Likert Scale vs Conjoint Analysis. Which to use?

#1
Dear Forum professors,

I am planning to conduct a questionnaire to obtain information of a tenant's perspective on management in a building. I would need to ask the tenants to fill up the questionnaire about how important they think each element in a management is. The sample size will be (n = 300)

I am not sure if a likert scale would work better or a conjoint analysis in this situation.

The outcome of the questionnaire must allow me to rank the elements in the building management according to the importance perceived by the tenants. I would have roughly 20 items to rank

For example, Items to rank are : Cleanliness, Security, Maintenance, Repair. When i get back the data, i would have to be able to rank them according to importance 1) Cleanliness 2) Repair, 3) Maintenance 4) security

So would like to know how do i structure my questionnaire to achieve this. And could you suggest me any test that allows me to analyse the data to rank them.

Your kind suggestion is much appreciated =)
 

Miner

TS Contributor
#2
You would probably get better results using a conjoint approach. The underlying issue that you will encounter is best expressed by the Kano model (see http://en.wikipedia.org/wiki/Kano_model). The items that you listed would be described as "Must-be Quality" in the model. Consequently when rated by a Likert scale, all would be rated very high and be difficult to differentiate. The only way to get differentiation is to force a comparison as is done in Conjoint analysis.