Without exaggeration, I have surveyed tens of thousands of hard-to-get-to people, and I have had fair success.
Here is my tried and true workflow.
- Write a research statement to establish what you need to learn.
- Develop a hypothesis on what you think you might learn. Example: Shareholders don’t read management biographies.
- Identify the target recipients.
- Decide on the survey format that will best engage your taregets — online, phone, mail, in-person, etc.
- Plan the questions and their flow.
- Consider question order and possible order bias.
- Diagram questions in a flow chart to see the logic.
- Keep like questions together.
- Start with non-threatening items.
- Don’t make your questionnaire too long.
- Start with general questions and then go to specifics.
- Reverse scale order in some items to eliminate habitual responses.
- Ask demographic questions last. They can be considered threatening.
- Write the non-threatening questions.
- Use closed-ended responses (Yes/No or 1, 2 or 3) for responses that tabulate quickly.
- Use open-ended responses to get a range of data and bring up ideas you had overlooked.
- Make questions and responses specific.
- Use terms everyone can understand; avoid jargon.
- Write the threatening (embarrassing) questions.
- Create an environment that lets people comfortably respond.
- Prefer open-ended items that let people decide how to respond.
- Include necessary qualifiers and context in questions.
- Phrase your question in terms of “most people you know.” People are usually more willing to talk about others than themselves.
- Ask about past behavior before you ask about present behavior.
- Write questions that test knowledge
- Don’t make the questionnaire seem like a test.
- Simplify questions and answers.
- Leave questions with numerical answers open-ended.
- If you ask yes/no questions, use related questions later to double-check responses.
- Write opinion questions.
- Be very specific.
- Use closed-ended responses.
- Gauge the strength of responses by providing a response scale aka “1-10… how important is this issue to you?”
- Start with general questions and then move to specific
- Group questions with the same underlying values.
- Write questions so answers will be easy to tabulate.
- Format the questionnaire.
- Test the survey with people who know about the study. Adjust.
- Conduct a pilot test with a small number of people from the target. Adjust.
- Launch the survey to the complete target.
There are other tactics, of course. My results are only worth mentioning as I gave away NO participation "bribes or rewards" and many on the targets were just "names on a list." Yeah. I spammed them.
Carefully.