Learning patterns/Using ordered response scales

From Meta, a Wikimedia project coordination wiki
A learning pattern forsurvey
Using ordered response scales
problemSurvey questions often ask for ordered responses, but it can be hard to think of them on the spot
solutionUse these example responses scales for your surveys
creatorEGalvez (WMF)
endorse
created on20:23, 3 June 2016 (UTC)
status:DRAFT

What problem does this solve?[edit]

Survey questions often ask for ordered responses. These are questions that have responses like: strongly agree/strongly disagree, or Very important/Not important. These can be hard to think up when you are doing a survey.

What is the solution?[edit]

Researchers have done a lot of research on this topic and there are responses that are often accepted as established. View the list below to find responses that work for you! Note: If the words do not translate well from English into another language, use number scales instead of words. See this learning pattern. --learning pattern needed-- for more information.


Dichotomous (Forced –choice or anchor points)[edit]

NOTE: Dichotomous pairs are not recommended for most survey questions. These options should only be used for filtering people to the right questions (e.g. did you attend the event? Yes/No), or for voting. You should often include "not sure/don't know" or "not applicable" as an option as well for these lists."

Fair Unfair
Agree Disagree
True False
Yes No
Pass Fail
Minimally Maximally
Not at all Completely
Three-point scales[edit]
Agree Neither agree nor disagree Neutral Disagree
More than I would like About right Less than I would like
Too harsh About right Too lenient
Too heavy About right Too light
Too much About right Too little
Not at all Moderately Extremely
Too Strict About right Too lax
None Some A lot
Increased No Change Decreased
Improved Stayed the same Digressed
Not at all Occasionally Frequently

Four-point Scales (Forced -choice)[edit]

Poor Fair Good Excellent
Most of the time Some of the time Hardly ever Very seldom/Never
Strongly agree Agree Disagree Strongly disagree
Exceeded Met Nearly met Missed
Below standard Approaching standard At standard Above standard
Definitely will not Probably will not Probably will Definitely will
Not important Somewhat important Very important Essential
None Little Some Substantial
Not at all Not very well Fairly well Very well
A lot Quite a bit A little None

Five-point Scales[edit]

Excellent Above average Average Below average Very poor
Much better Somewhat better Stayed the same Somewhat worse Much worse
Much stronger Somewhat stronger No change Somewhat weaker Much weaker
Strongly agree (Somewhat) Agree Neither agree nor disagree/Neutral (Somewhat) Disagree Strongly disagree
Very high Above average Average Below average Very low
Way too little Too little About right Too much Way too much
Far too little Too little About right Too much Far too much
Much higher Somewhat Higher About the same Somewhat lower Much lower
Not at all Slightly Moderately Mostly Completely
One of the best About average Average Below average One of the worst
Completely satisfied Very satisfied Fairly well satisfied Somewhat dissatisfied Very dissatisfied
Very satisfied Satisfied Neither satisfied nor dissatisfied Dissatisfied Very dissatisfied
Almost always Often Sometimes Seldom Never
Always Very often Fairly many times Occasionally Never
All of the time Most of the time Sometimes Rarely Never
All A lot/ An extreme amount Quite a bit A little/ Some None
Very good Good Fair Poor Very poor
Very often Regularly Sometimes Once or twice Never
Very important Quite important Fairly important Slightly important Not at all important
Not like Somewhat Moderately Very much Totally
Definitely True True Do not know False Definitely false
Severe Moderate Mild Very mild None

Six-point Scales (Forced -choice)[edit]

Always Very often Fairly often Sometimes Almost never Never
Always Most of the time Often Occasionally Rarely Never
Completely satisfied Very satisfied Somewhat satisfied Somewhat dissatisfied Very dissatisfied Completely dissatisfied

Seven-point Scales[edit]

Very dissatisfied Moderately dissatisfied Slightly dissatisfied Neutral Slightly satisfied Moderately satisfied Very Satisfied
Far below Moderately below Slightly below Met expectation Slightly above Moderately above Far above
Very poor Poor Fair Good Very good Excellent Exceptional

Words to watch out for[edit]

The following words can bias scale if not careful to balance depending on number of responses:

  • Slightly vs. Somewhat
  • Very/Extremely
  • Completely/Totally
  • Excellent vs. Exceptional

Scales for international use[edit]

Translating these scales into another language can be very challenging. There are always slight differences in the meanings of words based on the language and culture. If a survey is being translated or used in multiple countries, it is best to use a number scale and show the extremes only. For example:

  • 0 - Very poor
  • 1
  • 2 - Average
  • 3
  • 4 - Excellent

Endorsements[edit]

See also[edit]

Related patterns[edit]

External links[edit]

References[edit]