Dear Rick,
I'm not sure whether there's any easy option in SPSS but you could use the OMS to create a Dataset containing correlations, N and p-values and edit this into the correlation matrix desired. The example below is one I used in order to replace all nonsignificant correlations by missing values (-99).
In your case it might be easier to obtain the correlations via the FACTOR menu since you'll have a correlation matrix without p-values and sample sizes. You could then use a recode function like (RECOD <varlist> (-.4 thru .4=SYSMIS)(ELSE=COPY)).
HTH,
Ruben van den Berg
* OMS.
DATASET DECLARE Corr.
OMS
/SELECT TABLES
/IF COMMANDS = ["Correlations"]
SUBTYPES = ["Correlations"]
/DESTINATION FORMAT = SAV NUMBERED = TableNumber_
OUTFILE = Corr
/TAG = "Corr".
COR V70_1 to V70_16 WITH V80_1 to V80_5.
OMSEND TAG=['Corr'].
DATAS ACT Corr.
COMP ID=$casenum.
SOR CAS by ID (D).
DO REP Vars=V80_1 to V80_5.
DO IF LAG(Vars)>.05.
DO IF Var2='Pearson Correlation'.
COMP Vars=-99.
END IF.
END IF.
END REP.
EXE.
SEL IF Var2='Pearson Correlation'.
EXE.
Date: Wed, 29 Jul 2009 08:48:36 -0400
From:
[hidden email]Subject: Suppressing Output (Pearson correlations)
To:
[hidden email]
Hi All,
Is there a way to suppress PASW output by the magnitude of correlation coefficients? For example, suppose I am running a number of correlations and only want to see those with r > .40 in my output. Is this possible? I don’t see anything in the syntax guide.
Thanks,
Rick
Richard W. Handel, Ph.D.
Associate Professor
Department of Psychiatry and Behavioral Sciences
Eastern Virginia Medical School
825 Fairfax Avenue, Hofheimer Hall
Norfolk, VA 23507
Phone (757)-446-7992
Express yourself instantly with MSN Messenger!
MSN Messenger
Locked
