A note on the interpretation of n x 2 tables.

摘要: Data are frequently acquired or extracted in the form of a contingency table two rows and several columns, it is sometimes necessary to consider distribution entries into various cells terms consistency with some hypothesis. Where ranking order columns no consequence, conventional widely recommended x2 tests may De appropriate, pro vided totals not grossly disproportionate number. If they number based on ratio observed expected variance assumption Poisson more efficient (Berkson, 1940; Cochran, 1954). These cases which inconsequential comparatively rare, except relation genetic segregation. Almost always meaningful, any test fails appreciate this clearly inefficient. The %2 gives result uninfluenced by arranging n\ ways, related n-1 degrees freedom, only one likely be biologically meaningful variation. In ad dition, unrelated intuitively estimator.