Information Technology Reference
In-Depth Information
Table 11.11
Example of choice set
Attribute
Online shop A
Online shop B
Using neither online shop
Store
No
No
-
Saturation
Low
High
Procedure
Easy
Easy
Postage
Free
500 yen
Price
3,000 yen
4,000 yen
negative. In contrast, it is affirmative for the online purchase if the value is
significantly estimated as positive.
Response data of 157 students from the Ehime University and 352 students from
the Ehime University Senior High School were analyzed. First, the dummy variable
university dummy, whose value is 1 for a college student, was added and the cross
term was multiplied by the university dummy; then, an attribute variable was
added. An important difference in the attribute between university and high school
students was analyzed by the time duration of online shopping. Second, the usage
dummy, whose value is 1 for those who have used online shops in the past, was
added and the cross term was multiplied by the usage dummy; then, an attribute
variable was added, and the differences in the preference by use of an online shop
was investigated.
The coefficient estimated in conjoint analysis indicates marginal utility, i.e., the
increment of utility. In analysis, all the attribute variables were calculated as a
random parameter, and the standard deviation parameters of the coefficient that
were not significant were re-calculated as a non-random parameter. NLogit 4.0 was
used as the analysis software.
11.5.2 Results and Analysis
The investigation results are illustrated in the Tables
11.12
and
11.13
. The bold
lettering in the table indicates the nonrandom parameters.
From the complete results, ASC is estimated as significantly negative, and
respondents are considered to be positive about using an online shop. A positive
and significant estimated result was obtained for saturation and purchase procedure
of the site. It is considered that a respondent's utility increases when the site
saturation becomes high and the site's purchase procedure becomes easy. A nega-
tive and significant estimated result was obtained for the goods price and postage,
that is, if the product price and postage increase, the respondent's utility decreases.
Because the value of the marginal utility of each attribute is considered to
change with the student education level and usage experience, these differences
are verified as described below.
