+- HP Forums (https://www.hpmuseum.org/forum)
+-- Forum: HP Calculators (and very old HP Computers) (/forum-3.html)
+--- Forum: HP Prime (/forum-5.html)
+--- Thread: Statistical inference app import buggity? (/thread-21868.html)
Statistical inference app import buggity? - cjsuk - 06-10-2024 01:06 PM
Doing some a simple one var z-test. Turns out the 1-var stats import is a bit, how shall we say, "hmm".
I'd expect the mu/sigma vars to be pulled from the 1-var stats app if you do an import but it brings the thing into the x-bar variable and completely misses sigma. n is at least populated correctly. This results in a little poking of the vars menu to get things into the app properly.
Or am I holding it wrong?
Rather like it though. It sucks considerably less than doing it in R!
RE: Statistical inference app import buggity? - jte - 06-11-2024 02:45 AM
(06-10-2024 01:06 PM)cjsuk Wrote: Doing some a simple one var z-test. Turns out the 1-var stats import is a bit, how shall we say, "hmm".
I'd expect the mu/sigma vars to be pulled from the 1-var stats app if you do an import but it brings the thing into the x-bar variable and completely misses sigma. n is at least populated correctly. This results in a little poking of the vars menu to get things into the app properly.
Or am I holding it wrong?
Rather like it though. It sucks considerably less than doing it in R!
Is it possible to give a short example of data and how interactions / results compare to expectations?
When I press "Import" from the Inference app's Numeric view (with a one-variable z test selected in the Inference app's Symbolic view), the screen that comes up has "Import Sample Statistics" as a title. As "sample" is in the title (rather than "population"), I'd think it is sample statistics (x bar and n) that are being imported rather than population parameters (mu and sigma). (Well, and x bar and n are shown in the import screen.) (That the data in the Statistics 1Var's Numeric view is sample data, not data for the entire population.)
Is a second type of import wanted (for when data is available for an entire population), or am I misunderstanding?