According to the USA Office of Research Integrity, misconduct in science means fabrication, falsification, plagiarism, or other deliberate misrepresentation in proposing that seriously deviate from those that are commonly accepted within the conducting, or reporting research. It does not include honest error or honest differences in interpretations or judgement of data. The difference between scientific misconduct and professional misconduct is sometimes slight and the subject of debate. Clearly scientific misconduct is behaviour that directly affect the integrity of research -- embodied in the three great scientific sins: falsification, fabrication and plagiarism; whereas there are other things involved with unprofessional behaviour as a scientist like non-collegiality, sloppy notebooks, poor laboratory supervision, nastiness, not sharing resources, not helping younger colleagues, and indifference to the acknowledging promptly the contributions of others and to the sharing of research materials of importance to others.
Dishonest scientists can mislead other researchers, stain their institutions' reputations, waste taxpayers' money, and sometimes even endanger the public. Occasionally researchers may falsify data in an effort to make their studies more exciting or plausible. Compounding this problem (which, in addition to publication stress, also has financial roots – large medical school debts to repay) is the reticence with which fraudulent practices are brought to public view. Scientists who suspect colleagues of fraud often turn a blind eye as they fear tarnishing their institution's standing, being blackballed by colleagues, and possible defamation lawsuits brought by the accused.
Fraudulent authorship is often the product of the highly competitive world of science. Researchers under great pressure to publish papers may ultimately resort to plagiarism and data faking. Grants and promotions are largely based upon publication productivity, thus it is an accepted practice for senior researchers to co-sign subordinates' studies, even though they may not know much about the data. This co-signing helps younger researchers to publish their research and boosts the publication numbers of their bosses. In addition to flagrant fraud, the practice of co-signing when one has done no research on the study involved, is another form of fraud. Senior researchers should be held accountable for any publications they sign their names to.
Scientific fraud is not limited to modern or minor figures: the verdict of history is that Ptolemy, Galileo, Newton, Dalton, Cyril Burt and Gregor Mendel all tampered with some of their data. Examples frequently cited of pathological science include N-rays, mitogenetic rays, the Allison effect, Piltdown man, Piri Reis map, extrasensory perception, and flying saucers.
A Yale University scientist was fired in 1980 for plagiarizing and falsifying his research work. Another institute found a researcher had faked evidence to show that human skin can be made transplantable without rejection (he was given a one year full-salaried 'sick leave'). A notable Harvard Medical School case involved the discovery of fraud, but the culprit was allowed to stay and continue to cheat.
In 1993, a key US AIDS researcher was found guilty of scientific misconduct for failing to reveal that his laboratory had grown a French version of the virus that causes AIDS (LAV), and in so doing impeded potential AIDS-research progress with LAV.
It has been alleged that initial reports on evidence of the polar ozone hole were discredited by undertaking research deliberately designed to be unable to detect the reported phenomenon.
Practitioners of any scientific discipline are aware of many prominent people in that discipline who have risen to eminence by faking data. They are protected only by the law of libel and by the fact that universities are reluctant to having made mistaken appointments. It has been estimated that possibly 25% of scientific papers have some (usually minor) discrepancy in relation to the original test results. In most cases these do not merit any form of rectification, but in a few cases the discrepancies are crucial.
Given the extreme pressures to publish in order to prove one's validity and in order to earn enough money to repay education loans, it is no wonder that scientific fraud is becoming more prevalent. Until some of the pressures are alleviated (for example, requiring scientists to cite only 3 publications a year when applying for funding, and lengthening the terms of federal research grants), this practice will not be curtailed.