A couple months ago the results of a study published in the journal Nature showed that about 1 of 3 scientists doing research admitted to poor ethics, which typically means they fudged data at some point in their studies. The study surveyed over one thousand researchers.
This is disturbing. Science has as a goal to find the truths in Nature. When scientists submit articles for publications in the top journals of their respective fields, it is assumed the primary research was completed with honor and respect for the truth, no matter if what was measured and observed fits in with preconceived ideas/attitudes/beliefs or not. While articles submitted to the major journals are peer-reviewed and go through a series of re-edits, it is impossible for the reviewers to be one hundred percent sure of whether an honest effort was made or not. If a dishonest paper makes it through the process, it may be years before that dishonesty is discovered, most likely through other independent checks of the experiments that lead to opposing or conflicting results and/or conclusions.
One trend that has developed in fields such as biochemistry, genetics, biomedical engineering, and so on, is large numbers of researchers going off and developing private companies that try to develop a new drug, procedure, or technology for specific needs, in addition to their university research. Private enterprise springs the ultimate motive to slightly tweak data or develop overly optmistic conclusions - of course, I mean monetary profit. In fact, when I was talking with acquaintances who are in the administration at the medical research facility of a major university hospital, I was shocked by the lack of collaboration and sharing of information on cancer research between their labs and another nearby major university facility. The reason given was largely because of the race to get patents and private funding from pharmaceutical and biotechnology firms. With potential billions of dollars in such areas of research for the next wonderdrug or life-saving procedure, of course some may weaken under the stress to win the race and miss something in the data or be a tad sloppy with analysis, or feel the need to make results just a tad better than they really are in order to get a piece of the money pie. Such actions not only go against the science code of ethics, but also can conceivably delay the process that can mean unnecessary deaths or progress in an entire field. It is likely that real collaboration and sharing of knowledge between the two university facilities could lead to breakthroughs in a more timely manner than the two working in secrecy, but the need for money and glory get in the way, and weaken all of science in the process when unethical behavior enters the fray.