How to write an effective survey

Surveys are a low budget, easy to administer option for collecting feedback. A good survey is simple to fill out; therefore, writing them effectively takes thought. See below and download our guide to discover methods and considerations to make surveys honest, fair, and efficient.

Pros/cons of surveys

Before you commit to a survey as your research method, consider the pros and cons:

Pros

  • Realistic to produce: many survey-making tools are available
  • Low cost: several of these tools are free
  • Have the potential to collect large amounts of data

Cons

  • Hardly anyone wants to take a survey
  • Inaccurate answers can be collected due to participants' lack of effort or desire to seem a certain way
  • Occurrence of response bias: The people who respond to the survey may not be representative of the total population. Those who respond may do so because they have a strong opinion, and the neutral population is left unheard.

If it seems a survey isn't the best method for achieving your research objectives, instead consider conducting a focus group, in-depth interviews, audience observance, or media analytics.

Pre vs. Post Surveys

Surveys can provide useful information both before and after a project is completed. School projects, formal research, and communications campaigns alike can employ surveys as a means of collecting information to inform efforts and evaluate effectiveness.

 

Sharing a survey before a project's completion can answer questions like the following:

  • How can I make my project appeal to my audience?
  • Does my audience prefer X or Y?
  • What visual elements can I use to best engage my audience?
  • To what tone does my audience connect?

 

Sharing a survey after a project has been launched can answer questions like these:

  • How well did I achieve my project's objective?
  • What did people learn from my project?
  • Which elements of my project performed well? Which performed poorly?
  • What could be done differently or better next time?

Research ethics

Before creating a survey, ethical standards should be considered.

  • Identify the research sponsor
  • Get informed consent
  • Be clear that participation is voluntary
  • Allow participants to withdraw at any point
  • Give details on the survey's purpose
  • Do not deceive, mislead, influence, harm, or falsify data
  • Uphold respect, honesty, and confidentiality and avoid bias
  • Ask permission to use recording equipment if necessary
  • In general, be as transparent as at all possible!

This list is not exhaustive; you can always define ethic code further depending on your organization, project, and personal values.

Parts of a survey

Surveys are usually structured into 5 sections:

  1. Consent form, involving contact info, option to withdraw, purpose of study, and indication of consent
  2. Introduction that describes the survey's format
  3. Main body questions addressing research concerns
  4. Demographic information if desired
  5. Concluding statement/thank-you note

Survey question do's and don'ts

Do:

  • Develop clear research objectives
  • Make sure all your questions relate to at least one research objective
  • Uphold strong ethical standards
  • Form your survey to meet research needs

 

Don't:

  • Ask leading questions
    • Do you think my presentation on how paint dries was good? Yes or no?

Since this question is somewhat personal to the presenter, it leads the audience to answer "yes." It's also difficult to determine whether something is objectively good or bad in totality without evaluating what elements were more positive or negative.

 

  • Use jargon or inappropriate terminology the audience wouldn’t understand
    • What is your favorite method of bioremediation? Explain why.

Unless this question was given specifically to people who work with or are educated on bioremediation, it would be hard to answer. The general population may not know enough about bioremediation to pick a favorite method.

 

  • Ask questions evoking response bias
    • Do you think it's OK to pick your nose? Yes or no?

This question is sensitive and influenced by social desirability, so it's unlikely that all participants will answer it honestly.

 

  • Assume the consumer thinks a certain way
    • Why are chili cheese Fritos better than classic Fritos?

This question assumes that all consumer prefer one product over the other, which is unlikely to be true. This leads to some participants being unable to answer the question.

 

  • Make the consumer feel dumb by patronizing or confusing them
    • Did you find this webpage on how to write a survey not unhelpful, or was it helpful?

It's hard to tell how to answer this question, which can be frustrating to consumers.

 

  • Use absolutes like "always" and "never"
    • Are you always a good person? Yes or no?

It isn't fair to ask participants to group themselves into such absolute categories. It's rare that a participant identifies as "always"
being a certain way.

 

  • Ask unclear questions
    • In your opinion, what is your experience with Terry Yaki's new sauce collection?

It's not clear what this question is asking. Using words like "opinion" and "experience" in the same question is conflicting.

 

  • Ask more than one question at a time
    • What is your experience with California Betty's new surf wax and when did you use it?

This question is also unclear. It asks about both the product itself and an element of consumer behavior. This question takes more effort to answer and should be separated.

 

  • Include bias
    • How great was your last jazzercise class at Fitness Palace?
      • It was okay.
      • Really great
      • Extremely great!

With this question, the survey is showing bias towards the service by excluding the idea that the consumer may not have liked the
product at all.

Analyzing results

Data collected from surveys can become actionable through proper analysis. When reporting on your data, consider and include:

  • The question as it was asked
  • Number of respondents
  • How respondents were selected
  • How respondents were contacted (if at all)
  • A visual representation of the feedback
  • Valuable data, like totals, averages, and differences

More resources