+ Reply to Thread
Results 1 to 4 of 4

Thread: Including cross-sectional variables in panel data analysis -- can I do that?

  1. #1
    Points: 4,664, Level: 43
    Level completed: 57%, Points required for next Level: 86
    kiton's Avatar
    Location
    Corn field
    Posts
    234
    Thanks
    47
    Thanked 51 Times in 46 Posts

    Including cross-sectional variables in panel data analysis -- can I do that?




    Hello dear forum members,

    I have a panel data set with a count outcome variable. As far as I understand, Yit is conditional on Xit in panel models.

    Yet, I have a set of control variables (e.g., age, graduation year, professional experience) which are cross-sectional. My question: is there a proper way to include these cross-sectional variables in the panel model?

    Thank you in advance!

  2. #2
    Points: 4,664, Level: 43
    Level completed: 57%, Points required for next Level: 86
    kiton's Avatar
    Location
    Corn field
    Posts
    234
    Thanks
    47
    Thanked 51 Times in 46 Posts

    Re: Including cross-sectional variables in panel data analysis -- can I do that?

    Here is some more info on the study and data:

    My goal is to explore the impact of online review-based rating on doctor performance.

    Yit -- "Total number of new patient referrals", where i=1,2,...2000, and t=2009, 2010, 2011, 2012, 2013.
    b1Xit -- "Review-based rating" score (min 1, max 5), where i=1,2,...2000, and t=2009, 2010, 2011, 2012, 2013.
    b2Xit -- "Distance" (min 0, max 9,000), where i=1,2,...2000, and t=2009, 2010, 2011, 2012, 2013.

    I declared the data set for the variables above as panel using: -tsset id year, yearly- (panel variables "strongly balanced").

    On top of that, I have b3Xi, b4Xi, b5Xi (i.e., age, graduation year, professional experience (years) -- suggested as controls by prior literature), which were collected in 2013, where i=1,2,...2000.

    Hope this provides more info on the subject. Your comments are greatly appreciated

  3. #3
    Points: 4,664, Level: 43
    Level completed: 57%, Points required for next Level: 86
    kiton's Avatar
    Location
    Corn field
    Posts
    234
    Thanks
    47
    Thanked 51 Times in 46 Posts

    Re: Including cross-sectional variables in panel data analysis -- can I do that?

    So, I figured that the variables I highlighted above could be "imputed" for the given preceding years. But here is a more tricky part:

    There are three more additional variables:

    b6Xi -- Number of journal publications, min 0, max 717), where i=1,2,...,2000.
    b7Xi -- Number of professional affiliations (min 0, max 5), where i=1,2,...,2000.
    b8Xi -- Number of hospital affiliations (min 0, max 40), where i=1,2,...,2000.

    These variables were collected in 2013 and its close to impossible to "impute" these for the preceding years. Is there a way to deal with these in panel model?

  4. #4
    Points: 4,664, Level: 43
    Level completed: 57%, Points required for next Level: 86
    kiton's Avatar
    Location
    Corn field
    Posts
    234
    Thanks
    47
    Thanked 51 Times in 46 Posts

    Re: Including cross-sectional variables in panel data analysis -- can I do that?


    Actually, I have figured out how to deal with these as well

    Those forum members facing similar issues, please do not hesitate to send a message and I will explain.

+ Reply to Thread

           




Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts






Advertise on Talk Stats