文献检索，用中文搜 PubMed

Suppr 超能文献

核心技术专利：CN118964589B侵权必究

核心技术专利：CN118964589B侵权必究

Comp Biochem Physiol A Comp Physiol. 1987;87(4):979-82. doi: 10.1016/0300-9629(87)90023-5.

Three tests of statistical significance: the confidence interval (C), Student's t, and Satterthwaite's corrected t were compared using a computer generated sampling experiment. 2. The C-test was shown to be extremely inefficient at detecting differences between pairs of means. The t-test performed as expected and Satterthwaite's correct t was slightly conservative. 3. When the population variances are different Satterthwaite's corrected t performed extremely well, the t-test was slightly liberal, while the C-test remained extremely insensitive. 4. It is concluded that the C-test should not be used (i.e. against Scheer, 1986, Comp. Biochem. Physiol. 83A, 405-408). 5. It is noted that when many t-tests are performed on one data set, alternative methods are appropriate.

进行了三项统计显著性检验：使用计算机生成的抽样实验比较了置信区间（C）、学生t检验和萨特思韦特校正t检验。2. 结果表明，C检验在检测均值对之间的差异时效率极低。t检验表现符合预期，萨特思韦特校正t检验略显保守。3. 当总体方差不同时，萨特思韦特校正t检验表现极佳，t检验略显宽松，而C检验仍然极度不敏感。4. 得出的结论是不应使用C检验（即与舍尔，1986年，《比较生物化学与生理学》83A，405 - 408相悖）。5. 注意到当对一个数据集进行多次t检验时，采用替代方法是合适的。

Comp Biochem Physiol A Comp Physiol. 1987;87(4):979-82. doi: 10.1016/0300-9629(87)90023-5.

Three tests of statistical significance: the confidence interval (C), Student's t, and Satterthwaite's corrected t were compared using a computer generated sampling experiment. 2. The C-test was shown to be extremely inefficient at detecting differences between pairs of means. The t-test performed as expected and Satterthwaite's correct t was slightly conservative. 3. When the population variances are different Satterthwaite's corrected t performed extremely well, the t-test was slightly liberal, while the C-test remained extremely insensitive. 4. It is concluded that the C-test should not be used (i.e. against Scheer, 1986, Comp. Biochem. Physiol. 83A, 405-408). 5. It is noted that when many t-tests are performed on one data set, alternative methods are appropriate.

进行了三项统计显著性检验：使用计算机生成的抽样实验比较了置信区间（C）、学生t检验和萨特思韦特校正t检验。2. 结果表明，C检验在检测均值对之间的差异时效率极低。t检验表现符合预期，萨特思韦特校正t检验略显保守。3. 当总体方差不同时，萨特思韦特校正t检验表现极佳，t检验略显宽松，而C检验仍然极度不敏感。4. 得出的结论是不应使用C检验（即与舍尔，1986年，《比较生物化学与生理学》83A，405 - 408相悖）。5. 注意到当对一个数据集进行多次t检验时，采用替代方法是合适的。