Hi All,
I have a some survey questions where a participant compares 2 things A and B
and makes a judgment on a -3 to + 3 scale with -3 meaning A is much better,
... to +3 B is much better.
This survey is repeated weekly and I have been asked to produce a summary
metric that can be tracked over time. Any thoughts/suggestions?
Do I take an average, median or something else..., a lot of the questions
are ties so I would like to take that into account too?
Here are the categories
-3 - A much better
-2 - A better
-1 - A a little better
0 - A and B the same
1 - B a little better
2 - B better
3 - B much better
Best regards,
Richard.
|