Microsoft Word 2007 ichrie conference Proceedings Final-Final 06-06-07. doc

 
2007 Annual International CHRIE Conference & Exposition 
61
Pre-survey 
The pre-survey was performed in order to make a decision about final menu selection attributes and the 
level of the attributes to be used in the main survey. The subjects were asked to rank the order according to the 
importance for the attributes presented (taste, price, nutritional content, speed of serving time, appearance of food, 
sanitation, variety of menu items ingredients, type of food) to elicit young adults’ general perception for the 
important attributes they use when selecting menu items. 
The point of “marginal cheapness” and the point of “marginal expensiveness” for the typical lunch menus 
(entrée, salad, pizza, sandwich combo) for on campus dining were examined through a PSM (Price Sensitivity 
Measurement) investigation (Westendorp, 2003). The information that customers want to have when selecting a 
menu and the detail of nutrition information as necessary information, was examined. Pictures were presented to 
describe the appearance level of menu items and were ranked according to the preference of the subjects.
Main survey 
The instrument of the main survey was developed based on the results of the pre-survey and the focus 
group interviews. Four attributes were elicited as most important from the pre-survey and focus group study: taste, 
price, appearance, and nutrition facts. In other words, the number of attributes was reduced to four due to a possible 
limitation of cognition by subjects. Research by Cacioppo & Petty (1982) and Bagozi & Utpal (1999), shows people 
to be “cognitive misers” so they often limit the number of factors used to make a decision. Table 1 presents 
attributes and levels of attributes used in this study. The four attributes selected had two levels respectively. The 
number of possible combinations of all attribute levels is 2
4 
= 16. But, this study used a fractional factorial design of 
11 combinations including 2 combinations as a holdout set. The holdout set was set in the orthogonal plan of SPSS 
in order to verify the validity and reliability for the conjoint model of this study (Addelman, 1962). Additional 
attributes would substantially increase the number of profiles needed for comparison and would, therefore, cause 
data collection difficulties (Green & Srinivasan, 1990).
The full profile method under a fractional factorial design was used. The fractional factorial design is 
usually utilized in order to reduce the number of evaluations collected while still maintaining orthogonal among the 
levels and subsequent part-worth estimate (Gustafsson et al.,1999). The SPSS 12.0 conjoint procedure can specify 
an orthogonal array of 11 profiles that present a complete concept. 
The survey for the conjoint analysis was performed on the 250 university students. All respondents were 
presented 11 menu profiles and asked to rank each menu profile in terms of their preference. An 11-point scale was 
used to rank the profiles. The best menu profile was ranked as “1” and the worst one as “11”. 
†
: 
based of PSM (Price Sensitivity Measurement)