There has been much talk of late about a "replication crisis" in psychology. Important experimental results, sometimes results that have generated considerable bodies of literature, peer-reviewed and otherwise, have turned out to have been resting on sand when subjected to rigorous efforts at replication. One notorious example involves the "power pose" (standing, feet apart, hands on hips, chin up). Striking the power pose was supposedly proven to have a range of valuable psychological effects, including the reduction of stress. Except that it doesn't. The effect was the consequence of data mining and statistical noise.
Such failed replication efforts have reopened talk about whether psychology deserves the title "science" at all.
Fortunately, for those of us who will to believe that there IS such a science, some important results have survived efforts at replication. This is especially so it appears in the study of cognition, a very Jamesian branch of psychology.
Nine findings in particular have been found to stand on a much sounder-than-sand empirical foundation. (There's a link to a list.)
The most fascinating of these, to me, involves "motor priming" and subliminal perception (third on the list). But I'll let you decided of the respective degree of fascination for yourself dear reader.
In the event you're wondering: I have illustrated this post with the visage of William Wundt, one of the pioneers of experimental psychology.