It depends a lot on which discipline are we talking about - a computer science journal is nothing like a biology one. Having said that, the requirements to pub

It depends a lot on which discipline are we talking about - a computer science journal is nothing like a biology one. Having said that, the requirements to publish in a good journal are quite high - reviewers will point out mistakes, demand more data if yours is not convincing enough, point out related literature that you might have missed, and so on. The journal will also check that the reviewers are not simply friends of the author. It can take about a year to get a journal article published. After seeing some of the stuff that gets submitted (and rejected), I don't feel I can trust research that has not been peer reviewed. A good journal ensures that a minimum standard of quality is met, and that I'm not building my research on lies and fluff pieces.