This thesis is concerned with the determination of the standard product sizes when customer's preferred sizes have a normal distribution, or bivariate normal distribution. The probability of saling a product is assumed to be the value of a function of the degree to which standard size deviates from the customer's real requirement. This can be converted to another assumption that the probability of sale is the value of a function of the degree of difference between two adjacent standard product sizes.
The optimal standard size set and the product scale are determined to maximize total expected profit during a planning horizon. This means that the standard size set is determined to satisfy as many customers as possible and at the same time to bring the company with maximum profit. Therefore, lost-sale cost and set-up cost are major considerations in this study.
In relation with customer's preference, the following two cases are considered:
(1) Within an acceptable range, the customer's preference is well satisfied and the probability of success in sale in 1(one) but beyond the range, the customer's preference is not satisfied, so the probability of success in sale is 0(zero).
(2) Probability of success in saling one product size is a function of the degree of difference between two adjacent standard product sizes.
For these, analytical formulae are derived and some numerical examples are given to illustrate the solution methodology for each case.