Comments on Presentation by Paul Cox
Abstract
As the first discussant, my main concern will be with what I consider the most important aspect of this problem. It involves the question of the basic underlying model and what the purposes of the investigator were in designing this experiment. As our chairman has just pointed out, this is our primary consideration. This example illustrates what I think is a truism in statistical design and experimentation, viz. that no amount of statistical knowledge and methodology, however great, is a sufficient substitute of substantive knowledge of the field of application and in being able to discuss with the investigator the important questions. For instance, Mr. Cox has indicated that his six points were arbitrarily selected and the analysis of variances relating thereto were not necessarily meaningful. This may be so in these six points but does the person who has the expert knowledge of thrust curves tend to associate important meanings to them, or any other set of critical points. It may very well be that point A has direct application to the understanding of the model underlying thrust curves and that the actual value of A, the time elapsed from the origin, or the rate of ascent from the origin to A are of primary concern to the rocket motor designer.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1963
- Accession Number
- ADP014603
Entities
People
- Bernard G. Greenberg
Organizations
- University of North Carolina at Chapel Hill