Weak research design, flawed data analysis, biased observation and poor discussions have encouraged the publication of nonreplicable or false findings in recent years. In the present study, different types of errors and biases and their role in prevalence of wrong data publications are discussed. The main reason for occurrence of wrong and nonreplicable findings is the great tendency for publishing favorable results and ignoring the unwanted ones. As most project sponsors and publication reviewers seek impressive results, it is becoming a routine procedure to either neglect the peculiar results or fix them. Many scientists publish erroneous results due to the lack of knowledge or cognitive biases. This can be avoided if the present science culture undergoes a substantive change. Researchers are usually under pressure to publish more papers, because institutions reward them on the quantity of their publications, this policy needs to change too. Here some methods are recommended to attain a decent science culture.