#Null-hypothesis
I am confused at why so many otherwise very bright people insist that the null hypothesis---things are just breaking because of complexity---must be true. I simply do not believe that we can trust this administration and these companies to say true things in response to very public failures.
November 17, 2026 at 12:16 AM
It's an unserious position because it's unfalsifiable. If the left loses, the answer is for Democrats to moderate. If the left wins, the answer is for Democrats to moderate. There's no null hypothesis and so we can throw the whole thing away
November 12, 2026 at 3:11 PM
You should always assume the null hypothesis (no connection, just a string of bad luck or coincidental errors that the media happens to latch onto) until proven otherwise.
November 16, 2026 at 10:30 PM
Folks just stating a null hypothesis or rebuttable presumption
November 14, 2026 at 2:20 PM
The Null Hypothesis
November 16, 2026 at 3:42 AM
Comparing to a synthetic null hypothesis is useful
November 17, 2026 at 7:01 PM
No, that isn’t what 7 billion people see and it’s pretty weird that you think you can speak on behalf of that many people. As for “no indication they’re wrong”, you might want to look into the phrases “argument from ignorance” and “null hypothesis.”
November 16, 2026 at 8:40 PM
Oooh boy hard agree.

I also think funnel plots are misleading as they are centred on the average reported effect rather than on the null hypothesis being tested.
An analysis of 1,000 meta-analyses across 10 different fields reports that almost all use publication bias detection methods that don't specifically evaluate publication bias and that the most popular method is essentially researchers making a call after eyeballing a plot doi.org/10.1017/rsm....
November 18, 2026 at 8:28 AM
i like this flowchart a lot because Enduring Complementarity is clearly, to me, the null hypothesis, the 'nothing ever happens' of the chart, but i think most people talking about this stuff start at "takeoff severity?" and regard everything else as improbable
everyone read this article & pick your top three in order of preference.

I think mine are:

1. orion's arm
2. enduring complementarity
3. the singleton
November 13, 2026 at 5:34 PM
As the null hypothesis for this was that the Book of Mormon was not plagiarized from Views of the Hebrews, based on this experiment, we must fail to reject the null hypothesis.
November 17, 2026 at 3:43 AM
T: in reality, we don't reject the null hypothesis much
S: wait, so this is a scam?
T: well, that's a good thing, the null hypothesis means there isn't salmonella in your lettuce!
S: ohhhh
November 17, 2026 at 12:56 AM
Just on the basis of these two pieces alone "that guy has absolutely horrific stuff in his digital archive" is the null hypothesis.
November 11, 2026 at 7:01 PM
I mean the null hypothesis is always bots, but mine seem like pretty typical Bluesky humans fwiw
November 18, 2026 at 6:10 AM
„A low p-value means your observed data would be unlikely to happen by random chance alone if the null hypothesis were 100% true.”
Taylor Swift is making a “big announcement” today on Good Morning America.
November 13, 2026 at 8:24 PM
I especially don’t think the null hypothesis is “voters will reward a steady hand on the ship of state.” What about the last fifteen years suggests that?
November 13, 2026 at 12:37 PM
Do we need a null hypothesis more complicated than simple cowardice?
November 10, 2026 at 9:39 PM
To me the random words thing is not mysterious at all. A string of apparently random words seems maximally detectable in principle. This isn’t exactly how Pangram works but conceptually, this prompt has a clean null hypothesis: a uniform distribution over words. Deviations are easy to detect.
November 13, 2026 at 2:26 AM
My group and I (in my experimental projects lab) got our research results back last class

We failed to reject the null hypothesis on one front, but accepted the alternative hypothesis on two others! (We had three hypotheses)

What an interesting class, and I’m so glad I made some friends in there!🩷
November 13, 2026 at 12:20 PM
True, uniform is not quite the right null hypothesis but I think the overall point stands
November 13, 2026 at 3:05 AM
This goes back at least to the 70s and, along with being left for a trans person, is absolutely the null hypothesis for political transphobia.
you cannot overstate how much of the past several decades of transphobia resulted from cis people not getting intimate access to a trans person's body and deciding to strip our autonomy so we could never say "no" again
November 10, 2026 at 8:05 PM
The null hypothesis is to forestall scrutiny of his own similar behavior... like his multiple known affairs with subordinates.
November 15, 2026 at 12:58 PM
I finally understand you nerds and your null hypothesis jokes
November 11, 2026 at 6:41 PM
The null hypothesis here is voters want change imo. They’re not going to mind change to senate procedure bc they have affirmatively have voted for change for fifteen years
November 13, 2026 at 12:47 PM
this is not a great direction for him but he clearly understands the hypothesis is null.
November 13, 2026 at 5:34 PM