Mathematics / Mathematik / Matemática

Mathematics / Mathematik / Matemática

Posts 1-2 of 2
  • Sandra Cafazzo
    Sandra Cafazzo    Premium Member
    The company name is only visible to registered members.
    I urgently need some help with dummy variables..!!!
    Hello

    I am currently writing my bachelor thesis. The topic is the challenge to efficiently allocate advisory resources in the portfolio advisory business. I am trying to identify which variables/criteria have an impact on the advisory time needed for the client and am developping a model through regression that 'predicts' the advisory time needed for specific clients.

    'Unfortunately' I have many qualitative values like 'is the client working', 'retired' or a 'professional trader'. Or for example the reachibility of the client (easily reachable by advisor, difficult to reach by advisor, only client contacts advisor)I need to use dummy variables (0 or 1). I have read a lot of theory, but I could not find a description on how to do it in practice. Can you help? Can you recommend me something? How can I define which criteria gets a 1 and which one a 0? How can I test it? I need application help, no more theory...

    Thanks in advance
    Sandra
    • 20 Nov 2006, 12:06 pm
  • Katharina Schüller
    Katharina Schüller    Group moderator
    The company name is only visible to registered members.
    Re: I urgently need some help with dummy variables..!!!
    Hi Sandra,

    is there still a problem? Sorry, I haven't read your article earlier, but I'm sure that I can help you.
    It makes no difference at all which criteria gets the 0 and which the 1. The only impact is that signs in parameter estimation are changing from + to - (or vice cersa) if you calculate the GLM.

    To help you in detail, just write me an e-mail and I will look at your data.

    Best wishes
    Katharina
    • 14 Dec 2006, 6:12 pm