+ Reply to Thread
Results 1 to 5 of 5

Thread: When log transformation

  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

    When log transformation




    Hello,

    I am a bit confused with WHY and WHEN you should take the log of the dependent variable and/or independent variables, after running a regression.

    Is there something in the (multiple)regression outputs in SPSS where you can see that it may be wise to take the logarithm of the variables? I came as far as: transforming log may reduce heteroscadicity

  2. #2
    Omega Contributor
    Points: 38,334, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    6,998
    Thanks
    398
    Thanked 1,186 Times in 1,147 Posts

    Re: When log transformation

    Decisions are based on the relationship (functional form) between the variables. Decisions are usually based the model's residuals (looking at graphs).
    Stop cowardice, ban guns!

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

    Jazz3 (03-14-2017)

  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: When log transformation

    This may stupid but..... at what do you look at exactly in the graphs? if it is linear?

  5. #4
    Omega Contributor
    Points: 38,334, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    6,998
    Thanks
    398
    Thanked 1,186 Times in 1,147 Posts

    Re: When log transformation

    Residuals are ideally examined creating a histogram to see if they are normally distributed. You can also employ a normality test, but these reject the hypothesis of normality when sample sizes get large. Most people like to also create a q-q plot where residuals should approximate a diagonal line on the graph. Lastly, a scatter graph can also be used to examine residual distribution against continuous variables in the model plotted on the x-axis, which should approximate a shotgun spray (be randomly distributed with no pattern).
    Stop cowardice, ban guns!

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

    Jazz3 (03-17-2017)

  7. #5
    Points: 533, Level: 10
    Level completed: 66%, Points required for next Level: 17

    Posts
    50
    Thanks
    20
    Thanked 0 Times in 0 Posts

    Re: When log transformation


    Thank you, very clear answer

+ 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