Sequential Testing for Determining Cat Chirality | Hacker News Reader

Sequential Testing for Determining Cat Chirality

blog.dimview.org

106 pointsgwern30 comments

The end result: sequential testing terminated without accepting either hypothesis http://blog.dimview.org/math/2017/10/15/cat-chirality.html

Hmm, there should be three hypotheses: cw, ccw, and no-preference.

No-preference is the null hypothesis.

Or a forth, subjects become aware of data collection and purposefully skew the data because lets face it, cats are jerks.

H0 is no preference, H1 is preference for either clockwise or counterclockwise. In this test with N=1, there was not enough evidence to reject H0.

I’m betting there may be preferences for individual cats, but not symbolic of the entire cat population. If anything, cats are almost pathological about generalizations. If you assume to know what cat will want to do then cat just might detect your assumption and decide to change. I’ve seen them do this too much to discount it.

It's because an observer collapses the wavefunction. Everybody thinks that Schrödinger was being metaphorical -- but no, he was actually talking about cats.

Crowdsource it. Run the exact same experiment on more cats.

Ok, I know this is a little bit silly but on the similar subject... but has anyone noticed pomeranian dogs tend to circle clockwise instead of counter-clockwise? My friend's dog does it so much we nicknamed her "The Anti-NASCAR".

Research of this importance is deserving of an ig-Nobel prize.

We need an AI chirality analyser from webcam data

We now need to determine the spin direction of buttered toast.

That field met its Einstein with Chris Smith's epochal 2013 work.

Google "chris smith manchester toast" and persevere longer than I did to find the paper the newspaper stories are talking about.

The spin direction is such that it will land on the buttered side if let fall.

I think that these observations are important. Imagine if some cats did have a preference for sleeping on one side rather than the other, 'clockwise' rather than 'anti-clockwise'. Why would that be, and why would nobody have noticed until now?

There must be important evolutionary benefits to sleeping on each side an equal amount, more study is definitely needed.

I'm reminded of the question of whether cattle prefer to stand aligned north-south:

https://www.nature.com/news/the-mystery-of-the-magnetic-cows...

The unfortunate thing is that to get complete data, you'd need full web cam coverage of your home in order to catch all instances of your cat sleeping over 24 hours. Otherwise your data risks being incomplete, or worse, skewed. The level of rigor required necessitates that only the most dedicated cat owners can study this phenomena.

I initially thought this has something to do with the concept of chirality in lifeforms...

https://en.wikipedia.org/wiki/Chiral_life_concept

You might like the book Axiomatic. It's a collection of short stories by Greg Egan, and one of them has a plot involving Chiral humans.

cc and ccw are not the only cat sleeping positions. I've seen some cats sleeping on their backs, and others sleep in a sphynx-like position.

I believe that the technical term for that last one is "loaf."

Rabbit parent, not cat parent here, but I hear "loaf" used to refer specifically to the pose in which the front feet are curled underneath the body, and thus hidden: http://i.imgur.com/Y2bBtBy.jpg

Whereas the Sphinx has its feet extended in front of itself: https://upload.wikimedia.org/wikipedia/commons/f/f6/Great_Sp... (which is indeed another rabbit sleeping pose; presumably cats do as well)

I've heard the "loaf" position called the "lozenge" (my fat barely-a-mouser does that a lot).

I dont think that I've ever seen cat really sleeping in one of these positions, they usually change position after they are really deep asleep and are then either curled up or in some totally ridiculous position.

My understanding was that it only becomes a 'loaf' when the legs get tucked underneath so that they look like a bun with rounded corners. When the paws are parallel out front, then you get the sphinx. See, of course, Reddit: https://www.reddit.com/r/Catloaf/

You're right, I stand corrected!

Those observations were excluded.

I would have expected the center area to get wider as we move up and to the right.

Because the sample number increases, the confidence improves and the test can detect lower and lower biases from the 0.5 chirality preference probability.

The exact shape depends on your cost function. If you want to stop as early as possible when a strong effect is present (at the cost of running the test longer when the effect is weak), you can make the area get wider.

Reply on news.ycombinator.com