What's the best way to interpret Likert Scales with equal negative and positive Likert Items that represent each construct presented?

Construct: Accuracy of Data
Scale: Agree-Disagree (1 to 5)

  • The data is accurate? (positive) Responses: 5(Dan),5(James),4(Sarah)
  • The data is not accurate? (negative) Responses: 1(Dan),1(James),2(Sarah)

I know that bar graphs can be used to illustrate how respondents felt about each Likert Item individually (e.g. 45% strongly agreed that the data is accurate), but assuming that the split half reliability method shows that the data I have collected can be trusted from face value, how could I best group the information by like-minded constructs (e.g. Data being the construct in the two questions above)?

1 Answer 1


I would suggest that you might be able to present it on a single scale as "perceived data accuracy", and then just reverse the values of the negative ones and combine them with the positive ones.

Presumably (depending on how your phrased the questions), a 5 for "The data is accurate" would map directly to a 1 for "The data is inaccurate," and likewise for the other four values on each: an inverse correlation. If the questions were phrased in a similar manner for both positive and negative, you might be able to simply group all the negative 1s with the positive 5s, negative 2s with positive 4s, and all 3s, etc. and present them on a single scale.

  • Yeah, that's what I was thinking. I think where I began to become a little confused was how to relay that data in a group setting with regards to overall satisfaction. So for instance if the constructs were data accuracy, ease of use, and functioned properly, how could I say that a user was happy with the product or dissatisfied? Should I set up a score based and use that relay the information?
    – Darryl R.
    Aug 22, 2016 at 12:53
  • Ah, that's different... I don't know your situation, but those three measures might or might not really correlate well, so just finding an average (add all three scores together and divide by three) might or might not give you a truly meaningful result. However, it could be a place to start, and it couldn't hurt.
    – Mattynabib
    Aug 22, 2016 at 19:38
  • You're right, it couldn't hurt. Thanks for the help.
    – Darryl R.
    Aug 23, 2016 at 12:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.