Choosing response scales for an event survey
The draft survey uses stars on one page, numbers from one to ten on another and agree or disagree on a third.
The short answer
Pick one or two scale types for the whole survey, label every point in words, keep the direction the same throughout and add a Not applicable option where some delegates could not have had the experience. Consistency matters more than which scale you choose.
Report the spread of answers and the comments, not a single average. Do not compare scores with other events' results, since scale, question and audience differ.
Scale choices at a glance
| Choice | Options | How to decide |
|---|---|---|
| Scale length | Four, five or seven points | Fewer points are easier to answer on a phone; more points rarely change what you will do. Five labelled points suit most event surveys. |
| Midpoint | Include a neutral point or not | Include one when neutral is a real answer. Remove it only if you need a clear lean and accept some delegates will feel forced. |
| Labels | Words on every point, or only the ends | Label every point so delegates read it the same way. |
| Direction | Low to high, always the same way | Keep positive on the same side across the survey to avoid slips. |
| Not applicable | Add a separate option | Use for sessions or services a delegate did not use. Do not count it as a score. |
| Yes or no | Binary question | Use for facts, such as Did you use the quiet room? |
Wording the labels
Use balanced labels with equal weight on both sides, such as Very poor, Poor, Acceptable, Good, Very good. Avoid labels that push towards praise, such as Good, Great, Excellent, Outstanding.
Agree scales suit statements, such as The programme matched my expectations. Rating scales suit quality questions. Pick the one that matches the question, then stay with it.
Do not read more into the numbers than they hold
A mean score hides the spread. Two groups scoring five and one can average three. Show how many delegates chose each point.
Treat the number of responses as part of the result. A high rating from a small, self-selected group says little about everyone who attended.
Comparing across years
Compare only where the question, scale and audience are the same. If you change a scale, mark the break in your records and do not draw a line across it.
Avoid comparing with other organisers' figures. You do not know how their questions were worded or who answered.
Putting it into the survey
Set scale types in the feedback questionnaire builder and map each scaled question to its objective using the objective-to-measure planner. Record observed results, estimates and gaps in the benefit evidence tracker.
Worked example · Fictional example
A seminar team standardises its scales
Fictional organisation and figures, written to show the level of detail that is useful.
The secretariat of a fictional insurance association had used a ten-point scale for the keynote, stars for sessions and agree or disagree for logistics. Responses could not be compared across sections.
For the next seminar it moved all quality questions to a five-point labelled scale from Very poor to Very good, kept three agree statements on expectations, added Not applicable to session questions and used yes or no for facts. Results are shown as counts at each point alongside comments.
Use this yourself
Scale decision table
Copy this table and fill in one row per question type used in your survey.
| Question type | Scale chosen | Points and labels | Midpoint? (yes or no) | Not applicable? (yes or no) | How results are reported |
|---|---|---|---|---|---|
| Session quality | |||||
| Programme as expected | |||||
| Logistics | |||||
| Facts (used, attended) | Yes or no | Yes, No | No | No | Counts |
| Open comments | None | Free text | Grouped by theme |
Handle it in-house, or bring in help?
Your team can usually handle this when
- The survey is short and one owner can fix scales across all questions.
- You are comparing only with your own earlier surveys.
- No one will see the results beyond your committee.
Outside planning help earns its fee when
- Several committees or sponsors want their own scales in one survey.
- Results go to a funder or head office who want comparison with other events.
- The survey runs across languages and formats and needs equivalent scales.
Need a consistent survey structure across events?
An Event Blueprint can set the survey design: which scales are used for which question types, how results are reported without a single average, and how scales stay the same from year to year so your own comparisons hold. The scoring conclusions stay with your committee.
Questions organisers ask
Is a five-point scale always best?
It suits most event surveys, but no scale is correct for every question. Choose one length and labelling style and keep it throughout the survey.
Should we show an average score?
Show the number of delegates who chose each point, with comments. A single average hides the spread and the number of responses.
Can we compare our scores with other events?
Not safely. Differences in questions, scales and who answered make such comparisons unreliable. Compare with your own earlier surveys using the same questions.
When should we remove the neutral point?
Only when you need a clear lean and accept that some delegates will feel forced. For many questions a neutral answer is honest, so keep it.
Related resources
Content record: Draft. Written from the cited sources and checked by automated rules; not yet independently reviewed.