+ Reply to Thread
Results 1 to 10 of 10

Thread: Log regressions

  1. #1
    Points: 533, Level: 10
    Level completed: 66%, Points required for next Level: 17

    Posts
    50
    Thanks
    20
    Thanked 0 Times in 0 Posts

    Log regressions




    So between these two log regressions:
    1) transform y into log but keep X normal
    2) transform both the y and x into log

    When or what would be a logical explanation for only transforming y and not c over transforming them both?

  2. #2
    Omega Contributor
    Points: 38,396, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    7,001
    Thanks
    398
    Thanked 1,186 Times in 1,147 Posts

    Re: Log regressions

    The functional relationship between them may only require one to be transformed in order to normalize the residuals or get a linear relationship.
    Stop cowardice, ban guns!

  3. The Following User Says Thank You to hlsmith For This Useful Post:

    Jazz3 (04-08-2017)

  4. #3
    Points: 533, Level: 10
    Level completed: 66%, Points required for next Level: 17

    Posts
    50
    Thanks
    20
    Thanked 0 Times in 0 Posts

    Re: Log regressions

    Quote Originally Posted by hlsmith View Post
    The functional relationship between them may only require one to be transformed in order to normalize the residuals or get a linear relationship.
    I've also heard that it depends on your research, for example if you expect diminishing returns? Are there more underlying reasons than just in order to normalize the residuals and to get a lineair relationship?

  5. #4
    Human
    Points: 12,676, Level: 73
    Level completed: 57%, Points required for next Level: 174
    Awards:
    Master Tagger
    GretaGarbo's Avatar
    Posts
    1,362
    Thanks
    455
    Thanked 462 Times in 402 Posts

    Re: Log regressions

    What matters is if the model fits to the data. If the model does not fit, then the model is rejected.

    I've also heard that it depends on your research, for example if you expect diminishing returns?
    If you have an hypothesis of "diminishing returns" then that is something that can be tested with the data. Models are not sacrosanct, holy scriptures. They can and should be tested.

  6. The Following User Says Thank You to GretaGarbo For This Useful Post:

    Jazz3 (04-18-2017)

  7. #5
    Omega Contributor
    Points: 38,396, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    7,001
    Thanks
    398
    Thanked 1,186 Times in 1,147 Posts

    Re: Log regressions

    Agreed. Some fields prefer to talk about relationships on relative instead of absolute terms. Though, regardless of the model used it must fit data and meet assumptions.
    Stop cowardice, ban guns!

  8. #6
    TS Contributor
    Points: 12,227, Level: 72
    Level completed: 45%, Points required for next Level: 223
    rogojel's Avatar
    Location
    I work in Europe, live in Hungary
    Posts
    1,470
    Thanks
    160
    Thanked 332 Times in 312 Posts

    Re: Log regressions

    Just to make it more clear, what hlsmith said about the functional form:

    1. if you only transform Y you have a relationship like log Y = a1*x1+ a2*x2 -> Y=exp(a1*x1+a2*x2)
    2. if you transform both then logY = a1*log(x1)+a2*log(x2) -> Y = x1**a1*x2**a2

    regards

  9. The Following User Says Thank You to rogojel For This Useful Post:

    Jazz3 (04-18-2017)

  10. #7
    Points: 533, Level: 10
    Level completed: 66%, Points required for next Level: 17

    Posts
    50
    Thanks
    20
    Thanked 0 Times in 0 Posts

    Re: Log regressions

    Thanks for all the information! Really helpful

    But to come back to diminishing returns to scale I know this applies for log log regression, but also for log linear regressions?

  11. #8
    Points: 533, Level: 10
    Level completed: 66%, Points required for next Level: 17

    Posts
    50
    Thanks
    20
    Thanked 0 Times in 0 Posts

    Re: Log regressions

    Quote Originally Posted by Jazz3 View Post
    Thanks for all the information! Really helpful

    But to come back to diminishing returns to scale I know this applies for log log regression, but also for log linear regressions?
    Or does the log linear only e.g. explains the marginal effect?

  12. #9
    Fortran must die
    Points: 58,790, Level: 100
    Level completed: 0%, Points required for next Level: 0
    noetsi's Avatar
    Posts
    6,532
    Thanks
    692
    Thanked 915 Times in 874 Posts

    Re: Log regressions

    In economics I think, where this type of analysis is most common I believe, there are theoretical reasons commonly for which to log.
    "Very few theories have been abandoned because they were found to be invalid on the basis of empirical evidence...." Spanos, 1995

  13. The Following User Says Thank You to noetsi For This Useful Post:

    Jazz3 (04-18-2017)

  14. #10
    Points: 533, Level: 10
    Level completed: 66%, Points required for next Level: 17

    Posts
    50
    Thanks
    20
    Thanked 0 Times in 0 Posts

    Re: Log regressions


    I've been going through some economics book but I can't find anything about diminishing returns and a log linear. I guess it only applies for log-log

    Can anyone confirm this?

+ 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