+ Reply to Thread
Results 1 to 3 of 3

Thread: Can I compare regression results with principal component analysis(PCA)

  1. #1
    Points: 109, Level: 2
    Level completed: 18%, Points required for next Level: 41

    Posts
    1
    Thanks
    2
    Thanked 0 Times in 0 Posts

    Can I compare regression results with principal component analysis(PCA)




    I'm trying to analyze survey data ( 31 response data, and 5 variables, dependent variable is non-parametric (distribution free)). I tried to use regression and PCA to select the most important independent variables to the dependent variable. My questions are:
    1) Is it legitimate to view regression in the same way as PCA (methods for variable selection)?
    2) Are there any sources to compare these two methods? For example: in terms of the easiness to interpret the results I feel that regression is easier than PCA, or if the method will give the same results if we use one-third or two-third the data points (insensitive to the data point change).

    HELP please. Thanks

  2. #2
    TS Contributor
    Points: 17,749, Level: 84
    Level completed: 80%, Points required for next Level: 101
    Karabiner's Avatar
    Location
    FC Schalke 04, Germany
    Posts
    2,540
    Thanks
    56
    Thanked 639 Times in 601 Posts

    Re: Can I compare regression results with principal component analysis(PCA)

    PCA isn't a method for assessing "importance" (however defined)
    of variables with regard to their predictive power, at least AFAIK.
    It's a data reduction tool.

    With kind regards

    K.


    BTW, only 31 subjects while there are 5 variables wont't give
    you generalizable results in most cases, regardless of what your
    method of analysis will be.

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

    rosaie (05-01-2014)

  4. #3
    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: Can I compare regression results with principal component analysis(PCA)


    Structural equation models use a form of regression, but PCA does not. It's components do reflect variance in the variables, but unlike regression it is not assumed that one variable is explaining another, the relationship can flow both ways (that is rather than X-> Y X->Y and Y->X).
    "Very few theories have been abandoned because they were found to be invalid on the basis of empirical evidence...." Spanos, 1995

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

    rosaie (05-01-2014)

+ 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