Picking the right question types
Scale and ranking questions
Scale and ranking questions can be used to ask respondents how they feel about each item, to rate items on a scale, or to rank items in order of importance or preference. This is where answer options are redefined but on a scale. Answer options can be in words or in numbers, if the meaning of the numbers is explained.
For example, ‘How useful was the patient information provided to you at time of diagnosis: a) very useful, b) useful, c) neutral, d) unuseful, e) very unuseful, f) prefer not to say’ is an example of a scale question.
‘Please rank each of the following items in order of importance, with 1 being most important and 4 being least important: easy to read, use of diagrams and images, accessible online, available in print’ is an example of a ranking question.
- Gives more information on each item
- Allow respondents to give more detailed answers rather than ‘Yes/No’ answers
- Provides average and measure of spread
- Scale questions allow you to ask more questions with the same measure, making it easier for respondents to answer
- Takes more time to answer
- Places equal importance on each option and may force differences that don’t exist
- Can be complicated to analyse data
- Ranking questions don’t capture distance (i.e. how much more important one choice is over the other)
Things to watch out for with scale questions
- Make sure the answer scales correspond to the question
- The answer options must be balanced so that it doesn’t lead respondents to answer in a certain way
- Don’t use mid points that aren’t meaningful (e.g. ‘neither likely or unlikely’ or ‘neither important or unimportant’). 'Neutral' can be an example of a meaninful midpoint depending on the question. You can also include no midpoint.
- Make sure that the answer options don’t overlap
- Each option must be of equal magnitude
- Clearly label number scales with what the number means