Publication in peer reviewed media is a necessary step as a standard of quality for science, but it is not sufficient. Everybody reading any publication must also check those standards of research excellence such as full description of hypotheses, availability of study protocol (e.g. description of sampling methods, full description of data and assumptions used), adherence to accepted statistical standards, and even availability on demand of data/models/econometric routines or programs for replicability purposes. These are in my mind the golden standards for peer reviewed publications. It all boils down to transparency.
There have been articles published in high caliber journals which have been later withdrawn due to improper research methods or even in some cases researcher impropriety. Furthermore, there is another development which requires even more vigilant efforts to examine evidence presented to an audience and that is the resurgence of publishers and journals of questionable quality and/or standards for publication. The blog
Scholarly Open Access contains a list of publishers and individual journal which are of a questionable status and quality. You can peruse their lists and make your own opinion about these publications.