Hi guys, so I've been researching sleep studies and I've been looking at comparison on devices used, and most of them seem to obtain a magnitude value based on the acceleration seen on the device by the user. What I'm wondering about though is that the comparison between devices seem to be using the Pearson Correlation. Example: http://www.journalsleep.org/Articles/290214.pdf http://www.sciencedirect.com/science/article/pii/S0165027000003642 http://www.itc-ltd.co.jp/pdf/Kushida_et_al.pdf But from what I read, the Pearson correlation only works with linear correlations, while I would assume the values based here would be quite random values, and while I would imagine there is a correlation say between the devices, I wouldn't imagine it to be linear which I thought was what the Pearson correlation needed? Am I wrong? I will admit, I don't understand the correlation when I tried to read up on it.