Start with the question, not the first page
Before you read a word, write down what you want from the paper: a method you might borrow, a result you need to cite, or simply whether it changes what you believe. A paper read without a question takes twice as long and leaves half as much behind.
Then read the title and the abstract, and stop. Can you say, in one sentence, what they did and what they found? If the abstract will not let you, the paper rarely will either.
Look at the figures before the text
Go straight to the figures and tables, with their captions. They are where the evidence lives; the text is the argument about it. Ask of each figure what is being compared with what, and whether the difference you are being shown looks large, small or within the noise.
If a key result appears only in the text and never in a figure or a table, notice that.
Read the conclusion, then check it against the results
Now read the discussion and the conclusion. Do the claims stay inside what the figures showed, or do they reach further? Watch for the step from “was associated with” in the results to “causes” or “improves” in the conclusion.
Find the limitations paragraph. Authors usually know the weak points of their own study better than anyone; a paper that admits none deserves extra care.
Only then, the methods
If the paper has earned it, read the methods for the parts that decide whether you can trust the result: who or what was studied, how many, what was measured, and what it was compared against. You do not need every detail to judge whether the design can answer the question.
Finally, look at who funded the work and what the authors declare. It does not make a result wrong, but it is part of reading it well.
