Original work with own language vl make the article best but repeated or copy paste in literature vl make it worst.Now a days one software is available in govt institutions we can identify
It depends on how the authors answered the research questions they propagated in their paper. The results, discussion and conclusion or recommendations they wrote should be inline or coherent. Although statistics are usually used but, we shouldn't realy only in statistics, we have to use our "scientific-common sense".
The term you used for judging paper is "scientific evidence". Thus, the focus go to methodology and methods. There are debates between natural science approach and social science approach, between qualitative and quantitative and many other. Depending on the type of discipline, there are expectations of evidences. If you are too ahead of time in novelty, the other would think you are bad. They would just realize much later. Another possibility that you perceived it wrongly. Thus, there is truly no easy you to suggest about this issue. However, if the research question, methodology, analysis and results are coherent (just as the previous scholar commented) the article, with "scientific evidence" should be found good to best.
Firstly, in order to judge how good a research article is, the most important factors are, the relevance fo the research question (how it helps answer part of questions that may assist in the completion of a puzzle) and how the research question is investigated (the methodology) and finally how the investigation is presented
This is a tough question. The best evidence always has also limitations. There are several recommendations for each design and research object . Another important related question is how to make any level of evidence useful in clinical practice? Even if there caveats, good reporting is a strong point.
Take a look at.
http://www.cebm.net/ # oxford evidence based website
http://www.cebm.net/index.aspx?o=5653 # levels of evidence updated
http://www.equator-network.org/home/ # good reporting practice