24小时热门版块排行榜

返回列表

当前只显示满足指定条件的回帖，点击这里查看本话题的所有回帖

吉吉爱100

金虫 (小有名气)

应助: 0 (幼儿园)
金币: 485.5
散金: 9
红花: 4
帖子: 64
在线: 169.5小时
虫号: 1344046
注册: 2011-07-12
性别: MM
专业: 种群生态学

[求助] 关于选择什么统计分析比较好

各位大师，求帮助~~！
最近的一个文章返回，审稿人的意见是我的ANOVA用的不合适。我的原始意思是有三组不同类型的湖泊，然后我想看这三组类型的湖泊中，里面的环境变量（a1，a2,a3) (b1,b2,b3)、(c1,c2,c3)在各组之间是否是否有显著性差异。我弄完分析之后，编辑的回复是：
The use of ANOVA tests, as has been done, does not seem right to me: the lakes appear to differ quite dramatically in the key variables used to group them (i.e. fish yield). The groups should be made based on the variation of the data, for instance, by ordination techniques that clearly identified the importance or association of the different characteristics (fish yield, cyanobacteria biomass and TP) with each lake. The belonging of each lake to each particular group should then be tested, if ANOVA was going to be applied. Besides, there is an overlap between fish yield and nutrient concentrations (e.g. type 2 lakes), which makes the use of ANOVA even more complicated since the two factors cannot be fully disentangled. A variance partitioning technique should be used here to determine how much of the variance in zooplankton size depends on fish and how much on cyanobacteria, for instance. However, the application of regression techniques, as done afterwards, seems more appropriate.

不知道现在做什么样的分析比较合适，请教各位大师了

回复此楼