If what you mean by studying cheaters is "interview" them, be careful they might cheat you of their answers. But yes, whether they're cheaters or presumably honest individuals, you still need to secure their consent to be part of your study.

you need consent for them to take part, there is a rich literature about what amount of deception or 'ambiguity' is allowed

Cheating is a very context specific and cultural thing. For instance manufacturing a patented medicine in the US/Europe which gets manufactured in India, South Africa, Brazil, Thailand without consent or following the Global TRIPS regime gets considered as cheating or forgery( See Texmati case, haldi powder patent). In India, Nepal and many Asian, African countries there was no notion of patent copyright etc, but in the West it has been existing for hundreds of years.

Now for instance I want to know how the same doctor treats the patients who works both in the government and private sector, taking his consent will destroy the whole purpose of the study. Consents were primarily aimed at clinical trials and any bodily interventions and later shifted to all aspects of research. I can go on and on,,,,, but I hope this much is useful.