Covid-19 and NGOs – Impacts Now and Later
November 26, 2020Programming and Monitoring Systems to Coordinate Public and Private Action to Achieve Sustainable Development Goals
December 9, 2020Questionnaire Design: 4 Principles for Preparing Good Questionnaires
Well-designed questionnaires are a critical component of high-quality and “useful” research. Your analysis and findings are only as robust as your data is relevant, valid, and credible. And collecting such data is nearly impossible without a well-designed data collection instrument – the questionnaire! (and of course, high-quality data collection).
While some general guidelines exist on the design of a questionnaire, only through application and experience this art is learned. That is why field managers and enumerators often have many useful suggestions and deeper insights on how to better collect the same data. In this blog, I share four principles I learned while conducting interviews myself and from my amazingly experienced field team.
1. Make your Questionnaire conversational
The questionnaire should flow like a conversation because it helps the interviewer to build a rapport and gain cooperation. All the questions should flow in a logical order so that one leads to the next naturally.
- Open with simple and interesting questions to engage the respondent. Don’t let them lose interest or be defensive.
- Follow by the most important questions related to “key variables” you will use in the analysis. Respondents can remain attentive for a maximum of 45-60 minutes and then get impatient. Respondents also find it distracting to juggle between topics so sequence and group questions accordingly.
- Towards the end, include “action” questions that require the respondent to walk around or demonstrate something. Mix-up type of questions which could shake them up from boredom and routine that sets in after 60 minutes of talking. Ask administrative questions and basic factual questions you can run through towards the end. Perhaps, you may also want to schedule “sensitive” questions towards the end to avoid respondents backing out in the middle.
2. Don’t be greedy; Keep the Questionnaire “Relevant”
Include only those questions which are relevant to the research topic and your analysis plan. As researchers, we are sometimes tempted to get “extra” information from the survey. But ask, “are these questions really needed at the cost of quality of other data?”
We also tend to over-explain the questions which actually confuses the respondents and enumerators both! I often see multiple-choice questions with 25 options listed. Respondents rarely give an answer that fits nicely into one of the 25 options, there is always overlap. This confuses the enumerator more and she ends up coding options she feels are appropriate. Instead, do a good pre-test and figure out that there are <10 “mutually exclusive” answers or option codes. You can always have an open-ended “Specify” option, but if your data has more than 5% of these, then know your design and pre-test wasn’t really that good.
My own lesson: It is best to ask only relevant data and ask it really well and in detail. Quality over quantity!
3. Think from the respondent’s perspective
It is always better to think in the local context and language while developing questionnaires. We tend to develop the questionnaire in English and then translate it into a local language, which sometimes results in accurate but complex and subtly different meaning questions. At NEERMAN, we overcome this issue by getting field managers and executives from relevant local areas and backgrounds involved in the questionnaire design stage itself. Their practical, no-nonsense review has made my questionnaires so much better.
4. Paper is dead! Use Technology
I almost certainly won’t go back to paper-based surveys especially because Computer Assisted Personal Interviews (CAPI) Apps are freely available. CAPI makes administration flow smoothly with the least logical errors. It also helps us design the “appearance” and “content” of the questions better, which helps both the enumerator and the respondent. Keep the following tips in mind:
- Use bold, italics, and colour coding to differentiate/highlight keywords, hints, instructions, etc. better
- Customize question-wording by inserting the answer to a previous question, name of the respondent (easily possible in most free CAPI platforms such as ODK).
- Always give a brief introduction to a new topic or line of questioning.
- Use tables and matrices to administer similar questions which helps enumerators move a lot faster and without errors
Display only those options or allow only those answers which are logically feasible based on previous answers. And when there is an error, give a clear error message to the enumerator to fix any logical fallacy.
I hope you will remind yourselves of these four principles the next time to undertake a questionnaire development! Check out our Resources page for information that can help you improve your research design and implementation. You can also schedule a free micro-consulting on our website to get answers to your pressing questions on questionnaires.