Date: Tue, 15 Feb 2005 15:04:43 +0100
Reply-To: Antoon <email@example.com>
Sender: "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From: Antoon <firstname.lastname@example.org>
Content-Type: text/plain; charset="iso-8859-1"
I hav a question about weighting.
I work with a file of young respondents that have committed minor delinquency. Half of the group got a learning and or work penalty, the other half (control group) got no (official) punishment.
Regretfully the groups are not completely comparable as to etnicity (6 categories) and wether the act was committed alone, or with others (4 categories: alone, with one friend, with more friends, with people I do not know).
Now I would like to create weights to adjust for these differences. So I made a crosstabulation of etnicity by group offendership, outputing the total percent, then split the file by treatment group and made the same crosstabulation.
Now computing the weights was done by dividing the total percent for all subjects by the total percent of the respective treatment groups. In pseudo code:
DO IF condition = punishnent.
DO IF etnicity = Dutch.
DO IF offendership = alone
COMPUTE weight = TotalPercentOfDutchOffendingAlone / TotalPercentOfPunishmentgroupDutchOffendingAlone
ELSE IF offendership = WithOneFriend
COMPUTE weight = TotalPercentOfDutchOffendingWithOneFriend / ....
ELSE IF... etc
ELSE IF etnicity = ...
ELSE IF condition = NoPunishment
I have two questions:
1 Is my reasoining conceptually valid?
2 Is there a less error prone way to get the same results?
Thank you in advance.