论文部分内容阅读
一、2×2列联表的一个简算式在医学科研中,由两个组按照两个定性变量进行比较时通常称为2×2列联表(表一)当两组总观察数n<40,理论数E<5时,一般的X~2检验或Yates连续性修正式均不适用,此时不能以卡方分布来近似计算,而需要直接计算的精确概率来代替,这就是Fisher倡用的精确法(exact methed)。它是假定两个独立变量在边缘总计保持不变的情况下得到频数a、b、c、d任一特定排列的概率,其公式为:当表中任一频数(?)时,需要计算累积概率才能下结论。同时Fisher检验表明是在一特定方向上与零假设有无偏离,因此它是单侧检验,此时p的显著水平为0.05和0.01。若不考
A simplified formula of 2x2 contingency tables In medical research, when two groups are compared according to two qualitative variables, they are usually called 2x2 contingency tables (Table 1) when the two groups of total observations n< 40. When the theoretical number E<5, the general X~2 test or the Yates continuity correction formula is not applicable. At this time, the chi-square distribution cannot be used for approximate calculation, and instead of the exact probability of direct calculation, this is Fisher’s proposal. The exact method used (exact methed). It is the probability of assuming that any two independent variables get any specific arrangement of frequency a, b, c, d when the edge total remains the same. The formula is: When any frequency (?) in the table, it needs to calculate and accumulate. Probability can only be concluded. At the same time, Fisher’s test shows that there is a deviation from the null hypothesis in a specific direction, so it is a one-sided test, and the significant levels of p are 0.05 and 0.01 at this time. If not