|
Hi,
I am currently working on a geographical dental survey in which weights are used for different regions/areas to counter the differing populations. I therefore want to apply the relevant weight to the raw data and then aggregate the data to give means, SDs, etc... of a number of different variables. Unfortunately, the SD is often not generated for certain variables in certain areas so, for example, 17 out of 22 areas have a SD for age and the other 5 are just missing. Just to test, I took the weighting off and rand the syntax again and all the SDs were generated (though they were obviously meaningless) suggesting that applying the weighting causes the problem. Any ideas how I can resolve this problem? Cheers, Tim. ===================== 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 |
|
Probably the weights are missing for cases in those regions, or they are
missing for some cases in such a way that the remaining cases have all the same value and then the SD would be zero. Weighting in SPSS means that each individual value is multiplied by the weight before being counted. When the weight variable is missing for a case, the multiplication has no result, and anything resulting from that procedure would be missing. However, when some cases have missing weights and others do not, the aggregation is carried out on the remaining cases. Since you do not obtain any result, it is probably all the relevant weights that are missing. Hector -----Original Message----- From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Tim Pickles Sent: 02 October 2009 12:09 To: [hidden email] Subject: Weighted SDs in Aggregate Command Hi, I am currently working on a geographical dental survey in which weights are used for different regions/areas to counter the differing populations. I therefore want to apply the relevant weight to the raw data and then aggregate the data to give means, SDs, etc... of a number of different variables. Unfortunately, the SD is often not generated for certain variables in certain areas so, for example, 17 out of 22 areas have a SD for age and the other 5 are just missing. Just to test, I took the weighting off and rand the syntax again and all the SDs were generated (though they were obviously meaningless) suggesting that applying the weighting causes the problem. Any ideas how I can resolve this problem? Cheers, Tim. ===================== 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 ===================== 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 |
| Free forum by Nabble | Edit this page |
