Skewness by splitting the scale parameter
Ingo Klein, Matthias Fischer
Abstract
Open-access reader
Ingo Klein, Matthias Fischer
Abstract
Open-access reader
There are several possibilities to introduce skewness into a symmetric distribution. One of these procedures applies two dfferent parameters of scale - with possibly different weights - to the positive and the negative part of a symmetric density. Within this work we show that this technique incorporates a well-defined parameter of skewness, i.e. that the generated distributions are skewed to the right (left) if the parameter of skewness takes values less (greater) than one. Secondly, we prove that the skewness parameter is compatible with the skewness ordering of van Zwet (1964) which is the strongest ordering in the hierarchy of orderings discussed by Oja (1981). Hence, the generated (skewed) distributions can be ordered by the skewness parameter.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
There are several possibilities to introduce skewness into a symmetric distribution. One of these procedures applies two dfferent parameters of scale - with possibly different weights - to the positive and the negative part of a symmetric density. Within this work we show that this technique incorporates a well-defined parameter of skewness, i.e. that the generated distributions are skewed to the right (left) if the parameter of skewness takes values less (greater) than one. Secondly, we prove that the skewness parameter is compatible with the skewness ordering of van Zwet (1964) which is the strongest ordering in the hierarchy of orderings discussed by Oja (1981). Hence, the generated (skewed) distributions can be ordered by the skewness parameter.
Key concepts: Skewness, Scale (ratio), Hierarchy, Shape parameter, Statistical physics, Mathematics, Distribution (mathematics), Statistics