Doris,
A variation on Albert-Jan's suggestion would be:
Aggregate
/ outfile = *
/ break = Stud_ID predictor_name
/ count = n.
This would give you one record for each variable per study. I am assuming
that the file is already sorted on the break variables Stud_ID and
predictor_name.
HTH,
Stephen Brand
For personalized and professional consultation in statistics and research
design, visit
www.statisticsdoc.com
-----Original Message-----
From: SPSSX(r) Discussion [mailto:
[hidden email]]On Behalf Of
Doris Gerstner
Sent: Tuesday, February 06, 2007 4:40 AM
To:
[hidden email]
Subject: summarize
Hi list,
I`m working on a data base containing of different studies about predictors
in the context of a literature review. The collected data show multiple
records per single study in the way that a specification of a variable e.g.
stress occurs several times. For example:
Stud_ID predictor_ID predictor_name
1 1 stress
1 2 stress
.
.
.
25 37 age
25 38 age
25 39 age
This is because a special predictor could have different categories. But
these categories aren`t the same over the studies, so I focused on the
variable specification per se what arose the following question: how can I
summarise the multiple records in that way that a special predictor like
age appears only 1x per study, in the above example that study 25 counts
the predictor "age" only 1x?
Thanks for your help, Doris
Locked
