Computer-assisted personal interviewing (CAPI) is a survey method where an interviewer sits down with someone face-to-face and reads questions off a laptop, tablet, or phone, typing the answers straight into the device. No paper, no separate data-entry step later.
The short version:
- CAPI is face-to-face — an interviewer is physically present, unlike phone or web surveys.
- The questionnaire is software — it controls question order, hides irrelevant questions, and checks answers as they're entered.
- It's the standard for big government surveys — national censuses, labor surveys, and large household studies run on CAPI.
- It replaced paper — the older method, paper-and-pencil interviewing (PAPI), is mostly gone in well-funded surveys.
If you've ever been stopped by a census worker holding a tablet, or sat through a household survey where the interviewer tapped through screens, you've seen CAPI in action. Below is what it is, how it works, where it's used, and how it stacks up against the other ways surveys get done.
What CAPI actually means
CAPI stands for computer-assisted personal interviewing. Break the name down and it explains itself:
- Computer-assisted — the questions live in software, not on a printed form.
- Personal — it happens in person, face-to-face.
- Interviewing — a trained interviewer asks the questions and records the responses.
As Wikipedia's survey-methods entry puts it, CAPI is data collection where an in-person interviewer uses a computer to administer the questionnaire and capture answers directly onto the device. That last part is the whole point: the answer goes into a database the moment it's spoken. There's no stack of paper forms to key in afterward.
CAPI is one branch of a broader family called computer-assisted survey information collection (CASIC). Its close relatives differ mainly by *who* operates the device and *where* the interview happens:
- CAPI — interviewer present, in person.
- CATI — interviewer present, over the phone.
- CAWI — no interviewer; the respondent fills it out online.
- CASI — interviewer hands the device to the respondent to answer sensitive questions privately.
That last one, computer-assisted self-interviewing, often runs *inside* a CAPI session. When a question is sensitive — drug use, income, sexual behavior — the interviewer passes the tablet over so the respondent can answer without saying the words out loud, then takes it back to continue.
How a CAPI interview works
From the respondent's side, a CAPI interview looks like a normal conversation. Under the hood, the software is doing a lot of work that paper never could.
The questionnaire is a program. The interviewer opens a survey application and the screen shows one question at a time. As answers go in, the program decides what comes next.
Three features do most of the heavy lifting:
- Skip logic (routing). If someone says they don't own a car, the survey skips all the car questions automatically. On paper, the interviewer had to read "if no, go to question 14" and hope they didn't misread it. CAPI removes that error entirely — respondents only see questions that apply to them.
- Real-time validation. Type an age of 200 or a date that hasn't happened yet, and the software flags it on the spot. According to the World Bank's Development Impact Evaluation team, this kind of built-in checking — logic checks, range limits, consistency rules — is one of CAPI's biggest quality advantages over paper.
- Direct data capture. The response is recorded electronically as it's given. No transcription, no second person re-keying paper forms, no transcription typos.
Modern CAPI tools layer on more: GPS coordinates to confirm the interview happened where it was supposed to, audio audits that record snippets of the conversation for quality control, timestamps on each question, and the ability to attach photos or show images and video as part of a question. Because the data syncs to a central server, supervisors can spot a struggling interviewer or a bad question while fieldwork is still running, instead of finding out months later.
The hardware is whatever's practical: laptops were the original choice in the 1980s and 1990s, but tablets and smartphones now dominate because they're cheap, light, and last all day on battery.
Where CAPI is used
CAPI shows up anywhere a survey is long, complex, or too important to risk paper errors.
National statistics. Government statistical agencies are the biggest users. The U.S. Census Bureau runs major surveys this way — its field representatives use laptops and tablets to conduct studies like the Current Population Survey, the source of the official monthly unemployment rate. The Bureau even builds its own free CAPI tool, CSPro / CSEntry, used for censuses and surveys around the world.
International development research. When the World Bank, UN agencies, or research teams run household surveys in dozens of countries, CAPI keeps the data consistent and lets them monitor remote field teams in near real time.
Market and social research. Consumer studies, public opinion polls, and academic behavioral research use CAPI when they need the depth of a face-to-face conversation plus clean, structured data. It's common in shopping-mall intercepts and in-home interviews where a survey runs 45 minutes or longer — too long to hold someone's attention over the phone or online. For researchers, the structured questionnaire is only half the job; the other half is recruiting the right respondents in the first place, which is closer to the relationship-management and outreach problem than to survey design.
