See first, think later, then test. 

But always see first. 

Otherwise, you will only see 

what you were expecting. 

Most scientists forget that.