| Date: | Fri, 13 Dec 2002 15:29:41 -0800 |
|---|
| Reply-To: | Becky Wu <bwu@luthresearch.com> |
|---|
| Sender: | "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU> |
|---|
| From: | Becky Wu <bwu@luthresearch.com> |
|---|
| Subject: | Conjoint - Importance Score Calculation |
|---|
| Content-Type: | text/plain; charset="iso-8859-1" |
|---|
Dear List,
I am struggling with some very confusing results from a full-profile rating-based conjoint study. We are using SPSS 11.5.
It is my understanding that you can calculate the importance scores by:
For example, you have six attributes, and for attribute A,
(Attribute A Maximum Utility Score - Attribute A Minimum Utility Score)/Sum of all six attributes' ranges from the highest utility score to the lowest utility score
I am inserting the SPSS summary output here for all my six attributes, please see below. However, my calculated importance score for the attribute "Hard" should be (0.3575 +0.1887)/2.5902 *100= 21, instead of 17.68. Similarly, Digital will not be the most important attribute based on the above calculation. Am I wrong or is SPSS wrong? Can some one here share some guidance and insights with me? Thanks a lot!! - Becky
*******************************************************************
Averaged
Importance Utility Factor
+--------+ HARD Hard Copy Media
I17.68 I -.1887 --I Birthday Cards
+--------+ .3575 I---- Newsletter
I -.1752 --I Letter
I .0064 I Postcards
I
+---------+ DIGITAL Digital Media
I20.40 I -.0879 -I Birthday Cards
+---------+ .2238 I--- Newsletter
I .0555 I- Letter
I -.1914 --I Postcards
I
+-------+ CONTENT Nature of Content
I16.21 I .0844 I- Industry Informative
+-------+ -.0965 -I Seasonal
I -.0022 I Homey/Folksy/Persona
I .0143 I Slick/Glossy
I
+--------+ FORMAT Format
I18.18 I -.2185 --I None/Generic
+--------+ -.1025 -I Sig/logo/pic
I .2214 I-- Sig/logo/pic/custome
I .0996 I- Sig/logo/pic/name/se
I
+----+ CORRES Purpose for Correspondence
10.57I I .0153 I Prospect/lead follow
+----+ -.0375 I Market closed loans
I .0222 I Status in-process bo
I
+-------+ FEEDBACK Feedback Capability
I16.96 I .0966 I- Track if customers r
+-------+ .3482 I---- Commission tree to t
I -.2813 ---I Notification of birt
I -.1635 --I Number of mailings a
I
5.8758 CONSTANT
Pearson's R = .976 Significance = .0000
Kendall's tau = .886 Significance = .0000
Kendall's tau = 1.000 for 4 holdouts Significance = .0208
| 
