Writing Better Scale Questions

Scale questions are the basics, the bread and butter of online surveys. They allow a single-select response with the options representing a range, scale or continuum. Scale questions are designed to capture the survey taker’s opinion or sentiment as a point along a “scale” of options. The units of the scale can vary and are usually measure factors such as:

  • Acceptability (Not at All – Completely Acceptable)
  • Agreement (Completely Disagree – Completely Agree)
  • Awareness (Not Aware at All – Extremely Aware)
  • Concern (Not at All Concerned – Extremely Concerned)
  • Familiarity (Not Familiar – Very Familiar)
  • Frequency (Never – Always)
  • Importance (Not Important – Extremely Important)
  • Likelihood (Not Likely – Extremely Likely)
  • Quality (Poor – Excellent)
  • Satisfaction (Not Satisfied at All – Completely Satisfied)

Though SurveyGizmo does not favor one type of scale over another, we found from the SurveyGizmo Benchmark Guide Survey, that 41 percent of survey builders prefer a numbered one to five scale.1

Text labels are being used, but a numeric value, also called a reporting value may be tied to each word or group of words for reporting purposes. For example, on a 3-point scale with the labels Good, Fair and Poor, good may have the value of 3, with Fair being 2 and Poor being 1. These values can help in reporting and analysis to calculate scores based on these text-labeled responses.

The best response scale is one that is easy to understand, clearly discriminates between respondents’ perceptions, and is easy to interpret. We do recommend that you choose a scale and label set and stick with it. This is for three reasons:

  • It’s less confusing to the survey taker if the scale questions you ask do not vary wildly across the survey.
  • Consistent scale use means that your clients (both internal and external) do not have to learn the psychology of different scales.
  • You’ll have consistent data over time so you can compare trends that occur between months, years and across different surveys easily.

You may notice that many scale questions do not include a neutral option (an option placed directly between the two extremes with a neutral option label). When selecting which scale to use, you’ll also need to decide whether you find a neutral option valuable.

  • Those in favor feel that including the neutral option will gather cleaner data without forcing someone to answer on the positive or negative side of the scale.
  • Those who prefer scales with no neutral option, feel that scales that require a choice, positive or negative make action-focused decisions from the survey clearer.

One important thing to note, studies have shown that when lacking a neutral option, a survey taker is more likely to be “nice,” answering in favor of the positive side of the scale more than the negative.

Maybe those powerful scale questions aren’t as basic as we original thought!

Happy surveying!



1Source: SurveyGizmo Market Research Benchmark Guide: 2012, Comparative Analysis on Survey Metrics, Techniques and Trends, What type of scale do you use most often? n=876 Total Sample

Join the Conversation
  • http://twitter.com/rags Rags Srinivasan

    I think, the fact that most survey designers choose a 5 point scale does not say much about whether it is better to use over other scale.
    The broader question is, should the question give middle option at all. That is use a even point scale and require the survey takers to think and take a position one side or another. Odd scales gives a way out for them especially when they see a series of questions.
    -Rags

    • sgizmo

      Great comment! Well, the answer, like so many market research answers,
      is that it depends. Those in favor of a neutral option will gather
      cleaner data without forcing someone to answer on the positive or
      negative side of the scale. But, those who prefer scales with no neutral
      option feel that scales that require a choice (positive or negative)
      make action-focused decisions from the survey clearer. One important
      thing to note is studies have shown that when lacking a neutral option, a
      survey taker is more likely to be “nice,” answering in favor of the
      positive side of the scale more than the negative.

  • Rags Srinivasan

    I think, the fact that most survey designers choose a 5 point scale does not say much about whether it is better to use over other scale.
    The broader question is, should the question give middle option at all. That is use a even point scale and require the survey takers to think and take a position one side or another. Odd scales gives a way out for them especially when they see a series of questions.
    -Rags

    • sgizmo

      Great comment! Well, the answer, like so many market research answers,
      is that it depends. Those in favor of a neutral option will gather
      cleaner data without forcing someone to answer on the positive or
      negative side of the scale. But, those who prefer scales with no neutral
      option feel that scales that require a choice (positive or negative)
      make action-focused decisions from the survey clearer. One important
      thing to note is studies have shown that when lacking a neutral option, a
      survey taker is more likely to be “nice,” answering in favor of the
      positive side of the scale more than the negative.

';