[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: bug in Runs p-value
From: |
Ben Pfaff |
Subject: |
Re: bug in Runs p-value |
Date: |
Sat, 13 Sep 2014 11:06:45 -0700 |
User-agent: |
Mutt/1.5.21 (2010-09-15) |
On Sat, Sep 13, 2014 at 09:30:37AM -0700, Ben Pfaff wrote:
> On Sat, Sep 13, 2014 at 07:53:37AM -0700, Douglas Bonett wrote:
> > The p-value reported in the Runs procedure is not computed correctly. The
> > z-value appears correct. The error is easy to find with almost any data.
> > In my case, z = 1.26 and the reported p was 1.794 (it should be .206).
> > Note that 2 - 1.794 = .206.
> >
> > This should be a simple fix.
>
> Thanks for the report.
>
> RUNS has few test cases in the PSPP source tree. Do you have a small
> test case to contribute (with syntax, data, and expected output)? Then
> I would feel much more confident about the fix.
Thanks for providing your example (off-list).
I pushed a fix. Now .206 is reported in your example.