Replication of results

Replication of results, in experimental science, is essential for scientific progress. If a scientist makes a claim, and writes it up in the proper format for publication, it may pass peer review, but this does not make it valid. It might get some media play, but it's not going to be added to textbooks on the high school or university level unless other scientists are convinced.

What is most convincing to a scientists is seeing for himself. If it's an experiment which can be performed again, other scientists can satisfy their curiosity by doing the experiment over themselves. If they get the same results, that's a confidence booster. If their results are sufficiently different to shed doubt on the the original scientific paper, then we get a scientific dispute.