Product Failure Rate Research Articles

Exploration of Variables Inherent in Adopting an Industrial Product

Innumerable authors have indicated that new product failure rates have been alarmingly high [3, 10, 11]. With new product development becoming vital to a firm's maintenance or enhancement of its market position, it is increasingly necessary to seek new ways to minimize new product failure. Confrontation of this problem involves developing a more comprehensive understanding of why and how new products are adopted by consumers so that marketing strategy may be more effectively designed and implemented. Research studies attempting to understand the process of new product adoption have been substantial in areas such as rural sociology and the marketing of consumer goods [8, 10, 15, 17]. However, industrial products have received only limited attention. Recent investigations [4, 14, 16, 20] indicate that industrial product buyers may be affected by many of the same variables as the buyers of consumer goods. In these exploratory studies, variables such as perceived risk, self-confidence, seller image, and prior experience were proffered as important factors in the industrial adoption process. Other evidence reported by economists [2, 5, 12] indicates that variables such as the size of the firm and growth rate were relevant to industrial product adoption. To thoroughly understand why and how new industrial products are adopted, it is necessary to include firm variables, such as growth rate and size of firm, as well as individual behavioral variables. Apparently, industrial product research has been limited because cooperation from firms involved in the adoption process of an industrial product is often difficult to obtain, and individual behavioral variables pertinent to the industrial adoption process are frequently difficult to identify and measure. The objective of this exploratory research is to identify the relevant variables needed in this market to provide a better understanding of the adoption process and to establish an information base for future industrial market research in this area.

The effect of ignoring a simple dichotomy in sequential analysis

A modified sequential analysis is described as a means of testing the hypothesis of no change in product failure rates when the incoming sample is stratified into a simple dichotomy such that conditions which may affect the probability of product failure differ between the two strata. The effect of applying the classical wald sequential analysis which ignores the concomitant information on the dichotomy on Type I(α) and Type II (β) error frequencies and average sample size is studied for a given set of parameters by means of computer simulation. The modified test is shown to maintain the desired α and β levels and a stable mean sample size even when the incoming sample is form a population whose composition differs from past experience. The classical tests actual α and β levels and its average sample size are shown to be dependent on the composition of the incoming sample, which hypothesis is true, and when or not a correct decision is made.

Charts for Confidence Limits and Tests for Failure Rates

Simple charts for determining upper confidence limits for a product failure rate are given here, and their use is illustrated. They are also used to determine statistical demonstration tests. These charts are applicable when the number of failures has a Poisson or binomial distribution or when the distribution of time to failure for a product is exponential.