Date:         Mon, 24 Jun 2002 16:23:23 +0400
Reply-To:     Andrew <a3000@imail.ru>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Andrew <a3000@imail.ru>
Subject:      Re: Reliability and listwise exclusion of observations
In-Reply-To:  <5.0.2.1.1.20020624110151.00a6c5c0@cres.anu.edu.au>
Content-Type: text/plain; charset=us-ascii

Hi Rosemary,

  SPSS have one function - NMISS(variable[,..]), which returns a count
of the arguments that have missing values. This function requires one
or more arguments, which should be variable names in the working
data file.

  And now you can use it in more variants with and without filters.
I use filters (be careful with deleting unselected case!!!):
  - Use menu Data/Select cases/ ... If ... type NMISS(var1 to var40)<=0
  - Use Syntax (see below).

  Also you can create variable with number of miss vars in case and
then filter cases...
  Test and use syntax code. Step by step.

Good luck, Andrew.
Monday, June 24, 2002

P.S.
  It is very fast and easy answer.
  There are many ways to do task more exactly and useful :)

/* START SYNTAX SELECT FULL DATA CASES */.

/* In process will be added next vars: filter_$ */.

DATA LIST FIXED /id 1 var0 3 var1 5 var2 7 var3 9 var4 11.
BEGIN DATA
1 3 2 2 2 2
2   2 2 2 2
3 2   2 2 2
4 0 3 9 9 9
5 1 3 0 0 0
6 0     0 0
END DATA.
LIST.

* STEP 1.
* Rosemary, CHANGE var4 on var40 in your task and go on.
USE ALL.
COMPUTE filter_$=(NMISS(var0 to var4) <= 0).
VARIABLE LABEL filter_$ 'NMISS(var0 to var4) <= 0 (FILTER), USING FULL DATA CASES'.
VALUE LABELS filter_$  0 'Not Selected' 1 'Selected'.
FORMAT filter_$ (f1.0).
FILTER BY filter_$.
EXECUTE .

/* END SYNTAX */.

RESULTS
ID VAR0 VAR1 VAR2 VAR3 VAR4  FILTER_$
 1    3    2    2    2    2  Selected
 4    0    3    9    9    9  Selected
 5    1    3    0    0    0  Selected


> Original message
>
> Date:    Monday, June 24, 2002
> From:    R F James <rfjames@cres20.anu.edu.au>
> Subject: Reliability and listwise exclusion of observations
>
> I need to conduct a reliability analysis using only the 6000 observations
> (of 12000) for which I have complete data on all of 40 variables.  Subsets
> of the variables will be used in each reliability analysis.

> Can anyone tell me how to instruct SPSS 11.0 to use only the observations
> for which there is data on all variables (not just the ones in  particular
> reliability analysis)?  There must be a way to do this with the drop-down
> menus  .... musn't there?

> Many thanks in advance.

> Rosemary James
> Australian National University
> Australia
> Rosemary F.James
> Centre for Resource and Environmental Studies
> Australian National University
> Phone:  61 (0)2 6125 2635

> Executive Committee member
> Australia New Zealand Society for Ecological Economics

> Mailing address:

> CRES
> ANU
> ACT 0200
> Australia