Surveys are powerful tools for gathering insights and collecting customer feedback. Many researchers look to develop, program, and run their own surveys, but not all surveys are created equal. A well-designed survey not only gathers relevant data but also keeps respondents engaged and ensures the reliability of the information collected. Based on my experience, I’ve developed a five-step process to prepare surveys for programming.
1. Define Clear Research Objectives
The foundation of any successful survey always comes back to the ‘why.’ Clearly define your survey’s purpose and business objectives and decisions you plan to make from the research. Without well-established goals your survey could be too broad or too long. Start by asking:
- Which question or group of questions will be used to meet the research objectives?
- What critical decisions or actions will come from the results?
- Determine which questions provide the “need-to-know” and the “nice-to-know” information.
Tip: Seriously consider cutting the “nice-to-know” questions!
Once objectives have been defined you can frame questions and begin to build out your logic.
2. Question Order
The order in which questions are presented can significantly affect how respondents engage with the survey. A well-organized flow makes it easier for respondents to follow the survey logic, improving the quality of responses.
- If you’re using screening questions to get to your target audience, ask the broadest questions first. Follow those with ever-more specific questions to achieve the precise audience you’re targeting.
- Start with simple questions and place heavy-lifting questions later. Ensure the survey flow is in an easy-to-follow order. Begin with easy questions, group similar questions together and put more complex questions and open-ends later in the survey. Keep the non-screening demographic questions at the end.
- Be sure to provide participants clear instructions and a sense of survey length. Introductory text sets the tone for your survey and informs participants of survey length. Additional text helps transition between survey sections and sets expectations for questions within the sections.
3. Programmer Instructions
Programmer instructions are the bridge between the survey designer and the programmer who will build the survey. These instructions need to be clear and specific to ensure the survey functions as intended once programmed. Key elements to consider:
- Skip logic and branching – as determined by previous responses to ensure relevant questions are shown
- Piping instructions – inserting text from a previous question into the current question
- Conditional questions/responses – items only shown based on previous responses
- Single/Multiple response questions
- Open end response questions
- Anchored and/or exclusive responses
Use consistent programmer language throughout your survey.
4. Question Review
Before programming the survey, it is critical to review all the questions for clarity, coherence, and alignment with the research objectives. This step ensures that the questions are not only correctly worded but also ready to be translated into the programming language.
- Focus: Each question should have a singular focus, not asking double-barreled/multiple questions. It should be as clear and simple to answer as possible.
- Format: Ensure each question will give you useful data. Consider single/multiple response questions, open end response questions, or other question types.
- Text: Use clear, audience-friendly language with correct grammar, spelling, and punctuation.
- Font: Consider underlining, bold, or italic fonts for clarity or emphasis. You can make some questions clearer by formatting or adjusting font size.
- Engagement: Are you able to keep the respondent engaged or have a variety of question types? Incorporating things like card sort or ranking can ensure respondents answer each question thoughtfully.
5. Response Review
Closed-end question responses should be reviewed for clarity and completeness. This step ensures you’ve provided the right choices for participants or given them the opportunity to suggest another option.
- Content: Ensure your response options cover the majority of answers you would expect receive. Consider a write-in option to account for a response you may not have anticipated.
- Text: Use clear, audience-friendly language with correct grammar, spelling, and punctuation.
- Order and flow: Determine when responses remain in a set order (such as counts, months, or a list or countries) vs when to randomize to avoid any biases. Identify responses to be anchored within a list or when responses should be exclusive.
- Conditional formatting: Identify responses which may only be relevant based on responses to previous questions.
- Scale questions: When using a bi-polar scale, make sure it’s balanced. Then decide if your scale will have a middle/neutral option. When using a uni-polar scale, ensure the steps are differentiated. Determine if it’s appropriate to include a ‘not applicable’ option.
By following this 5-step process, you’ll ensure that your survey is thoroughly prepared for programming and data collection. From setting clear objectives to ensuring flawless survey functionality, these steps will help you avoid common pitfalls and deliver high-quality data.
Hemispheres helps people understand their customers as well-rounded individuals. It’s an approach we call Research for Humans. Let’s continue the conversation…reach out at hello@hemispheresinsights.com
Holly Anselm, Senior Strategist
Holly Anselm is a Senior Strategist at Hemispheres Insights. Holly is an expert in leveraging quantitative data to uncover valuable insights, with a sharp focus on project management and attention to detail. She excels at presenting complex data in visually clear formats and training team members in data analysis, quality review, and data presentation. Since joining Hemispheres in 2012, Holly has played a key role in driving strategic projects and delivering impactful results.