The common thread: CAPI is chosen when the interview is complex enough to need a human guide but structured enough to run as software.
CAPI vs. CATI vs. CAWI vs. paper
Each survey mode trades off cost, reach, and data quality differently. Here's how the four main approaches compare.
| Factor | CAPI (in person) | CATI (phone) | CAWI (web) | Paper (PAPI) |
|---|---|---|---|---|
| Interviewer present | Yes, face-to-face | Yes, by phone | No | Yes, face-to-face |
| Cost per response | High | Medium | Low | Medium |
| Data entry | Automatic | Automatic | Automatic | Manual, later |
| Skip logic & validation | Built in | Built in | Built in | Manual, error-prone |
| Good for long/complex surveys | Excellent | Limited | Limited | Workable but slow |
| Reaches people without internet/phone | Yes | No | No | Yes |
| Handles sensitive questions | Yes (via CASI) | Weak | Good (private) | Weak |
| Speed to clean data | Fast | Fast | Fastest | Slow |
A few takeaways:
- CAPI and paper are the only modes that reach people without phones or internet — which is exactly why they dominate in rural and developing-country fieldwork.
- CAWI is the cheapest and fastest but you lose the interviewer, so it suits short, simple surveys with a connected audience.
- CATI sits in the middle — cheaper than CAPI, but it struggles with long questionnaires and anything that needs visual aids.
- Paper is the baseline everyone moved away from. It's still used where devices aren't practical, but the manual data-entry step and lack of validation make it the slowest and most error-prone option.
CAPI is the most expensive mode, full stop. You're paying for interviewers, travel, and devices. Organizations choose it anyway when data quality and reach matter more than cost.
Common CAPI software
You don't build a CAPI survey from scratch — you use a platform that lets you design the questionnaire, push it to field devices, and collect the results. The widely used tools include:
- Survey Solutions — the World Bank's free CAPI platform, popular for large international surveys.
- CSPro / CSEntry — the U.S. Census Bureau's free tool for census and survey data entry on Windows and Android.
- SurveyCTO and ODK (Open Data Kit) — flexible mobile data-collection tools widely used in research and development work.
- Blaise — a long-running survey system from Statistics Netherlands, used by several national statistical agencies for their flagship surveys.
The World Bank publishes a comparative assessment of these programs if you need to pick one for a real project — it weighs cost, offline support, question types, and quality-control features.
FAQ
What does CAPI stand for?
CAPI stands for computer-assisted personal interviewing. It's a survey method where an interviewer meets a respondent in person and uses a laptop, tablet, or phone to ask the questions and record the answers directly into software.
What's the difference between CAPI and CATI?
Both use software to run the questionnaire and capture responses automatically. The difference is the channel: CAPI is conducted face-to-face in person, while CATI (computer-assisted telephone interviewing) is conducted over the phone. CAPI is usually preferred for long or complex surveys and for reaching people without reliable phone or internet access.
Is CAPI better than paper surveys?
For most large surveys, yes. CAPI eliminates the separate data-entry step, applies skip logic and validation automatically, and lets supervisors monitor data quality during fieldwork. Paper (PAPI) is still used where devices aren't practical, but it's slower and more prone to entry errors.
What software is used for CAPI?
Common platforms include the World Bank's Survey Solutions, the U.S. Census Bureau's CSPro/CSEntry, SurveyCTO, ODK (Open Data Kit), and Blaise. They let researchers design questionnaires, deploy them to field devices, and collect results centrally.
Who uses CAPI?
National statistical agencies (like the U.S. Census Bureau), international organizations (the World Bank, UN agencies), and market and academic researchers. It's the standard for censuses, labor-force surveys, and large household studies.
The bottom line
CAPI is the workhorse of serious face-to-face survey research: an interviewer, a device, and software that keeps the data clean from the first question. It costs more than phone or web surveys, but it buys you depth, reach, and quality that the cheaper modes can't match — which is why censuses and major household studies still run on it.
If your work involves research or outreach where the hard part isn't the survey tool but finding the right people to talk to in the first place, that's a different problem. Articuler uses semantic matching across 980M+ professional profiles to surface the specific people who fit what you're researching — then helps you find the right people and reach out without the guesswork. It's the same intent-based approach behind the best AI networking apps for anyone who needs fewer, better conversations.