Further Examination of Simulated Percentage Points.
Abstract
In a previous report, Juritz, Juritz and Stephens (1981) discussed methods of obtaining percentage points of distributions by Monte Carlo studies, either by using one long run, or using the average of several smaller runs. The single run has certain advantages--in particular, less bias. The two methods were illustrated using Monte Carlo studies. In this follow up, the comparison is continued, using as illustrations the normal and exponential distributions, for which exact values of order statistics can easily be found. An amalgamation of the earlier report and this one is now published in Juritz, Juritz and Stephens (1983). (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 15, 1983
- Accession Number
- ADA130338
Entities
People
- Michael A. Stephens
Organizations
- Stanford University