[英]R rmcorr incorrect degrees of freedom number
When I analyse raz2005
data from mrcorr
package for repeated measures correlation with the following code:当我分析来自mrcorr
包的raz2005
数据以与以下代码相关联时:
brainvolage.rmc <- rmcorr (participant = Participant, measure1 = Age, measure2 = Volume, dataset = raz2005)
as given in the paper https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5383908/pdf/fpsyg-08-00456.pdf , I get df=71, which is correct since the number of participants is 72 in this data.正如论文https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5383908/pdf/fpsyg-08-00456.pdf 中给出的,我得到 df=71,这是正确的,因为参与者的数量是该数据为 72。
However, when I use the same code for my data, which is structured exactly like raz2005
, I get df=more than 400, although my participant number is 25. I use:然而,当我对我的数据使用相同的代码时,它的结构与raz2005
完全一样,我得到 df=more than 400,虽然我的参与者人数是 25。我使用:
mydata.rmc <- rmcorr (participant = participant, measure1 = measure1, measure2 = measure2, dataset = mydata)
From the paper (page 7), the degrees of freedom should be calculated like this:从论文(第 7 页)中,自由度应该这样计算:
N = n_distinct(raz2005$participant)
k = nrow(raz2005)/N
df = N*(k-1)-1
声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.