四格表卡方检验的第二类错误

• 作 者: 徐秦 施济民
• 机 构: 新疆医学院卫生统计学教研室,新疆医学院卫生统计学教研室,
• 出 处: 新疆医学院学报
• 发表日期: 1987年02期
• 关键词: 第二类错误 显著性检验 备择假设 微型电子计算机 β值 不知道 检验假设 样本含量 脑膜炎双球菌 阴转率

对于四格表资料显著性检验的结果,大多数调研者是依靠α(犯第一类错误的概率)水平来作判断的;而β(犯第二类错误的概率)通常被省略掉了。发生这些现象,除了极少数人不知道β的存在而外,多数人是由于不关心或无从关心所致。本文通过pc-8000型微型电子计算机抽样得到20,000个四格表资料,据此对β值进行初步探讨。

For the results of the significance test of the four-table data, most researchers rely on the level of α (probability of the first type of error) to make judgments; and β (the probability of making the second type of error) is usually omitted. The occurrence of these phenomena, except for a very small number of people who do not know the existence of β, is due to the lack of interest or concern. In this paper, 20,000 data of four grids were obtained by sampling with a pc-8000 microcomputer, and a preliminary discussion was made on the value of β.