p-value of <.001 appeared as "1" in one correlation matrix

khagan · Post by **khagan** » Tue Jul 26, 2022 6:59 am

Hello,

I am teaching a class, and we generated a correlation matrix (pictured). One student showed me an otherwise identical correlation matrix generated in Jamovi but the p-value for each of the correlations read as "1" instead of <.001. Are you able to give me any ideas about why this might have happened? We repeated the correlation matrix and it gave to (expected) p-values pictured here, so I am not sure if there were settings or something that might account for the p-value previously showing up as one.

With appreciation,
Katherine

Post by **jonathon** » Tue Jul 26, 2022 7:04 am

hi,

if you provide the steps to reproduce the problem, we'll be able to take it from there.

cheers

jonathon

khagan · Post by **khagan** » Tue Jul 26, 2022 7:28 am

Thank you for your reply. I just saw the output on another person's screen, so I am unfortunately not able to reproduce the problem (though I did try, just by selecting different options to see if I could recreate the output I was seeing). Thank you, though!

Bobafett · Post by **Bobafett** » Wed Jul 27, 2022 3:21 pm

At a guess, could they have selected a one-tailed p-value to be produced - and it is pointing in the 'wrong' direction to what would be expected? Rather than p>.999 it could then simply be rounding up?

MAgojam · Post by **MAgojam** » Wed Jul 27, 2022 7:35 pm

I realized that Bobafett has already given the explanation, but I leave this screenshot that I had prepared anyway.

: Correlation.png (102.54 KiB) Viewed 4822 times

Cheeers,
Maurizio

khagan · Post by **khagan** » Wed Jul 27, 2022 7:46 pm

Thank you, MAgojam. This is very helpful to see. I really appreciate the example.

jamovi