You don’t have to be great to start, but you have to start to be great.
在观察性研究中,我们常常不得不依赖于无法直接检验的假设。