The following example involves a 2x2x3 design with Color (Red vs Green)
x Shape (Round vs Rectangular)xPrice ($2 vs $4 vs $7) as the factors. Generally,
the steps involve first creating an orthogonal design to find out which
product profiles to actually test in your survey. Then, after the
profiles have been ranked by the respondents in your survey, run conjoint
analysis on the rankings to obtain the utilities for each level of each
attribute (factor).
Data/Orthogonal
Design/Generate
Name
& Label your first factor
Go
to "Define the value labels" for your first factor
Define
those value labels
Name
& Label your second factor
Define
the value labels for your second factor
Name
& Label your third factor
Define
the value labels for your third factor
Click
on "Create New Data File" to save your orthoplan to disk
Create
the New Data File on disk
Click
on "OK" to actually generate and save the orthoplan onto your disk
Output of orthoplan run.
IMPORTANT-- View of the SAV file created by Orthoplan procedure.
Note. This SAV file contains the 8 profiles you need to have your respondents rank or rate. In this example, the respondents must rank or rate only these 8 in your survey even though there are actually 12 possible combinations of the 2x2x3 product profiles. (Orthoplan is telling you that having respondents rank these 8 profiles will probably suffice.)
Enter
the Survey Results--here, the Rankings of Profiles by Subjects (create
the SAV file).Note that the data shown in this SAV file are RANKINGS where a "1" was the most favorite profile and "8" was the least favorite profile.
First,
get into Syntax mode in SPSS
Create and save the Conjoint Analysis Syntax file.
Execute
the Conjoint Analysis Syntax file.
Note. To execute the syntax file, highlight the stuff you typed into the syntax file and then click on the arrow icon (execute icon).
Summary utilities and importance scores output.
Sample of utility file (SAV) created by the Conjoint run.
Sample of a utility graph.
