This document describes best practices for drafting the introduction to a questionnaire. There isn’t one solution that is perfect for all scenarios. However, there are best practices and some common patterns to remember. The introduction is critical for achieving a successful response rate: people respond positively to brief introductions and make immediate judgments about whether the survey is trustworthy and important. In fact, the majority of people who break off a survey stop at the survey introduction step. Once people agree to participate in a survey, they are much more likely to complete it. Therefore, it is critical to write a compelling and brief introduction for SMS and phone call surveys.
Key dimensions of an introduction for a questionnaire include the following components:
Greeting: Where the survey identifies what it is, who is sending it, and context describing why it is important for an individual to participate.
Language selector: Allows you to self-select your preferred language for the questionnaire.
Consent: Where the questionnaire informs you of the conditions around a survey and allows them to opt in or out of the survey.
Incentive: Consider including incentives in the introduction. In Ask, this is the only place where respondents can be informed of any survey token and the amount.
Do Balance the length and content of the greeting.
The greeting is critical. It needs to identify who is sending out the survey and encourage participation. A long or confusing greeting will result in decreased participation, but writing something concise and engaging will encourage participants to stay involved. The ideal length of the greeting is under 8 seconds for a phone call, and within 1-2 messages for an SMS version.
Do test your introduction with a small group of people.
Before main data collection, test the introduction with a small group of people to adjust to social and cultural expectations and reactions to the survey. No one approach is perfect for all situations, but by testing early and often, costly mistakes and issues that depress response rate can easily be resolved.
Do use the lingua franca for greetings in multilingual surveys.
In multilingual countries, it can be difficult to convey the greeting in each language. For example, people may quickly lose interest in a survey when greeted in another language. In these situations, it is recommended to start with a lingua franca - i.e. a language that is widely understood by a broad section of the population. Examples of lingua francas are English in Zambia and Swahili in Kenya. Using a lingua franca often mirrors the usual style of government communications, which can build trust and encourage participation.
Do Provide a language selector; have it presented in all languages.
For each language that is offered, use the related language directly in the step or language selection questions. This comes after the lingua franca introduction. A successful language selector step that offered in English, French, and Arabic (via SMS) would read as:
“For English, press 1; Pour le François, appuyez sur 2; بالنسبة للغة العربية، اضغط 3”
This allows individuals who do not read the other languages offered or specifically the lingua franca, to find an option associated with a language they are literate in. For IVR, an individual would hear the script in the specific dialect they understand.
Do separate and sequence the language selector before consent.
Consent to participate can be technical in nature. For example, the institutional review board may have some specific language that must be included. Allowing you to select your language before hearing the consent question can help to improve the delivery and comprehension of the consent question.
Don't Use the same script for both IVR and SMS introductions.Using identical scripts for an introduction in IVR and SMS fails to take advantage of the strengths of each mode of communication. For example, IVR allows for longer, more fluid introductions, whereas SMS can take advantage of abbreviations. It is helpful to design the right introduction for each mode from their perspective of the respondent.