Comments on Presentation by Paul Cox
Abstract
The panelists so far have covered just about everything that I had in mind. I certainly agree in the main with the comments they have made regarding just what particular points on the observed curve or what particular function of the observations may be of interest. Also, I agree with the comments regarding the desirability of fitting a "rational" model, which presumably can be supplied, at least in approximate form, by the engineers. I wish, however to expand on a very important point. Many of the remarks of panelists about design and analysis have been engendered by the existence of "noise" along the curve for an individual motor and the probable lack of independence of successive observations. I wish to emphasize that there is another, and probably much -more important, "noise" component involved. The latter arises from the fact that a group of motors which are constructed and treated alike, insofar as can be managed, will, nevertheless, have inherently somewhat different curves. That is, there is "between-motor" noise as well as "within-motor" (along the-curve) noise. The existence of between-motor noise must be taken into account for proper experiment design and analysis.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1963
- Accession Number
- ADP014604
Entities
People
- H. L. Lucas
Organizations
- North Carolina State University