University of Jos Institutional Repository >
Natural Sciences >
Mathematics >
Please use this identifier to cite or link to this item:
http://hdl.handle.net/123456789/1856
|
Title: | On Regression Approach to ANOVA Designs |
Authors: | Ukwu, Chukwunenye |
Keywords: | Indicator Matrices Structure |
Issue Date: | 2014 |
Publisher: | International Journal of Mathematics and Statistics Invention (IJMSI) |
Series/Report no.: | Vol. 2;No.8; Pp 47-54 |
Abstract: | Regression approach to one-way (one-factor) Analysis of Variance problems was investigated using indicator variables to represent the selection of data from alternative levels (treatments) of the factor. A matrix, U of indicator variables Xi j was obtained, from which 1 () tt U U U y yielded the means of the output responses 1corresponding to given treatments rightoff- the-bat. The key benefit is that the matrices and , ( ) , t t t U U U U U y
1 () tt U U U y have very simple structures, which could be deployed for an easy generation of a desired ANOVA table and the establishing of Confidence and Prediction intervals. |
URI: | http://hdl.handle.net/123456789/1856 |
ISSN: | 2321 – 4767 |
Appears in Collections: | Mathematics
|
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.
|