Should I transform my raw data into logarithms before multilevel analysis?

是 否 +2 论坛币

k人 参与回答

经管之家送您一份

应届毕业生专属福利!

求职就业群

赵安豆老师微信：zhaoandou666

经管之家联合CDA

送您一个全额奖学金名额~ !

立即领取

感谢您参与论坛问题回答

经管之家送您两个论坛币！

+2 论坛币

I was ask to do a linear regression analyses of my variables. I have Reaction times. Can you recommend me a linear regression analysis (hierarchical?) and tell me whether I should transform my raw data into logarithms before that analysis?

My reaction times are in milliseconds.

扫码加我 拉你入群

请注明：姓名-公司-职位

以便审核进群资格，未注明则拒绝

分享0 收藏0 回帖

关键词：Multilevel transform Analysis Analysi Should recommend whether before should

Usually, log transformation is performed on Y to make its distribution look more like Gaussian. If your Y already has a nearly gaussian distribution, the transformation is not necessary. For the 'hierarchical' part, you need to distinct between different levels (see the classic students within schools example, or Goldstein good and free book: http://www.arnoldpublishers.com/support/goldstein.htm.)