/sci/ - Science & Math » Thread #13540490

37KiB, 600x502, 1617970767588.gif

View Same Google iqdb SauceNAO

I still don't understand p-values

Anonymous Fri 20 Aug 15:26:08 2021 No.13540490 View Reply Original Report

Quoted By: >>13540497 >>13540527 >>13540731 >>13540937 >>13541596 >>13542058

Anonymous

Anonymous Fri 20 Aug 2021 15:27:27 No.13540497 Report

Quoted By:

>>13540490
The p is short for penis. When you hear someone talking about a "low p-value", they're talking about how small and thin their penis is.

Anonymous

View Same Google iqdb SauceNAO 1623371399364.jpg, 49KiB, 498x573

Anonymous Fri 20 Aug 2021 15:36:40 No.13540527 Report

Quoted By:

>>13540490
p value is bascially how good your pee tastes when you drink it

Anonymous

Anonymous Fri 20 Aug 2021 16:36:54 No.13540731 Report

Quoted By: >>13540969

>>13540490
P means pissosterone. Low p-values mean low pissosterone, which is correlated with low T. Personally I think low T causes low P.

Anonymous

Anonymous Fri 20 Aug 2021 17:10:04 No.13540814 Report

Quoted By: >>13540820 >>13541496 >>13542066

A p value of <0.05 means that the results that you are seeing have a less than 5% chance of occurring if the null hypothesis is correct.
For example, if I flip a coin 10 times and it lands on heads 10 times, the probability of that outcome is 1/1024 ~ 0.00098 if the coin is fair.
The 0.00098 is the p value in this problem, and the null hypothesis is that the coin is fair.
Because 0.00098 < 0.05, by convention, we would switch our hypothesis to that which says the coin is not fair.

Anonymous

Anonymous Fri 20 Aug 2021 17:11:19 No.13540820 Report

Quoted By: >>13540847 >>13541586

>>13540814
Wrong. Kys pseud.

Anonymous

Anonymous Fri 20 Aug 2021 17:18:37 No.13540847 Report

Quoted By: >>13540905

>>13540820
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5804470/
>The p-value is the probability of the observed data given that the null hypothesis is true, which is a probability that measures the consistency between the data and the hypothesis being tested if, and only if, the statistical model used to compute the p-value is correct (9)

Anonymous

Anonymous Fri 20 Aug 2021 17:40:00 No.13540905 Report

Quoted By: >>13540936

>>13540847
That definition is correct. It's not the same as the one you gave.

Anonymous

Anonymous Fri 20 Aug 2021 17:47:07 No.13540936 Report

Quoted By: >>13540942

>>13540905
>The p-value is the probability of the observed data given that the null hypothesis is true, which is a probability that measures the consistency between the data and the hypothesis being tested if, and only if, the statistical model used to compute the p-value is correct (9)

>The p-value is the probability of the observed data [10 heads in 10 flips]
>given that the null hypothesis [the coin has a 50% chance of heads or tails]
>is true, which is a probability that measures the consistency between the data [10 heads in 10 flips]
>and the hypothesis being tested [the coin has a 50% chance of heads or tails]
>if, and only if, the statistical model [at least 10 heads in 10 flips = 0.5^10] used to compute the p-value is correct (9)

Anonymous

Anonymous Fri 20 Aug 2021 17:47:16 No.13540937 Report

Quoted By:

>>13540490
Basically it's the odds your experiment would give a result confirming your hypothesis, if it was wrong, just by pure random chance.

Anonymous

Anonymous Fri 20 Aug 2021 17:48:08 No.13540942 Report

Quoted By: >>13540955

>>13540936
Pseud. Leave the statistics to the mathematicians.

Anonymous

Anonymous Fri 20 Aug 2021 17:50:51 No.13540955 Report

Quoted By: >>13540976

>>13540942
Enlighten me on what precisely is incorrect.

Anonymous

Anonymous Fri 20 Aug 2021 17:53:56 No.13540969 Report

Quoted By:

>>13540731
Say pissosterone without "piss"

Anonymous

Anonymous Fri 20 Aug 2021 17:55:03 No.13540976 Report

Quoted By:

>>13540955
>Sealioning

Anonymous

Anonymous Fri 20 Aug 2021 20:12:46 No.13541496 Report

Quoted By: >>13541586

>>13540814
This is categorically false.

Anonymous

View Same Google iqdb SauceNAO argument_pyramid.png, 238KiB, 679x516

Anonymous Fri 20 Aug 2021 20:39:07 No.13541586 Report

Quoted By: >>13541601

>>13541496
>>13540820

Anonymous

Anonymous Fri 20 Aug 2021 20:42:14 No.13541596 Report

Quoted By:

>>13540490
The p-value is the probability of observing your result *or a result that's even more extreme*.
If your result has a p-value of 5% all it means is that if you did the experiment again, you have a 5% chance of the result being the same as before, or even more extreme.

Anonymous

Anonymous Fri 20 Aug 2021 20:43:27 No.13541601 Report

Quoted By: >>13541761

>>13541586
Granted, the arrow sits remarkably high up compared to most rebukes on /sci.

Anonymous

Anonymous Fri 20 Aug 2021 21:27:00 No.13541761 Report

Quoted By:

>>13541601
Really? Fuck, even when I try to post low quality it's still higher quality than other people's shit here.

Anonymous

Anonymous Fri 20 Aug 2021 23:11:20 No.13542058 Report

Quoted By:

>>13540490
p-values are basically pseudoscience. Don't worry about it.

Anonymous

Anonymous Fri 20 Aug 2021 23:14:21 No.13542066 Report

Quoted By:

>>13540814
wrong & low-p

Anonymous

View Same Google iqdb SauceNAO nothanks.jpg, 97KiB, 900x900

Anonymous Sat 21 Aug 2021 00:12:43 No.13542206 Report

Quoted By:

>frequentist statistics

Capcode	All Only User Posts Only Moderator Posts Only Admin Posts Only Developer Posts
Show Posts	All Only With Images Only Without Images
Deleted Posts	All Only Deleted Posts Only Non-Deleted Posts
Ghost Posts	All Only Ghost Posts Only Non-Ghost Posts
Post Type	All Only Sticky Threads Only Opening Posts Only Reply Posts
Results	All Grouped By Threads
Order	Latest Posts First Oldest Posts First

Your latest searches