One of the central tenets of the scientific method is that others must be able to reproduce an experiment and achieve the same results. For too many years, no one was performing this vital task.
Now Science has reported on the results of the Reproducibility Project. What they learned was sobering: 64% of the peer-reviewed results could not be reproduced even when working the authors of the original studies! The Project's results are here.
I can hardly imagine what the statistics would be for climate 'science.'