Satisfaction likert scale and response order bias


 Hi – the default order for the 1-5 Satisfaction likert is:

1 very dissatisfied; 2 dissatisfied; 3 neutral; 4 satisfied; 5 very satisfied

Is there academic research that validates having ‘very dissatisfied’ as the first response choice (left to right)?  Or is it equally acceptable to reverse the order and have ‘very satisfied’ as the first answer choice?  What’s the theory behind this?  I simply assume that the question library is informed by best practices.  Thanks!

metricsinstitute answered

    Hi Adam,

    I have “by accident” used both directions in the same survey 2 different years in a row. We were very surprised to find a sharp drop in scores in the 2nd survey despite no real change in the organization. 

    The conclusion was that when the left score is”Very satisfied”, it is the 1st choice that the respondent sees next to the question, thus favoring a more positive answer and vice versa.

    The bottom line is that there is not a “right” order of the scale it is crucial to keep consistency across the survey and between the surveys

    Bvi Benita (survey expert) from metrics institute

    metricsinstitute answered

      Hi Adam,

      We’ve found that for left-to-right languages, scales that are laid out from worst/least to best/most are a standard layout with which most respondents will be familiar. I can’t say that this is based on hard research but we’ve been doing surveys for a long time and this is the most common approach.

      A perhaps more important concern is to keep your scales consistent. Flipping the layout, after using one or the other, is generally not a good idea as respondents might miss the change.

      I hope this helps!

      Bri Hillmer
      Documentation Coordinator
      SurveyGizmo Customer Experience Team

      Community Admin answered
        Adam 13 Rep.

        Thanks, Jim.  Do you know if any expertise/research has been baked into the default order in the SurveyGizmo question library?  Said differently, is there a reason it defaults low-high, vs. high-low?

        Appreciate the insight!

        Jim Wetherill commented
          • I do not know. You would have to contact SurveyGizmo for that question. I am just another user like yourself.


          Hi Adam,

          A simple Google search on Likert Scale bias will yield several academic articles on the topic.

          Based on my personal experience with the survey designers in my organization, the Likert Scale order seems to depend on what professor they had for Research Methods in grad school.  I have had people ask to have the scales set up both ways and swear that their preference was the only correct way.

          Good luck!

          Jim W (Moderator) answered


            Question stats

            • Active
            • Views8595 times
            • Answers4 answers
            • Followers1 follower