On a General Class of Chi-Squared Goodness-of-Fit Statistics
Tarald O. Kvålseth
Abstract
Tarald O. Kvålseth
Abstract
A class of chi-squared goodness-of-fit statistics is presented as being based on the so-called divergence of one probability distribution from another. A still more general class of goodness-of-fit statistics is then presented by eliminating some of the restrictions required of divergence-based statistics. This most general class of statistics includes, as particular cases, a variety of statistics used in the published literature.
OpenAlex reports 1 citations for this work. Citation counts describe recorded attention and do not establish research quality.
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.
A class of chi-squared goodness-of-fit statistics is presented as being based on the so-called divergence of one probability distribution from another. A still more general class of goodness-of-fit statistics is then presented by eliminating some of the restrictions required of divergence-based statistics. This most general class of statistics includes, as particular cases, a variety of statistics used in the published literature.
Key concepts: Goodness of fit, Statistics, Mathematics, Divergence (linguistics), Class (philosophy), Probability and statistics, Econometrics, Computer science