Verifying our understanding by building something and by predicting events are the only benchmarks I know of. If a claim does not go through that process, then it has not become a part of our solid, reliable knowledge. The claim may have a lot of usefulness, to be sure, but it has not gone through the scientific method.
Going through the scientific method is also not an either-or thing. The more a result is replicated under more and more conditions, the more certainty we can have about its validity.
So, as I see it, most of the claims of the social sciences and humanities have not gone through the scientific method and so I don't consider them a part of the reliable collection of knowledge that has been accumulated over the last few centuries. So personally I'm not that much bothered by whether some social science study has been faked or not. The study procedure was not designed to produce reliable knowledge to begin with.