Abstract
In comparing and averaging different measurements of a particular quantity, the problem frequently arises of treating discrepant data. Ideally the evaluator should then study the experimental methods in detail, to try to resolve the discrepancies. This however is often either impractical or unsuccessful. An alternative statistical approach is outlined here, using the ``normalized residual``. The theoretical probability distribution of this quantity is compared with that observed for fission chain yield data from a recent evaluation by the authors. The use of the normalized residual in treating discrepant data is explained and compared with alternative methods. (author).
James, M F;
Mills, R W;
[1]
Weaver, D R
[2]
- AEA Reactor Services, Winfrith (United Kingdom)
- Birmingham Univ. (United Kingdom). School of Physics and Space Research
Citation Formats
James, M F, Mills, R W, and Weaver, D R.
The use of the normalized residual in averaging experimental data and in treating outliers.
United Kingdom: N. p.,
1992.
Web.
James, M F, Mills, R W, & Weaver, D R.
The use of the normalized residual in averaging experimental data and in treating outliers.
United Kingdom.
James, M F, Mills, R W, and Weaver, D R.
1992.
"The use of the normalized residual in averaging experimental data and in treating outliers."
United Kingdom.
@misc{etde_10135709,
title = {The use of the normalized residual in averaging experimental data and in treating outliers}
author = {James, M F, Mills, R W, and Weaver, D R}
abstractNote = {In comparing and averaging different measurements of a particular quantity, the problem frequently arises of treating discrepant data. Ideally the evaluator should then study the experimental methods in detail, to try to resolve the discrepancies. This however is often either impractical or unsuccessful. An alternative statistical approach is outlined here, using the ``normalized residual``. The theoretical probability distribution of this quantity is compared with that observed for fission chain yield data from a recent evaluation by the authors. The use of the normalized residual in treating discrepant data is explained and compared with alternative methods. (author).}
place = {United Kingdom}
year = {1992}
month = {Jan}
}
title = {The use of the normalized residual in averaging experimental data and in treating outliers}
author = {James, M F, Mills, R W, and Weaver, D R}
abstractNote = {In comparing and averaging different measurements of a particular quantity, the problem frequently arises of treating discrepant data. Ideally the evaluator should then study the experimental methods in detail, to try to resolve the discrepancies. This however is often either impractical or unsuccessful. An alternative statistical approach is outlined here, using the ``normalized residual``. The theoretical probability distribution of this quantity is compared with that observed for fission chain yield data from a recent evaluation by the authors. The use of the normalized residual in treating discrepant data is explained and compared with alternative methods. (author).}
place = {United Kingdom}
year = {1992}
month = {Jan}
}