Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Weights using the SIPP Database

    Hello,

    I am working with the SIPP (Survey of Income and Program Participation) database, and I am having trouble dealing with the weights.
    Note: I am using data from the NBER server: http://www.nber.org/data/sipp.html.

    I want to use the panels 1996, 2001, 2004 and 2008. I need to combine all waves from the core files, and wave 2 from the topical modules, for each panel.
    In the end, I want to have 1 dataset with all panels (with all waves from core files and wave 2 of the TM files).

    Could anyone help me with this?

    I was thinking about doing the following: 1) For each panel:
    i) Append all waves of the core files
    ii) Merge the dataset resulting from i) with the topical module wave 2 file
    iii) Append all waves of the replicate weight files (containing the variables repwgt1-repwgt120)
    iv) Merge the dataset from resulting ii) with the dataset resulting from iii)
    2) Append the datasets resulting from iv) for every panel 1996-2008
    3) Declare a panel as follows: svyset [pweight= whfnwgt], brrweight(repwgt1-repwgt120) fay(.5) vce(brr) mse
    Does anyone know if it would be correct in terms of weights?

    Also, I am working with data from the NBER. For 1996, they only provide the Panel Longitudinal Core Files (Description: http://www.nber.org/sipp/1996/sipp96l.pdf), but do not give access to replicate weight files. When using PLC files, do I not need to use replicate weights?

    If anyone familiar with the SIPP database could give me guidance on this, that would great!

    Thank you very much in advance!

  • #2
    Hi Ana!
    I just read your post, 2 years after you have posted it. I have to construct a panel exactly like yours. Any advice about the procedure?
    Thanks a lot!

    Comment

    Working...
    X