+ Reply to Thread
Results 1 to 3 of 3

Thread: least squares method

  1. #1
    Points: 4,982, Level: 45
    Level completed: 16%, Points required for next Level: 168

    Thanked 0 Times in 0 Posts

    Unhappy least squares method

    Has anyone ever heard of using this?? It has to do with the regression equation... pllllllllease explain with lots of detail. statistics is not really my thing. thank you

  2. #2
    TS Contributor
    Points: 17,636, Level: 84
    Level completed: 58%, Points required for next Level: 214
    JohnM's Avatar
    Thanked 6 Times in 5 Posts
    The least squares method is the method of drawing the line of best fit through the points on a scatter plot, to show the relationship between two variables. The term "least squares" means that the squared differences between:

    -the actual points, and
    -the line drawn through them

    are minimized.

    The regression equation is the equation of this line - i.e., what will the value of the dependent variable (y) be if plug a certain value into the independent variable (x).

    Here's a link to an image of a regression line of best fit drawn through the points in a scatter plot. As you can see, the distances (squared) between the points and the drawn line are kept to a minimum - any other line drawn through the points would not meet this "least squares" criteria.

  3. #3
    Points: 5,699, Level: 48
    Level completed: 75%, Points required for next Level: 51

    Thanked 0 Times in 0 Posts

    Say you have a scatterplot and you are told to draw a regression line through it. In primary school science class you just roughly draw a line through all the dots. But the problem with this is that every student gets different lines.

    Least squares is objective. It minimizes the distance between the dot and the line. The reason why it is squared is because sometimes the dot is below the fitted line and sometimes it is above, which gives negative and positive differences respectively. Squaring it makes it all positive.

+ Reply to Thread


Similar Threads

  1. Method to calculate the most accurate method of prediction
    By joshAU in forum Regression Analysis
    Replies: 1
    Last Post: 10-12-2010, 01:04 PM
  2. least squares
    By leo nidas in forum Probability
    Replies: 5
    Last Post: 04-20-2010, 09:34 AM
  3. Least squares & 95% CI
    By cutiepiene in forum Statistics
    Replies: 3
    Last Post: 12-09-2009, 08:05 PM
  4. Replies: 0
    Last Post: 11-09-2009, 04:00 PM
  5. Chi Squares
    By iggins in forum Statistics
    Replies: 1
    Last Post: 09-07-2008, 05:27 PM

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