Doris
This is quite easy to do. Use Transform-->Rank Cases, and rank on
"willingness to pay" into Ntiles. For 10% splits, set the number of Ntiles
to 10. This will create a new variable which you can use to select or filter
on. Eg select if(Newvar = 1) would select the bottom 10%.
If you have a small sample, and tied data, you may need to experiment with
the way ties are ranked, because the lowest ranking algorithm may not always
be 1, and the highest may not always be 10 in such cases. Otherwise it is
quite straighforward.
Garry Gelade
Business Analytic
-----Original Message-----
From: SPSSX(r) Discussion [mailto:
[hidden email]] On Behalf Of
Doris Gerstner
Sent: 28 October 2008 07:45
To:
[hidden email]
Subject: selct a subset
Dear all,
how can I select a subset of observations at each extreme; e.g. the 10%
highest and 10% lowest values of a variable "willingness to pay" to further
analyse these extreme groups?
Many thanks in advance, Doris
=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command SIGNOFF SPSSX-L For a list of
commands to manage subscriptions, send the command INFO REFCARD
__________ NOD32 3561 (20081027) Information __________
This message was checked by NOD32 antivirus system.
http://www.eset.com=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
Locked
