1. Knowledge Base
  2. Building Great Questions

Choosing Response Options for Survey Questions

Looking for tips to write the very best questions on your Polco polls and surveys? Below you will find the best response options that scientists at Polco/National Research Center recommend.

 For more information on crafting quality survey questions, please see our webinar How to Craft Great Survey questions.

A couple of notes:

  • The use of a DON’T KNOW is not listed in these scales, but please consider adding this option at the end if you are asking a question where the respondent really could not know. 
A number of the scales we recommend are skewed positively or negatively.  (This means there are more positive or negative points rather than a symmetrical number.) We often skew scales on quality, problems, frequency and importance scales but not on support/favor questions. We create these skewed scales because we are countering social desirability bias (tendency to give more positive or negative responses that are socially acceptable) and/or want to make sure the scale is sensitive enough to capture full sentiment.  (If every item in a list is considered minimally important, it does not make sense to provide 2 points on how unimportant an item is.) 

Quality scales: good for evaluating services or community characteristics

  • Excellent, Good, Fair, Poor
  • Very good, Good, Neither good nor bad, Bad, Very Bad
  • Very well, Somewhat well, Somewhat poorly, Very poorly
  • Very satisfied, Somewhat satisfied, Neither satisfied nor dissatisfied, Somewhat dissatisfied, Very dissatisfied

Problem scales: good for identifying concerns not sufficiently addressed by service delivery

  • Major problem, Moderate problem, Minor problem, Not a problem
  • Large problem, Small problem, Not a problem at all
  • Major concern, Moderate concern, Minor concern, Not a concern

Change scales: good for estimating impacts of programs, interventions

  • Much better, A little better, About the same, A little worse, A lot worse
  • Much better, A little better, No change, A little worse, Much worse

Amount scales: good for estimating consumption or use  (consider a "Don’t know" response option here)

  • Too much, About right, Too little
  • Too many, About right, Too few
  • A great deal, Quite a bit, Some, Not too much, Very little
  • None, A few, Some, Most, All
  • None, A little, A lot
  • Far too much/many, Somewhat too much/many, Right amount, Somewhat too little/few, Far too little/few
Importance scales: good for determining resident values
  • Essential, Very important, Somewhat important, Not important
  • Extremely important, Very important, Somewhat important, Not at all important
  • Very important, Moderately important, Slightly important, Not important

Frequency scales: good for estimating use of services (if you are asking a series of questions using a grid format, make sure the frequencies make sense.  Some types of activities make sense on a weekly level while others make sense on a monthly or annual level.)


  • Never, Sometimes, Usually, Always
  • Never, Almost never, Sometimes, Almost always, Always
  • Never, Seldom, Sometimes, Almost always, Always 
  • Never, Rarely, Sometimes, Frequently, Always


  • Never, Once or twice, More than twice
  • Not at all, 1 time, 2 times, 3 or more times 

 [times a year]

  • Never, Once a year, Once a month, Once a week, Every day

 [times a month]

  • Never, One or two times a month, About once a week, A few times a week, Every day
  • Never, 1 to 12 times, 13 to 25 times, 26+ times

Likelihood scales: good for projecting use of services not yet offered (consider a "Don’t know" response option here)

  • Very likely, Somewhat likely, Somewhat unlikely, Very unlikely
  • Very likely, Likely, Unlikely, Very unlikely
  • Definitely will, Probably will, Probably not, Definitely not
  • Definitely would, Probably would, Not sure, Probably would not, Definitely would not

“Polar” scales: good for determining policy options (consider a "Don't know" response option here)

  • Strongly support, Support, Neither support nor oppose, Oppose, Strongly oppose
  • Strongly support, Somewhat support, Somewhat oppose, Strongly oppose
  • High priority, Medium priority, Low priority, Not a priority,
  • Support/Oppose 
  • Strongly favor, Favor, Neither support nor oppose, Oppose, Strongly oppose 
  • Strongly favor, Somewhat favor, Somewhat oppose, Strongly oppose

Agree/Disagree: good if no better questions, as these require a positively or negatively skewed stem/list of items. (consider a "Don’t know" response option here)

  • Strongly agree, Agree, Neither agree nor disagree, Disagree, Strongly disagree 
  • Strongly agree, Somewhat agree, Somewhat disagree, Strongly disagree

Yes/No: For youth or respondents with minimal literacy.

Yes, No

Yes, No, Maybe

Yes, Sort of, Not really

Source: Citizen Surveys for Local Government: A Comprehensive Guide to Making Them Matter

© National Research Center, Inc., 2008, updated in July 2020