Discussion: View Thread

  • 1.  Regression and prediction for aggregate data with corresponding measures of variation

    Posted 11-10-2013 20:12
    Hi everyone,

    The following may be a simple question but I appreciate any answers you may provide. 

    Assume we are in a simple linear regression setting where we need to predict Y from X based on aggregate data available for a set of consecutive calendar years (e.g., 2000, 2001, 2002, ..., 2012).  The aggregate data consist of mean and/or coefficients of variations for X and Y across the set of available years.   

    What kind of frequentist/Bayesian regression method(s) should we use when regressing the outcome Y on the predictor X and we have:
            
              a)  Means and coefficients of variation for X and means only for Y;
              b)  Means for X and means and coefficients of variation for Y;
              c)  Means and coefficients of variation for both X and Y.

    The underlying purpose is prediction of the value of Y for the next year given the available data.    

    Many thanks,

    Isabella

    -------------------------------------------
    Isabella Ghement
    Ghement Statistical Consulting Company Ltd.
    -------------------------------------------


  • 2.  RE:Regression and prediction for aggregate data with corresponding measures of variation

    Posted 11-10-2013 21:41
    A similar problem is more commonly studied in Econometrics.

    From a google search, An article from the FED

    http://research.stlouisfed.org/wp/2002/2002-024.pdf

    and citations


    http://ideas.repec.org/p/ner/tilbur/urnnbnnlui12-153298.html


    -------------------------------------------
    Chris Barker, Ph.D.
    Consultant and
    Adjunct Associate Professor of Biostatistics
    www,barkerstats.com

    ---
    "In composition you have all the time you want to decide what to say in 15 seconds, in improvisation you have 15 seconds."
    -Steve Lacy
    -------------------------------------------








  • 3.  RE:Regression and prediction for aggregate data with corresponding measures of variation

    Posted 11-22-2013 16:44
    Hi Chris,

    Thank you very much for your references - I am in the process of reading them and deciding whether I can use them in my work.  

    Kind regards,

    Isabella
     

    -------------------------------------------
    Isabella Ghement
    Ghement Statistical Consulting Company Ltd.
    -------------------------------------------






    Original Message