I have responses from participants for two versions of a survey. In each version, the questions, or prompts, are identical - for example, "I am satisfied with my telephone service provider". One survey gives answer options as:

"Strongly Disagree", "Disagree", "Neither Agree nor Disagree", "Agree" and "Strongly Agree".

The older version of the survey has the options:

"Strongly Disagree", "Disagree", "Slightly Agree", "Agree" and "Strongly Agree" (which is neither balanced or symmetrical).

Is it possible to compare data between the two surveys by using the argued assumption (Labovitz, Traylor) that "Likert-type items perform closely to scales that are perceived as equal intervals" and then converting answers to numerical scale data and combining the numerical data?

  • $\begingroup$ Yes, you can compare them using this argument, and many people will disagree with your analyses (because you very well may be comparing apples to oranges...the truth is, you simply don't know). More importantly, ¿what types of comparisons are you trying to make? (This may lead to suggestions for other ways to use "flawed" data like this.) $\endgroup$ – Gregg H Apr 26 '18 at 18:34

Your Answer

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

Browse other questions tagged or ask your own question.