24小时热门版块排行榜    

查看: 476  |  回复: 2

cpu_luzg

金虫 (小有名气)

[求助] 求助统计学问题

前段时间投了一篇文章,其中一个审稿人给了这样一个问题:“The use of the Students t-test for comparisons between a control and multiple treatments is not appropriate unless conducted using a family level of confidence.  The chance of a Type 1 error for alpha=5% and comparing the control to 5 exposure concentrations is a 23%.  A parametric ANOVA approach should be use if all assumptions can be properly met.  I would suggest the researchers seek help from a qualified statistician.” 不知道怎么回答审稿人?求哪位虫友能够相助!
回复此楼

» 猜你喜欢

» 本主题相关价值贴推荐,对您同样有帮助:

已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖

feixiaolin

荣誉版主 (文坛精英)

优秀版主

用[学生氏]t检验比较对照组和综合组并不合适,除非采用家族水平的对比。当α= 5%时与对照组5人染毒相比(?我猜的),发生第一类1类错误的机会是23%.  如果所有的假设都能很好地满足,应采用参数化方差分析方法(ANOVA)。我建议研究人员向一个有水平的统计学家寻求帮助。

你的方差分析和数据解释有问题。
2楼2014-03-03 22:03:17
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖

13874995200

金虫 (初入文坛)

★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★
cpu_luzg(feixiaolin代发): 金币+25 2014-03-05 23:20:30
Response: Thanks a lot for your professional advices. Based on your suggestion, we asked statisticians for help. one-way ANOVA, followed by a post hoc Newman–Keuls multiple comparison test, was used to detect statistical significance of results. Although Student's t-test is mathematically identical to a one-way ANOVA done on data, it is limited to just two categories of data. The information has been revised in the manuscript as: ......                                                                              希望对你有用,祝你好运!
3楼2014-03-05 21:47:36
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖
相关版块跳转 我要订阅楼主 cpu_luzg 的主题更新
信息提示
请填处理意见