This paper aims to offer some clarification on buying behavior of some brands of children's clothing by finding a method to improve Fishbein's multiple-attribute theory. Conclusion 1: An Application of the Fishbein's Theory
(1) We found that the value of general expectation a_i, where i was the number of an evaluated item, was dependent upon the value of expectation b_ij In this paper, we referred to the sum of the values of expectation b_ij as A_j. : the attitudes on a brand_j.
(2) We found that the value of expectation showed significant difference among brands and that each brand had the each purchaser's belief.
(3) We found that almost all brands show the same tendency as a result of plotting brands on the plane coordinates where the axis of ordinates was the number of brand purchasers and the axis of abscissas was the simple sum of values of expectation. And we explained why a little exceptional brands deviated from the trend.
(4) We used a multi-regression analysis where the object variable was the number of purchaser of the brands and independent variable was the value of general expectation. Precision of estimation was 0.97 and we obtained the following a multi-regression model:
y₁=-157.37+13.56x₁₁+15.28x₂₁+12.51x₃₁+12.67x₄₁
where y₁ was the number of purchasers of brand 1; x₁₁ was the constant cheap price; x₂₁ was the constant of good color/design; x₃₁ was the constant of ease wash with low wrinkle; and x₄₁ was the constant of large discount rate.
Conclusion 2: Considerations on the brands of children's clothing
(1) The value of expectation had no significant difference among tank tops, T-shirts and trainers of the same brand names.
(2) The differences among b_ij led to the characteristics of the brand purchasers. The cheap price of the brands had a relationship to the characteristic that their children were males and had lower average age. Good colors and designs had a relationship to the characteristics that their children were females and had higher average age at the significant level of 0.05.
(3) We used cluster analysis to brands in order to explain the differences between theoretical and real values. The obtained four groups of brands are characterized by their attributes of the brand items. The first group was referred to as natural children type, the second group as regular-directed type, the third group as children's clothing brand type and the fourth group as fashion brand type.

View full abstract

An application of the regression analysis has been made to the foot data of 241 young women ranging from 18 to 23 in age and 205 elder women ranging from 60 to 89 in age. In order to make clear their foot form characteristics, shoe sizes of ball joint girth to foot length and foot breadth to foot length have been discussed with the regression analysis. The results are as follows:
1) A foot length proportion, in terms of the ratio of ball joint girth to foot length, has been classified into three types: stubby, normal and slender. For the young, a stubby type conforms to E-3E; a normal type conforms to C-E and a slender type conforms to B-C. By contrast, for the elder, conforms to 2E-F, D-3E and C-D respectively. This classification covers 96% of the subjects i.e., 21-25 cm in foot length. The size designation of ball joint girth is indicated according to the Japanese Industrial Standard for shoes.
2) A foot breadth proportion, in terms of the ratio of foot breadth to ball joint girth, has been classified into three types: wide, normal and narrow. A discussion has been made within 203-245 mm in ball joint girth which covers 95% of the total subject. The normal type for the elder corresponds to the narrow type for the young. The narrow type for the elder corresponds to the normal type for the young. A foot breadth in the elder is thus wider than that in the young by one rank.

View full abstract