How to publish a scientific paper
Once a scientist has written a paper about his research (see 'How a science paper is written'), he then has to get it published. This is not as easy as it may sound and can involve a lot of work. First he will select a journal that is suitable for the subject he has been studying. So a paper about dinosaur evolution could be sent to the Journal of Evolution, or Palaeontology, or the Journal of Vertebrate Paleontology.
Once it has reached the journal, the editor will read it. If he thinks it is interesting and is suitable for the journal then it will be sent out for review. The review process is one of the most import features of scientific study, something often overlooked. It is there to make sure that no one published false or misleading results as well as to check the suitability of the analyses used and the conclusions drawn by the researchers.
The reviews are carried out by at least two experts in the filed that the researchers have written about. The reviewers are also active researchers and so are familiar with the existing work and methods used in this area. They will read the paper very carefully and check everything about it. Are the any major mistakes in the introduction about the current state of the subject, has the data been collected properly with appropriate methods, are the statistics correct, is their discussion justified? Reviewers will check all of this in detail and report any mistakes to the journal's editor.
If there are too many mistakes, or some serious problems with the methods and results then the editor will reject the paper and ask the researchers to start again. However, if the mistakes are not too serious, then he will send the reviewers' comments back to the researchers and ask them to do some more work. They might need to collect some more data, or look at some other papers on the subject, or just use some extra statistics to back up their arguments.
Once this extra work has been done the revised paper can go back to the editor. If he and the reviewers are satisfied with the changes then it can be made ready for publication and will eventually appear in the journal. All of this takes time though. Reviews often take several months if the reviewers are busy, and its not uncommon for a paper to appear in print 2 years after it was first sent to the editor and some can take even longer.
However, what all this really means is that science and scientists are very good and checking their facts and making sure all their results are correct. Everyone contributes to check everyone else's work and to ensure that the rules are obeyed and that the work is done properly.