Product Development Studies
Companies invest considerable resources in developing new products and services. Research studies play a critical role in providing information about many aspects of a product or service: its effectiveness, safety, reliability, and—ultimately—its value to the prospective customer. Good design of these studies, based on statistical principles, and powerful analyses of the data they generate can save companies time and money by providing accurate information in a timely manner, before the products are marketed.
Statistical science contributes to the product development cycle in several important ways:
- Design and analysis of individual experiments, field tests, and surveys
- Synthesis of research results from multiple studies
- Development of statistically-based systems for prediction, classification and diagnosis
Exponent statisticians assist clients to insure that a study’s design and scope will generate sufficient and pertinent information to resolve key research and development issues regarding the product or service. Care must be taken to ensure that the problem is defined with sufficient precision, so that meaningful research hypotheses can be formulated and evaluated with the data that will be gathered. The correct quantities must be measured on a sufficient number of test units. Studies with too few observations lack statistical power and cannot yield convincing answers to the questions that originally motivated the work. Studies with too many observations may misallocate precious resources and risk confusing statistical and practical significance.
Controlled experiments, sample surveys, and observational studies are different ways of collecting data on products and services. Researchers must be cognizant of the strengths and limitations of each approach in determining which is most suitable given the characteristics of a particular project. Controlled experiments, such as clinical trials, offer the strongest basis for causal inference, but they are not always feasible in projects involving human subjects. Sample surveys or observational studies may be used in such cases, but their conclusions require more qualifications and attention to uncontrolled and potentially distorting (or “confounding”) factors.
The development of a product or service may be advanced not only by designing individual studies but also by integrating and synthesizing information from previously executed studies. Because of organizational changes, personnel turnover, or shifting objectives, companies may find that development problems may be understood better and addressed more effectively through more rigorous analysis of existing data. Using state-of-the-art statistical methods and models for “meta-analysis,” Exponent statisticians pool historical data and examine the evidence for clues to potential causes of unsatisfactory outcomes. The findings of these analyses can influence the design of further studies to confirm cause-and-effect mechanisms and improve the next generation of the product or service.
Statistical Products and Services
Some components of products or services are essentially statistical in nature. Companies in such diverse industries as medical diagnostics, health surveillance, spend management, advertising, and fraud detection seek to develop products and services that efficiently analyze large amounts of data to extract and communicate valuable information to customers. In assisting client with these development projects, Exponent statisticians employ classification and regression trees, neural networks, support vector machines, and other contemporary methods for data mining.