Power analysis of independence testing for contingency tables

• Autorzy: Sulewski P. , Motyka R.

Identyfikatory

DOI

10.5604/0860889X.1161260

Warianty tytułu

PL

Analiza porównawcza testu niezależności dla tablic dwudzielczych

• Języki publikacji: EN

Abstrakty

EN

Six tests for independence in a two-way contingency table namely chi-squared test, log likelihood ratio test, Neyman-modified chi-squared test, Kullback-Leibler test, Freeman-Tukey test, Cressie-Read test, were examined. It was accomplished with the Monte Carlo method. The Goodman-Kruskal τ index was used to fix dependence in two-way contingency table in Monte Carlo experiments. The examination consisted in determining power functions of the tests. Next, the power functions were compared to each other. It was revealed that differences in power are negligible.

PL

W artykule zbadano sześć testów niezależności dla tablic dwudzielczych, do których należą: test chi-kwadrat Pearsona, test największej wiarygodności, test Neymana, test Kullbacka-Leiblera, test Freemana-Tukeya, test Cressiego-Reada. Dokonano tego metodą Monte Carlo. Wyniki są do siebie podobne. Na wyróżnienie zasługują test chi-kwadrat Pearsona oraz test Cressiego-Reada. Można uznać, że jakość tych dwóch testów (wyrażona w funkcji mocy) jest porównywalna, ale lepsza od pozostałych. Index τ Goodmana-Kruskala wykorzystano do badania zależności w tablicach dwudzielczych za pomocą eksperymentów Monte Carlo.

Słowa kluczowe

EN

two-way contingency table; independence test; Monte Carlo method

PL

tablice dwudzielcze; test niezależności; metoda Monte Carlo

• Wydawca: Akademia Marynarki Wojennej im. Bohaterów Westerplatte
• Czasopismo: Zeszyty Naukowe Akademii Marynarki Wojennej
• Rocznik: 2015
• Tom: R. 56 nr 1 (200)
• Strony: 37--46
• Opis fizyczny: Bibliogr. 13 poz., rys., tab.

Twórcy

autor

Sulewski P.

• The Pomeranian Academy, Institute of Mathematics, Arciszewski ego 22 Str., 76‐200 Słupsk, Poland

autor

Motyka R.

• The Pomeranian Academy, Institute of Mathematics, Arciszewski ego 22 Str., 76‐200 Słupsk, Poland

Bibliografia

• [1] Cressie N., Read T., Multinomial Goodness of Fit Tests ‘Journal of the Royal Statistical Society’, Series B (Methodological), 1984, Vol. 46, pp. 440–464.
• [2] Fienberg S. E., The use of chi squared statistics for categorical data problems , ‘Journal of the Royal Statistical Society’, Series B, 1979, Vol. 41, pp. 54–64.
• [3] Freeman M. F., Tukey J. W., Transformations related to the angular and the square root Annals of Mathematical Statistics’, 1950, Vol. 21, pp. 607–611
• [4] Goodman L., Kruskal W., Measures of Association for Cross classifications ‘Journal of the American Statistical Association’, 1954, Vol. 49, pp. 732–764.
• [5] Horn S. D., Goodness of fit tests for discrete data. A review and an applicatiob to a health impairment scale ‘Biometrics’, 1977, Vol. 33, pp. 237–248.
• [6] Kullback S., Information Theory and Statistics Wiley, New York 1959.
• [7] Lancaster H. O., The chisquared distribution , Wiley, New York 1969.
• [8] Light R. J., Margolin B. H., An Analysis of Variance for Categorical Data ‘Journal of the American Statistical Association’, 1971, Vol. 66, No. 335, pp. 534–544.
• [9] Moore D. S., Recent developments in chisquare tests for goodness of fit, Mimeograph series 459, Department of Statistics, Purdue University, 1976.
• [10] Neyman J., Contribution to the Theory of the x2 Test, Proc. (First) Berkeley Symp. on Math. Statist. and Prob. (Univ. of Calif. Press), 1949, pp. 239–273.
• [11] Pearson K., On the criterion that a given system of deviations from the probable in the case of a correlated system of variables is such that it can be reasonably supposed to have arisen from random sampling ‘Philosophy Magazine Series’ 1900, Vol. 50, pp. 157–172.
• [12] Sokal R. R., Rohlf F. J., Biometry: the principles and practice of statistics in biological research, Freeman, New York 2012.
• [13] Watson G. S., Some recent results i chisquare goodness of fit tests ‘Biometrics’, 1959, Vol. 15, pp. 440–468.