JASP Student t Distribution probability calculations are doubled

[mspezio]

• JASP version: 0.14.1
• OS name and version: Mac 10.15.7
• Analysis: Distributions > Normal
• Bug description: The probability density that JASP reports for df = 1000, ncp = 0, for 2.58 to infinity is 0.01. But this value should be 0.005010866. There seems to be a doubling of the probability density.
• Expected behaviour: As in R, 1-pt(2.58,1000) = 0.005010866

Is this fixed in the next version that is already in development?

Thank you.

[Kucharssim]

Thank you for this report! The value is not doubled, but rounded to two decimal places... I'll see how we can handle this better

[mspezio]

Thank you Kucharssim, that's helpful. I found this for several of the other distributions as well, for binomial and Poisson. Others did not have such issues. For instructional purposes, to help students understand probabilities as areas under the pdf or pmf curves, it would be helpful to have JASP report the numbers rounded to 2 significant figures, after all of the decimal positions, equivalent to the R functions for the distributions. Thanks again.

[JoKeyser]

Maybe the module could use the number of decimal points that are wanted by the user's settings?

[tomtomme]

@mspezio @Kucharssim This rounding-bug is still present in jasp 0.18.3. The distribution module should honor the user settings from "preferences => results => table options => fix number of decimals" I realize this setting was meant for tables only. If it is hard to expand to plots, please consider adding a table of the results below the distribution plot, like also requested here for discrete distris: https://github.com/jasp-stats/jasp-issues/issues/1597

[Kucharssim]

please consider adding a table of the results

Yes, this is in the works