As I am developing my research proposal, I encounter the problem of finding relevant literature that treats sentiment as an indicator of people's attitude/perception. Is it then theoretically reasonable? If not, what studies have argued this position? Please help me!