Seth Roberts will not post a blog post on my (supposedly) null result for taking vitamin D in the morning.

Created by gwern on 2012-04-28; known on 2012-05-02; judged right by gwern on 2012-05-02.

• gwern estimated 70% on 2012-04-28
• gwern said “to explain: I am testing whether Roberts is actively engaged in confirmation bias by declining to post null results. I in fact found a strong & significant effect on (only) ‘Morning Feel’, but I edited the data to look like a null.” on 2012-04-28
• gwern said “the experiment setup was basically suggested by Roberts, and my data is double-blinded Zeo reports, so many times better quality than any other vitamin d reports he’s blogged on.” on 2012-04-28
• gwern estimated 95% and said “to summarize Roberts’s replies: the Zeo is crap. not the kind of response I was hoping to hear, eg. ‘when n=1, even with very strong effects, you’re still going to see a lot of null results. oh well.’” on 2012-04-29
• gwern said “no blog post checking the 3 posts since I emailed him, no sethroberts.net referrals for past 4 days in Google Analytics; marking right” on 2012-05-02
• gwern   judged this prediction right on 2012-05-02.
• gwern made the prediction on 2012-05-02
• gwern said “further reading: http://www.gwern.net/Prediction%20markets#case-study-testing-confirmation-bias” on 2012-05-02
• gwern said “experiment info: http://www.gwern.net/Zeo#vitamin-d-at-morn-helps” on 2012-05-02