how many scientists fabricate and falsify research?

Well, this illuminating study by Daniele Fanelli suggests that quite a few (without giving us any exact numbers of course). It all depends on how you formulate your question. For instance,

scientists were less likely to reply affirmatively to questions using the words “fabrication” and “falsification” rather than “alteration” or “modification”. Moreover, three surveys found that scientists admitted more frequently to have “modified” or “altered” research to “improve the outcome” than to have reported results they “knew to be untrue”.

That does not surprise me at all, but it’s good to have something like that published in peer-reviewed journal. Speaking of peers,

The grey area between licit, questionable, and fraudulent practices is fertile ground for the “Mohammed Ali effect”, in which people perceive themselves as more honest than their peers.

The decrease in admission rates observed over the years in self-reports but not in non-self-reports could be explained by a combination of the Mohammed Ali effect and social expectations. The level and quality of research and training in scientific integrity has expanded in the last decades, raising awareness among scientists and the public. However, there is little evidence that researchers trained in recognizing and dealing with scientific misconduct have a lower propensity to commit it. Therefore, these trends might suggest that scientists are no less likely to commit misconduct or to report what they see their colleagues doing, but have become less likely to admit it for themselves.

And now, from the past to the future (misconduct):

There seems to be a large discrepancy between what researchers are willing to do and what they admit in a survey. In a sample of postdoctoral fellows at the University of California San Francisco, USA, only 3.4% said they had modified data in the past, but 17% said they were “willing to select or omit data to improve their results”. Among research trainees in biomedical sciences at the University of California San Diego, 4.9% said they had modified research results in the past, but 81% were “willing to select, omit or fabricate data to win a grant or publish a paper”.

Now, really difficult question. Are San Diego guys more fraudulent than their San Francisco colleagues? Or more honest because thay admit being more dishonest?