Ep 246: Why Replications Sometimes Don’t Agree with the Original Study

MichaelResearch and Stats2 Comments

Replications in Research


What’s the connection between life on Mars and the need to properly replicate scientific studies? Well, replicating a study one time and not finding the results the original researchers find is NOT proof that the original study was a fraud. Let’s be clear about what replication really means. Here’s a bunch of examples. This topic ties into the so-called “woman on mars” that so many people are seeing. I would love it if we had solid proof that there was life on Mars. So far we have “pretty darn good” evidence that the conditions were right for life on Mars.



Excellent video on the topic of reproducibility in science

Related Posts Plugin for WordPress, Blogger...

2 Comments on “Ep 246: Why Replications Sometimes Don’t Agree with the Original Study”

  1. Thanks Uni for the link to this video. It’s excellent. I’ll add it to the notes for this episode. Highly recommended.

  2. Things might be much worse regarding biases in research, and I strongly recommend research done by John Ioannidis (from Standford Univ.). He mainly concentrated on Medicine, as he is a professor of Medicine, but it seems that the conclusions and points that he makes are valid for Psychology and Social research as well.
    There are some lectures by him on Youtube. See, e.g. https://www.youtube.com/watch?v=KEAAHJDNCMQ

Leave a Reply

Your email address will not be published.