turn on suggestions

Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

Find a Community

- Home
- /
- Analytics
- /
- Stat Procs
- /
- How to convert Z-scores to p-values?

Topic Options

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

06-06-2017 03:56 PM

I have some data for which I have calculated z-scores -does anyone know if there is a function or built-in format that will return the p-values? It's data I have standardized, so I end up with just a number and a standard error, so I can't run it through any of the statistical procs. I'm using v 9.4.

Thanks -

Accepted Solutions

Solution

06-06-2017
04:59 PM

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to mmwhite1

06-06-2017 04:06 PM

PROBNORM

But PROC TTEST can take statistics and do calculations. See the example in the documentation.

All Replies

Solution

06-06-2017
04:59 PM

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to mmwhite1

06-06-2017 04:06 PM

PROBNORM

But PROC TTEST can take statistics and do calculations. See the example in the documentation.

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to Reeza

06-06-2017 05:00 PM

Thanks! PROBNORM is giving me values similar to those in the z-score table.

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to mmwhite1

06-06-2017 04:06 PM

One-sided or two-sided? Use the CDF function. See "Four essential functions for statistical programmers."

z = -1.96;

prob = 2 * cdf("Normal", z); /* two-sided when z < 0 */

z = 1.96;

prob = 2 * (1 - cdf("Normal", z)); /* two-sided when z > 0 */

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to Rick_SAS

06-06-2017 05:46 PM

One-tailed - so looks like I just drop the '2*'. And thanks for the reference!