Monday, September 17, 2012

Day 11 - Team H


CONJOINT ANALYSIS

Conjoint analysis helps organizations understand which factors drive decisions. Analysts determine factor preferences by presenting multiple combinations of factors and asking respondents to rank them. Market researchers often use conjoint analysis to determine which product features are most critical to purchase decisions.

Conjoint analysis must currently be run using syntax. Unlike most procedures in SPSS for Windows, conjoint analysis requires the user to invoke two files:

1.Plan File:The plan file contains the combinations that will be presented to the participants.

2.Data File:The data file contains the participants' responses.

The syntax includes the full location of the plan file, but uses an asterisk to alert SPSS to use the file in the data editor as the data file. 

You can also save utilities—the values assigned to each factor from the conjoint—using the utility command. Make sure to include the utility command at the end of the syntax run and indicate where to save the file. You can use this to segment customers based on their preference patterns.

Step 1: Generating the Plan file:
Open SPSS ->Data > Orthogonal Design-> Generate 

Define factors :

Example : In a Hi-Tech hotel survey, IMNU students defined LAPTOP_CARRY, INT_CONNECT_VIDEO_DEMAND,VIDEO_CONF_VOIP, PRICE_PREMIUM as different factors.
They got 9 profiles by orthogonal design.

Step 2:Data -> Orthogonal Design->Display 

This will give you profiles (Multiple combination of factors).Go to your subject and find out their ranking for the given profiles.

Step 3:Generate the data file
This file is generated on basis of ranking provided to different preferences.

Step 4:Run a conjoint Analysis:CONJOINT PLAN='C:\Documents and Settings\Administrator\Desktop\VXLPLAN.SAV'
/DATA=*
/SUBJECT=ID
/FACTORS=LAPTOP_CARRY INT_CONNECT_VIDEO_DEMAND 
VIDEO_CONF_VOIP PRICE_PREMIUM
/RANK=PREF1 TO PREF9
/UTILITY='C:\Documents and Settings\Administrator\Desktop\OUTPUT.SAV'
/PLOT=SUMMARY
/PRINT=SUMMARYONLY.

Step 5: Analyse the output

Author-
Manish Lath (14145)

No comments:

Post a Comment