must travel, recruit locally, secure devices, supervise field staff and accommodate call-backs. Interviewer presence can also create social desirability bias, particularly on questions about income, political attitudes, services received or other sensitive matters. An impressive CAPI dashboard offers little protection if interviewers select convenient households instead of sampled ones.
CAPI often suits NGO baselines and population studies requiring a defined geographic sample. It may be excessive when the target population consists of identifiable professionals who are widely dispersed and easy to contact by telephone.
CATI: efficient when reliable contact details exist
CATI can be valuable for business-to-business interviews, customer experience studies, beneficiary follow-ups and rapid monitoring when a usable phone frame exists. Interviewers can clarify questions, make scheduled call-backs and apply centralised quality supervision. There is no need to deploy a field team to every state simply to ask a short set of questions.
However, possession of a phone number does not guarantee eligibility, availability or willingness to participate. Incorrect numbers, shared phones, call screening and repeated unanswered calls can reduce effective coverage. Telephone interviews are also difficult for very long questionnaires, visual product comparisons or tasks that require observation. Poorly designed call scripts can feel intrusive and lead to early termination.
For CATI, track contact, eligibility, refusal and completion outcomes separately. A completed interview count is not enough to assess the quality of the sample. Attempt protocols and a genuine respondent-verification process are as important as the software.
CAWI: convenient but vulnerable to selection bias
CAWI works well when the study population can be reached online and is comfortable completing forms independently. Examples include an existing customer panel, a verified professional association or participants recruited from a known email list. Respondents can answer at a convenient time, while visual prompts, branching and multilingual interfaces can be programmed efficiently.
The danger is assuming that a survey link circulated through WhatsApp, social media or email represents the whole Nigerian market. Respondents may be disproportionately connected, literate, younger or more motivated by the subject. Duplicate entries, fraudulent incentives and unknown exposure to the invitation create further challenges.
A CAWI sample can still be rigorous when recruitment follows a credible sampling frame and participation is carefully controlled. Conversely, a large opt-in sample remains non-probability research unless the design provides a defensible basis for population inference. Volume does not automatically remove selection bias.
Compare the methods against research objectives
For geographic household coverage, CAPI commonly has the strongest practical fit, provided the listing and selection procedures are sound. For a well-maintained database of customers or service beneficiaries, CATI can reduce travel requirements while preserving interviewer support. For a defined digital audience or concept test with images, CAWI may provide speed and a better visual experience.
There is no universally cheapest option. CAWI generally has low cost per additional completed form once respondents can be recruited, but obtaining a representative online sample may be expensive. CATI saves travel costs but can require many attempts per completed interview. CAPI incurs mobilisation costs yet may offer better access to households that phone and web frames miss. Compare cost per *valid eligible completion*, not simply cost per click, call or visit.
Consider a mixed-mode design
A mixed-mode survey may combine methods to improve accessibility. For example, an NGO could interview a probability sample of households with CAPI and subsequently conduct short CATI follow-ups with consenting participants. A company could begin with CAWI among verified customers, then telephone groups with low response rates. The objective should be better coverage or lower burden, not simply adding modes for their own sake.
Mixed-mode designs create new questions. The same item may produce different responses when answered privately online versus aloud to an interviewer. Where comparisons matter, keep wording, response options and reference periods consistent, test for mode effects and document recruitment by mode. Never merge convenience web responses with a probability household sample and report one supposedly representative national percentage without a justified analytical approach.
Build data quality into each method
For CAPI, use field supervision, selective GPS review, interview-duration checks and back-checks with appropriate consent. For CATI, monitor call dispositions, recording permissions, interviewer performance and refusal patterns. For CAWI, consider unique links, bot and duplicate safeguards, device compatibility and attention checks that do not unfairly exclude genuine respondents.
Across all modes, pilot the questionnaire, train personnel, document the sampling process and protect personal information. Disclosure should include the target population, sampling frame, field dates, languages, mode, response outcomes, weighting and limitations. A transparent methodology is more persuasive than a claim of '100% accurate data'.
A practical decision test for Nigeria
Ask five questions before selecting the platform. First, do we have a defensible way to identify and contact eligible respondents? Second, will the method systematically exclude important groups? Third, must an interviewer clarify complex questions or show physical materials? Fourth, is the expected interview duration realistic? Fifth, can we quality-check the collected information within the budget?
Suppose a study explores consumer attitudes towards a new packaged drink in Lagos and Kano. CAWI may be appropriate for an initial digital concept test, but it cannot alone confirm product availability in informal outlets. Retail observation or in-person intercepts could add that evidence. A separate CATI distributor survey might clarify ordering patterns. Each method contributes a different piece of the decision, and each has its own limits.
Conclusion
The best survey mode is the one that matches the population, question and inference required. CAPI offers field access and observation; CATI offers interviewer-led reach through usable phone contacts; CAWI offers self-completion and efficient digital engagement. Good Nigerian research begins by confronting the biases and operational demands of each choice, then designing quality assurance around them.
Surveysphere Africa supports CAPI, CATI, CAWI and mixed-mode survey research, from sampling and questionnaire development to field supervision and analysis. Discuss a study with us at info@surveysphereafrica.com or surveysphereafrica.com.
Related guidance
For the continent-wide view, read our guide to CAPI, CATI and CAWI.
Our services for research in Nigeria
We run research across African markets. Start with these services for work in Nigeria:
- Outsourced fieldwork in Nigeria: CAPI, CATI and CAWI data collection with live quality control.
- Market and consumer research in Nigeria: Market-entry, consumer and retail research.
We also offer: Monitoring, evaluation and learning, Social and development research, Public health research, Policy, economic and institutional research, Data analytics and insights visualisation.
