Changing values with every new analysis

Markus

Hello!

When I do the same analysis (e.g. repeated measures ANOVA with the Bugs example in JASP 0.8.0.0) repeatedly in succession, I always geht different BFs. I think this shouldn´t be, right?

Thank you very much!
Best,
Markus

EJ

Hi Markus,

Some analyses require a numerical approximation. The quality of the approximation is indicated in the output tables as a "%error". If I'm not mistaken the current version of JASP allows you to improve the quality of the approximation by increasing the number of draws (under "Advanced Options"). Bottom-line: because of the numerical techniques that we sometimes have to use, the results may vary, but only slightly. You can also see this when you do one-sided t-tests (the one-sided adjustment requires a numerical trick but this introduces slightly different results every time).

Cheers,
E.J.

Markus

Hi EJ, thank you very much for your helpful response!
Best,
Markus