What determines the Sample Size for Cross-Sectional Surveys?
January 6, 2021Hiring Field Survey Enumerators: 5 Good Practices You Can Use
January 27, 2021Questionnaire Design: Best Practices for High-Quality Measurements
After we finalize research questions and methods to answer those, the next step is data collection. Quantitative research uses structured and standardized questionnaires to collect data. These questionnaires ensure that all respondents are asked questions in the same manner and in the same sequence, and the answers are coded as a number. Even for a multiple-choice type of a question, each option has an underlaying numeric code. And open-ended questions which collect text data are often coded into a number during data processing stage. Once finalized these structured and standardized questionnaires are administered to hundreds of respondents as per the required sample size. Therefore, it becomes important to get the design and administration modality right. Otherwise, we are stuck with a large data set you cannot use or trust.
Today, technology makes us administered well-designed questionnaires with even better accuracy, validity, and speed than what is possible with paper-based surveys. Below, I highlight a few aspects you must consider in designing a questionnaire.
Questionnaire Design Best Practices
1. Duration or Length
We researchers are greedy for more data, but more does not mean good! Nowadays researchers are moving towards lean data collection that forces them to ask how the data will be used, and question the need for extra information often collected but rarely used. We advise to keep the questionnaire length to <60 minutes in a rural setting and <30 minutes in cities.
2. Language
Write a question the way you intend it to be asked without the need for repetition or explanation in the enumerator’s own words; that beats the purpose of a standardized questionnaire. Our tip – first think of the question in the local language and then translate into English, not vice versa.
3. Questionnaire Revalidation
Some questions are just too important to get wrong. Ask such questions twice after some time gap and validate that the answers match. If they don’t, the respondent will be a lot more attentive. But, please use this tip sparingly.
4. Flow like Makkhan
Sorry for breaking into Hinglish but the questionnaire should run without hiccups and ask questions in a flow that is logical and ‘customized’ to previous answers. Administer a designed questionnaire to your non-research friend and if he yawns after 10 minutes, know you got it wrong. Fix section heading, section intro text, instructions for skipping-asking-customizing the question, question’s wording, and proper set of answers which can be entered quickly.

5. Pre-test and field tests
Always test the entire questionnaire for content and flow before proceeding with CAPI development. Rigorous research should ideally do a pre-test with target respondents, but at least trouble a few friends. A group meeting of researchers is a great way to dissect and finalize questionnaires.
6. Update and Changes to the Questionnaire
A questionnaire is never final until a few days of the fieldwork is done. So you should be prepared for such changes but if they are major, then programming of your CAPI App and field trainings can be affected by these last-minute changes. Our rule of thumb is <5% questions and <10% ‘words’ are changed during field training and initial fieldwork.
To dive deeper into the topic, watch this space for the second part of this blog: ‘Tips for Developing a CAPI App that works like a charm’
Also read “Principles for Designing Good Questionnaires”