We want to conduct a survey asking participants to rank a set of 10 fundamental engineering concepts. Ranking a concept 1 would mean the participant believes this concept is the most important engineering concept for an engineer to know, while a ranking of 10 would mean the concept is of least importance.
If roughly half the participants are practicing engineers and the other half are engineering instructors, we would like to know the extent to which these groups might differ in their overall rankings. That is to say, we don't just want to know how these groups differ in their ranking of each individual concept, but rather how one groups aggregated ranked list differs from the others. What would be an appropriate statistical measure for such a comparison?
Thanks for your reply, Nolberto! We were leaning towards using the Mann-Whitney U test. Would that seem reasonable to you? Or are there major differences between the Mann-Whitney U test and Kendall Tau Correlation Coefficient that we should be aware of?
Dear Matthew
I am not a reliable judge for the MWW test because my knowledge of it is very poor
However the MWW test is an application of the Hyphotesis Test and thus working with probabilistic values, which does not appear to be your case
Thank you for your reply, Sarbast! As I understand Kendall's W, it is used for determining the degree of agreement amongst n participants ranked lists within one group. But can/should it be used for holistically determining the degree of agreement across two or more groups aggregated ranked lists?
- Badges
- Science topic
- Similar topics
- Engineering