Project

General

Profile

Support #1286

Appropriate weights to use with Wave 7 "chmain" dataset

Added by David Irvine 8 months ago. Updated 8 months ago.

Status:
Feedback
Priority:
Urgent
Assignee:
Category:
Weights
Target version:
-
Start date:
11/27/2019
Due date:
% Done:

70%

Estimated time:

Description

Hi,

Sorry to bother you again. I work for DWP and have been trying to do some cross sectional analysis of the wave 7 Child Maintenance datafile (“g_chmain”). As far as I can tell this dataset is at the child level with one row per relevant child (is that right??) but I wasn’t sure how to go about weighting these child-level observations if, for example, we wanted to produce a cross tab of the total number of children in separated families by the amount of time the children spend with non-resident parent (using g_cseeterm variable).

Each child-level record in the “chmain” dataset has a PIDP person identifier but this seems to relate to the parent of each child rather than the child themselves. There is a child-level person number variable called “g_childpno” in chmain so my plan was to use this to link each child’s record from chmain to their corresponding record in the “indall” table, assuming g_childpno in chmain corresponds to g_pno in the indall table - is this correct??. From the children’s records in the indall table I could get the g_psnenub_xw cross sectional person weight for the weighting and also the g_strata and g_psu variables to take stratification and clustering into account (and then use proc surveyfreq in SAS to produce the crosstabs). Does this seem like a reasonable approach and would these be the right variables to use?

Grateful for any advice you can offer.

Many thanks

David

History

#1 Updated by Stephanie Auty 8 months ago

  • Private changed from Yes to No
  • Assignee set to Olena Kaminska
  • Category set to Weights

#2 Updated by Olena Kaminska 8 months ago

  • Assignee changed from Olena Kaminska to Alita Nandi

David,

Thank you for your question. I suggest you use g_psnenui_xw weight as it has a higher sample size, but g_psnenub_xw is also correct. The psu and strata variables that you chose are correct.

Someone else will come back to you about the pidp question.

Best wishes,
Olena

#3 Updated by Stephanie Auty 8 months ago

  • % Done changed from 0 to 30
  • Assignee changed from Alita Nandi to Stephanie Auty
  • Status changed from New to In Progress

#4 Updated by Stephanie Auty 8 months ago

  • % Done changed from 30 to 70
  • Assignee changed from Stephanie Auty to David Irvine
  • Status changed from In Progress to Feedback

Dear David,

Yes, w_childpno in w_chmain corresponds to w_pno in w_indall (within a wave) so, along with w_hidp, can be used to link these files.

Best wishes,
Stephanie

#5 Updated by David Irvine 8 months ago

Perfect. thanks a lot for your help!

Stephanie Auty wrote:

Dear David,

Yes, w_childpno in w_chmain corresponds to w_pno in w_indall (within a wave) so, along with w_hidp, can be used to link these files.

Best wishes,
Stephanie

Also available in: Atom PDF