Hope you all are fine. I just want to ask a question about cross validation "leave one subject out". I have confusion about that. e.g. I have 10 subjects and I took 9 subjects data as training and last subject for testing and repeat it 9 times. It means, I have to do second level analysis in SPM.  

In 10 cross validation we divided data into training and testing data of every subject individually. There is no cross validation between subjects in 10 cross validation. I want to ask that 10 cross validation is better or leave one subject out.

Please guide me whether I am right or wrong regarding cross validation  "leave one subject out" and "10 cross validation".

Thanks in advance Madam,

Regards