One of the necessary steps involved in bringing a new drug to market is an investigation into its effectiveness and possible harmful side effects. This research is performed in two basic stages: animal studies and the clinical stage of experimentation involving human subjects. Preclinical drug testing is based on a number of assumptions: (1) if a drug is safe for two other species it is also safe for humans, and (2) humans are approximately as sensitive to the drug being tested as the animal species used in the test. The weaknesses of these assumptions means that even the most careful drug testing is far from perfect. There are a number of ways of obtaining favorable results in the preclinical safety testing of drugs without outright falsification. These include using too few animals in the experiment so that no statistically significant result appears, using modern microsurgery to alter results, and the manner of selection of the control group. Bibliography