Distribution of the Sample Version of the Measure of Association, Gamma

摘要: Abstract Random sampling experiments were performed to determine whether Goodman and Kruskal's asymptotic theory (2) for the sample version (G) of measure association Gamma (γ) is applicable small samples. First, was checked against results obtained with a series 5 × population cross classifications values γ covering full range. For samples size fifty, tails distribution (G — γ)/σ corresponded closely unit normal when σ2 appropriate variance derived in (2). Asymptotic normality still worked fairly well replaced by its maximum likelihood estimator, providing that |γ| <.50. There was, however, clear tendency have more observations than predicted from this became very marked large γ's. When σ an upper bound (2), course tightly concentrated. same popul...