Phone Surveys (CATI): Best Practices for Reliable Data

Surveys & Data - 2026-11-15 - ATTPL EMS

Survey interviewers with headsets working at computer stations in a Jaipur telephone survey centre

Phone surveys can reach large, spread-out populations quickly and at lower cost than face-to-face visits. With careful design and quality control, computer-assisted telephone interviewing (CATI) produces reliable data for research, programmes and public opinion studies.

When phone surveys work best

CATI works well for short surveys, follow-ups with known respondents, monitoring studies and quick feedback. It is less suitable for long, sensitive or complex questionnaires, or where many respondents lack phones. Compare it with face-to-face CAPI surveys when choosing a method.

Sampling and contact strategy

Use a clear sampling frame, such as a beneficiary list or a properly drawn phone sample
Plan for non-response and replace numbers only by documented rules
Schedule calls at times convenient to respondents, including evenings and weekends
Set a maximum number of attempts per number and record every outcome

Design a short, clear questionnaire

Keep interviews short, often within 15 minutes, use simple language and offer the local language. Begin with an introduction, the purpose of the survey and a consent script that explains how data will be used and that participation is voluntary. Our questionnaire design guide covers question wording and flow.

Training and quality control

Train interviewers on the questionnaire, probing, neutrality and handling refusals, with mock calls before launch; see enumerator training. Supervisors should listen to a share of calls or recordings (with consent), run back-checks, and watch indicators such as interview length, refusal rates and unusual answer patterns for each interviewer.

Protect respondents' data

Collect only the personal data you need, store it securely, restrict access and delete it when no longer required, in line with the Digital Personal Data Protection Act, 2023 and the client's requirements. ATTPL EMS runs survey operations including CATI and field surveys, and supports baseline and endline evaluations.

Frequently asked questions

How long should a phone survey be?

Usually short, often within about 15 minutes, to reduce drop-offs and fatigue.

Do phone survey respondents need to give consent?

Yes. Explain the purpose, how data will be used and that participation is voluntary, and record consent before starting.

How is quality checked in CATI surveys?

Through call monitoring or recordings with consent, back-checks, and monitoring interviewer-level indicators such as duration and refusal rates.

Sources: Ministry of Statistics and Programme Implementation, Ministry of Electronics and Information Technology