Experimental Correlogram Analyses of Artificial Time Series (with Special Reference to Analyses of Oceanographic Data)
Abstract
When dealing with data as they come from nature, the applied statistician is frequently faced with the prospect of selecting and modifying theoretical mathematicalmethods to fit the special conditions imposed by the data themselves. Since theoretical mathematical tools are developed for specific situations they may become unsafe for applications outside the framework for which they were originally intended. Observations on natural phenomena cannot usually be controlled in the sense of a controlled laboratory experiment, and for obvious reasons, they are finite in scope. Hence, some form of experimental modification of theoretical formulae isusually essential if the maximum correct information is to be obtained from the data.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 1951
- Accession Number
- AD1028741
Entities
People
- H. R. Seiwell
Organizations
- Woods Hole Oceanographic Institution