First, fraud does happen in the sciences. In general, though, it is discovered and corrected by other scientists attempting to replicate the results. There are situations where this process fails: in very large projects (like CERN) where there may be hundreds of names on a single paper, and areas where the research is unlikely to have people attempt replication.<quoted text>
Can we trust science?
"Cases of scientific misconduct tend to hit the headlines precisely because scientists are supposed to occupy a moral high ground when it comes to the search for truth about nature. The scientific method developed as a way to weed out human bias. But scientists, like anyone else, can be prone to bias in their bid for a place in the history books."
Now, for the large projects, the solution is to have more than one team working on different data sets and which are competing with each other. This is very common. While not perfect, it does bring out the biases and tends to correct them over time.
The lack of replication is a much, much more severe problem. This strikes at the core of what it means to be a science. The article you posted noted that replication tends not to happen in psychology and that, I would argue, places psychology outside the bounds of real science. Replication of the results of others is absolutely critical for doing real science. And, truthfully, psychology and sociology are at best borderline sciences. At worst, they are hockum.
Medical science is another place where I would suggest that changes need to happen. it is common for drug trials to have a built-in error rate of 5%. That is way, way too high and corresponds to a 2.5-sigma signal. In particle physics, we don't consider anything less than 5 sigma to be a 'real' signal. The difference is huge and leads to many false positives in medicine, even for those doing the science correctly.
Generally, it is a good idea to take anything said by a sociologist or psychologist with a grain of salt. Ask for multiple studies testing the ideas before belief. This is automatic in most area of real science. If you don't see it in some area, they are not really scientists.