This thesis is concerned with the determination of reject allowance in a job shop added to an order quantity to prevent losses due to defective products. Three cases are considered according to the classes of quality characteristic of the products and the existence of the learning effect in the process:
ⅰ) the quality characteristic of the product is categorized into two classes; conforming and nonconforming item.
ⅱ) the quality characteristic of the product is categorized into two classes and learning effect of the operator exists.
ⅲ) the quality characteristic of the product is categorized into multi-classes. For each case, cost (or profit) model is developed accounting for the cost and/or profit elements involved in the production and contract, and the optimal reject allowance minimizing (or maximizing) expected total cost(or profit) is obtained.
For cases 2 and 3, optimal reject allowances guaranteeing the given probability that the number of conforming products exceed the order quantity are obtained. Sensitivity analyses of optimal reject allowance to the parameters are also performed.