Maximal type test statistics based on conditional processes |
| |
| Affiliation: | 1. Department of Electrical, Electronic and Computer Engineering, University of Pretoria, cnr Lynnwood Road and Roper Street, Pretoria, South Africa;2. Department of Insurance and Actuarial Science, University of Pretoria, cnr Lynnwood Road and Roper Street, Pretoria, South Africa;3. Council for Scientific and Industrial Research, Meiring Naudé Rd, Lynnwood, Pretoria, South Africa;1. Henry B. Tippie College of Business, Department of Finance, University of Iowa, 108 Pappajohn Business Bldg., Iowa City, IA 52242-1994, United States of America;2. Peter T. Paul College of Business and Economics, Department of Accounting and Finance, University of New Hampshire, 10 Garrison Avenue, Durham, NH 03824-2325, United States of America |
| |
| Abstract: | ![]()
A general methodology is presented for non-parametric testing of independence, location and dispersion in multiple regression. The proposed testing procedures are based on the concepts of conditional distribution function, conditional quantile, and conditional shortest t-fraction. Techniques involved come from empirical process and extreme-value theory. The asymptotic distributions are standard Gumbel. |
| |
| Keywords: | |
| 本文献已被 ScienceDirect 等数据库收录！ |
|
