turn on suggestions

Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

Find a Community

- Home
- /
- Analytics
- /
- Stat Procs
- /
- Linear model after Conjoint Analysis

Topic Options

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

08-22-2016 06:47 AM

Hi,

I am running a conjoint analysis with, let's say 2 attributes with each 2 levels:

Att1: Color

Att1Lv1: Red

Att1Lv2: Blue

Att2: Size

Att2Lv1: small

Att2Lv2: large

Putting this in a linear model, this should look like:

"Choice"=beta0 + beta1*red+beta2*blue+beta3*small+beta4*large

Please correct me if I'm wrong.

Now if I want to analyse this through e.g. PROC MODEL I would put all of the levels after PROC MODEL.

Then SAS would show me part-worths for red, blue and small but not for large. I understand that this is chosen as the "base product" so the part worth of small is the utility compared to "large".

However, how can I also display the part-worth of "large"? Is this even possibile?

Thank you for your help.

I am using SAS Enterprise.

Best,

Alex

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

08-22-2016 09:04 AM

I recommend that you look at the paper "Conjoint Analysis" by Warren Kuhfeld. The example on p. 687 ("Chocolate candy") is essentially the same as the problem you describe. It uses PROC TRANSREG to analyze the data. The documentation for PROC TRANSREG contains additional examples.

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

09-20-2016 10:12 AM - edited 09-20-2016 10:36 AM

Hi Rick,

thank you very much for your answer. The example you mentioned in Kuhfeld (2010) however deals with a rating based conjoint analysis. I am conducting a choice based conjoint analysis. Is this still comparable then?

Update:

If I run a conditional logit model, for example with proc mdc I get part-worths for the attribute levels red, blue and small. "Large" is set to 0 which makes sense to me but I wonder how I can set "blue" to zero, as well to make comparisons possible.

Thank you for your help.

Best,

Alex