Discussion: View Thread

  • 1.  Regressing group-level DV on an individual-level IV

    Posted 08-01-2022 17:48

    Dear Strategy Researchers:

     

    I want to regress a group-level DV on an individual-level independent variable (IV). For example, regress an industry's growth in a year on its constituent firm's sales in the same year. That is, for a given year, my data set has rows where the IV varies, but the DV remains the same.

     

    If you know of the correct econometric estimator for the above specification, please consider responding.

     

    Thank you for guiding me.

    Vivek Astvansh

    Assistant Professor, Department of Marketing, Kelley School of Business

    Adjunct Professor of Data Science, Luddy School of Informatics, Computing, and Engineering

    Indiana University Bloomington

    https://kelleyschool.iu.edu/astvansh

     

     



  • 2.  RE: Regressing group-level DV on an individual-level IV

    Posted 08-02-2022 21:49
    You can use regular regression, but have to cluster the standard errors at the group level, i.e. the industry in your example.
    See: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3694931

    ------------------------------
    Phebo Wibbens
    Assistant Professor
    INSEAD
    Fontainebleau Cedex
    ------------------------------