It's been called the "decline effect," "the proteus phenomenon," and "the reproducibility crisis": the startling realization that a lot of seemingly solid scientific research doesn't pan out under repeated testing. The latest blow to scientific confidence comes from the Reproducibility Project, which attempted to replicate 100 published psychology studies and found that, when the experiments were repeated, half or more failed to uphold the original findings. So is it time to start doubting the credibility of research in general? Stanford University psychologist and Reproducibility Project participant Mike Frank joined us to explain what the results really mean, misconceptions about statistical rigor in science, the various ways experimenters blunder and sometimes delude themselves, and the gradual, cumulative nature of scientific progress.