Notices
Results 1 to 5 of 5

Thread: Correlation and regression question

  1. #1 Correlation and regression question 
    Forum Freshman
    Join Date
    Apr 2008
    Posts
    35
    Hi,

    I must make a correlation and regression assignment and I have to search
    for 2 different facts. And see if they have a connection with each other.
    So I thought about the Budget balance of GDP in % and the trade balance.
    Is there any mathematician that can tell me if it is possible to make this assignment with these two facts? (I know for example that the GNP per capita and the energy consumption has a good connection but I want something original like this)

    I have the idea from this table:


    Reply With Quote  
     

  2.  
     

  3. #2 Re: Correlation and regression question 
    . DrRocket's Avatar
    Join Date
    Aug 2008
    Posts
    5,486
    Quote Originally Posted by gamelogic
    Hi,

    I must make a correlation and regression assignment and I have to search
    for 2 different facts. And see if they have a connection with each other.
    So I thought about the Budget balance of GDP in % and the trade balance.
    Is there any mathematician that can tell me if it is possible to make this assignment with these two facts? (I know for example that the GNP per capita and the energy consumption has a good connection but I want something original like this)

    I have the idea from this table:
    Certainly you can calculate a correlation. You can calculate a correlation between any two sets of data with a common basis (like time).

    But the fact that you show a correlation does not mean that there is any meaningful relationship between the two quantities, and one ought never to confuse correlation with causality.

    Example. Old people tend to have more money than young people. Old people tend to die more often than young people. Conclusion: Wealth will kill you.

    If you buy that argument, then I will be happy to help you live longer by accepting a check for your net worth -- but I prefer cash.


    Reply With Quote  
     

  4. #3  
    Forum Freshman
    Join Date
    Apr 2008
    Posts
    35
    hahah :-D ,
    I see your point.

    Then propably I won't be able to use this table for a meaningfull correlation.
    Does anyone have a good original idea for an economical correlation.
    I really can't find something in economy that hasn't anything to do with causality.
    Reply With Quote  
     

  5. #4  
    Forum Freshman
    Join Date
    Apr 2008
    Posts
    35
    I saw this peace of information.


    Does the population and
    the total land/water area have a connection for correlation or is there
    another factor that can be a cause for the expansion of the us area?
    Reply With Quote  
     

  6. #5  
    Forum Freshman
    Join Date
    Oct 2007
    Posts
    39
    One famous example of a misinterpretion of correlation is the famous Phillips Curve.

    http://en.wikipedia.org/wiki/Phillips_Curve

    In fact, it was argued that one can exploit the relationship between the rate of uneployment and inflation, meaning that there would be some kind of a trade off. In this way a policy maker could simply choose a higher inflation rate to lower unemployment ... it turned out that this reasoning was too short sighted

    there are several problems with correlations, as e.g. the Lucas Critique

    http://en.wikipedia.org/wiki/Lucas_critique

    Trending, seasonaliy etc are dangerous issues too as they can lead to spurious regression.

    if its only for the sake of calculating a few corr coefficients you could use the phillps curve example (unemployment vs inflation). you might find it intersting to look at lagged effects e.g. uneployment in period t vs. inflation in period (t-1)

    i do not know how far you have to go with your assignment. calculating correlations are possible in both of your examples, the point is to be careful when interpreting the results as causal effects> to tackle this problem you need some techniques from statistics and probability theory such as hypothesis testing and stationarity ...
    Reply With Quote  
     

Bookmarks
Bookmarks
Posting Permissions
  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •