I am wondering if someone can help me with the interpretation of the MaxDiff Score coming from SurveyGizmo.
The scores across all attributes sum to 0. I have a client that wants this re-scaled to be sum to 100.
So: My questions:
- Is there any rationale for the range of scores (i.e my scores range from +.5 to -.3, would it mean anything different if the range were +50 to -30?)
2. Also, any suggestions on how to re-scale this to 100?
3. What about the legacy reporting of the Bayesian average? Can I do something with that to re-scale?
The reporting section in this article provides a breakdown of how the MaxDiff score is derived:
(# times attribute was selected as best – (minus) # times attribute was selected as worst) / # times the item appeared to respondents
Hopefully that helps clarify with regard to the rationale for the score range?
With regard to re-scaling, I found some resources online but they are all pretty stat-heavy (maybe they will be helpful for you):
Hope this helps!