24小时热门版块排行榜    

查看: 1052  |  回复: 7
【悬赏金币】回答本帖问题,作者Clytze9将赠送您 14 个金币

Clytze9

金虫 (小有名气)

[求助] 关于投稿之后要求填写的checklist中的统计学问题 已有1人参与

1. Sample size
Describe how sample size was determined.
(Describe the statistical methods that were used to predetermined sample size or if no sample-size calculation was performed, describe how sample sizes were chosen and provide a rational for why these sample sizes are sufficient)

2. 如何说明文章中用的统计学方法比较合理?

有没有统计学方面的大神?关于sample size 和统计学方法,有没有推荐的文献什么的?

谢谢啦!
回复此楼

» 猜你喜欢

已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖
2楼2017-08-25 10:52:13
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖

ipublish_SCI

捐助贵宾 (正式写手)


【答案】应助回帖

感谢参与,应助指数 +1
Yes, there are. But we do not know about your experiments. But if you did not use statistical analysis to know sample size BEFORE conducting the experiments, then you cannot do it A POSTERIORI. What you can do is to explain you decision about the sample size: that is , why?
3楼2017-08-25 11:02:33
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖

holypower

至尊木虫 (知名作家)


Clytze9(liouzhan654代发): 金币+1, 感谢交流 2017-08-25 13:36:57
根据80%统计检验力,0.05显著水平,以及估计的效应量(比如中等0.25),就可以算出至少多少样本量。你搜索G*POWER程序吧!输入参数自动计算

发自小木虫Android客户端

» 本帖已获得的红花(最新10朵)

4楼2017-08-25 12:37:37
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖

Clytze9

金虫 (小有名气)

引用回帖:
4楼: Originally posted by holypower at 2017-08-25 12:37:37
根据80%统计检验力,0.05显著水平,以及估计的效应量(比如中等0.25),就可以算出至少多少样本量。你搜索G*POWER程序吧!输入参数自动计算

亲 有G Power安装包嘛?
5楼2017-08-25 13:51:52
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖

Clytze9

金虫 (小有名气)

引用回帖:
3楼: Originally posted by ipublish_SCI at 2017-08-25 11:02:33
Yes, there are. But we do not know about your experiments. But if you did not use statistical analysis to know sample size BEFORE conducting the experiments, then you cannot do it A POSTERIORI. What  ...

So you mean we need to do the pre-experiment to know the sample sizes before conduting the formal experiment? But how ? Do you have any sofware to calculate the sample sizes or do you have any good literatures about the statistical analysis?
6楼2017-08-25 13:55:55
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖

Clytze9

金虫 (小有名气)

送红花一朵
引用回帖:
4楼: Originally posted by holypower at 2017-08-25 12:37:37
根据80%统计检验力,0.05显著水平,以及估计的效应量(比如中等0.25),就可以算出至少多少样本量。你搜索G*POWER程序吧!输入参数自动计算

下载了 待我慢慢研究下 谢谢啦
7楼2017-08-25 14:37:32
已阅   回复此楼   关注TA 给TA发消息 送TA红花 TA的回帖

ipublish_SCI

捐助贵宾 (正式写手)


【答案】应助回帖

引用回帖:
6楼: Originally posted by Clytze9 at 2017-08-25 13:55:55
So you mean we need to do the pre-experiment to know the sample sizes before conduting the formal experiment? But how ? Do you have any sofware to calculate the sample sizes or do you have any good  ...

I do not know what your experiments are about. I do not know if they refer to dimensions of samples or if this is an animal or biomedical study and you needed to know how many animals you will need to use. So yes, you may need to calculate statistical power. I again should tell you that if you use now g*power (after you conducted the experiment) to find out that your sample size should have been, then it's a little bit of a problem. If you need further comments on this you can contact me here by message in the forum. Hope this comment helps for you. Best regards
8楼2017-08-25 23:14:53
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 Clytze9 的主题更新
不应助 确定回帖应助 (注意:应助才可能被奖励,但不允许灌水,必须填写15个字符以上)
信息提示
请填处理意见