mark-ER said:
That seems to be one of the most important and least 'taught' portions of our scientific careers. I can kinda tell what paper is crap, which paper is a classic and ones somewhere in between, but so far it's mostly by gestalt. I know a few criteria to judge by (is it physiological, do the authors understate/overstate their conclusions, were there appropriate controls), but I would appreciate some comments on this issue. Could someone add some criteria by which you judge the quality of any given publication? It doesn't have to be a thorough explanation (if it is, great; perhaps even a link to a website that has a commentary on the topic). Thanks
I think you have already hit some of the major criteria.
I would add:
Research question of significance
Acknowledgment of previous work in the field
Appropriate experiments to answer the questions
Good experimental design
Good statistical power
Sufficient description of methods to allow others to replicate results
Clear, logical presentation of results
Well-labeled and clear figures
Presentation of right amount of data (not too much or too little)
Conclusions are supported by the data
Consideration and explanation of why other interpretations are less likely
Discussion of implications of work, future directions
And to a lesser degree:
How well-written and clear the paper is (style, grammar, omissions, etc)
Impact factor/type of journal
Laboratory that the paper comes from
