作者: T. SCHWEDER , E. SPJØTVOLL
关键词: Null hypothesis 、 Statistics 、 Mathematics 、 Line (geometry) 、 Base (topology) 、 Covariance matrix 、 Per-comparison error rate 、 Correlation 、 Contingency table
摘要: SUMMARY When a large number of tests are made, possibly on the same data, it is proposed to base simultaneous evaluation all plot cumulative P-values using observed significance probabilities. The points corresponding true null hypotheses should form straight line, while those for false deviate from this line. line may be used estimate hypotheses. properties method studied in some detail problems comparing pairs means one-way layout, testing correlation coefficients matrix, and 2 x subtables contingency table. illustrated real data.