Date:         Fri, 17 Jan 1997 12:49:18 -0600
Reply-To:     "Nichols, David" <nichols@SPSS.COM>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@UGA.CC.UGA.EDU>
From:         "Nichols, David" <nichols@SPSS.COM>
Subject:      Re: Difference of proportions, etc.

No, it isn't. The square of a standard normal random variable is
distributed as a chi^2 random variable with one degree of freedom,
but that doesn't imply that a test statistic referred to a Z
distribution, when squared, will give the same value as another
test statistic that's referred to a chi^2 distribution with one
df. There are, of course, multiple chi^2 tests for independence,
and they don't give the same numbers in any particular case
either, though they have the same distribution under the null
hypothesis. I wouldn't have bothered to write a macro to do the
Z-test if you could simply take the square root of the chi^2
value from CROSSTABS and tell people to use that.
 
David Nichols
Senior Support Statistician
SPSS Inc.
nichols@spss.com
 
>----------
>From:  Scheltema, Karen[SMTP:Karen.Scheltema@STATE.MN.US]
>Sent:  Wednesday, January 15, 1997 9:24 AM
>To:    Multiple recipients of list SPSSX-L
>Subject:       Re: Difference of proportions, etc.
>
>it should also be noted that the square of the z-test for independent
>proportions is the same as the chi-square test of independence.
>
>Karen Scheltema, MA MS
>Statistician
>State of Minnesota
>Department of Human Services
>Division of Performance Measurement and Quality Improvement
>444 Lafayette Road
>St. Paul, MN 55155-3823
>(612) 296-4339         fax (612)297-1862
>karen.scheltema@state.mn.us
>
> ----------
>From: David Nichols
>To: "SPSSX(r) Discussion"
>Subject: Re: Difference of proportions, etc.
>Date: Tuesday, January 14, 1997 5:03PM
>
>In article <32D6E6DB.1BE2@spss.com>, David Marso  <marso@spss.com> wrote:
>>Ruth,
>>  The following will probably work.
>>(I don't know how your stuff is coded or I would probably remove
>>the word probably from the above sentence).
>>
>>COMPUTE Groupvar = MAX(Male_Var,Female_var).
>>Then run the T-Test using Groupvar as the grouping variable.
>>
>>This can be done in syntax or the dialog boxes under
>>Data>Transform>Compute...
>>Regards, David Marso
>>
>>Ruth Ann Stacy wrote:
>>>
>>> These are elementary questions but I am new to
>>> SPSS.  First, what test do I use to do a difference
>>> of proportions test:Independent Binomial
>>> Experiments.
>>> Second, I am working with a survey in which females
>>> answered quesA and males answered quesB. (They are
>>> the same question.  Hey, I would have designed the
>>> survey differently.  Anyway, I want to do an
>>> independent sample t-test of difference of means
>>> but using the windows version, it wants a grouping
>>> variable.  The two variables, quesA and quesB
>>> already form the grouping.  Any suggestions?
>>> Thanks in advance.
>
>If for the test of independent proportions you want the popular
>Z-test, this isn't programmed into any SPSS procedure at the
>current time. However, there is a macro available that will do
>it. It's available at
>
>ftp://ftp.spss.com/pub/spss/statistics/nichols/macros
>
>as ztest.sps, with ztest.rm as the readme file. You can get to
>it via our Web site by clicking on Tech Support, then Macros
>under Ftp Archives, and scrolling to the end of the macro section.
>
> --
>
> ----------------------------------------------------------------------------
> -
>David Nichols             Senior Support Statistician              SPSS,
>Inc.
>Phone: (312) 329-3684     Internet:  nichols@spss.com     Fax: (312)
>329-3668
>
> ----------------------------------------------------------------------------
> -
>