• Support PF! Buy your school textbooks, materials and every day products Here!

Correlation and gradient

  • #1
118
0

Homework Statement


I have have a set of data pairs (x, y);
(1, a)
(2, b)
(3, c)
(4, d)
(5, e)
(6, f)
(7, g)

The least squares regression line for the this set is y=3x-12

Determine the new gradient of this line if the original set of scores has been transformed to;

(6, a+3)
(12, b+3)
(18, c+3)
(24, d+3)
(30, e+3)
(36, f+3)
(42, g+3)

i.e. the x scores have been multiplied by 6, and the +3 has been added to the y scores.
Now from my statistical tables book; I have the formula;
[tex]m_{gradient} = \frac{{{\rm{covariance}}}}{{{\rm{variance}}}} = \frac{{S_{xy} }}{{S_{x^2 } }}[/tex]

how can I find the new gradient?

The answer says;
[tex]\begin{array}{l}
m_{gradient} = \frac{{{\rm{covariance}}}}{{{\rm{variance}}}} = \frac{{S_{xy} }}{{S_{x^2 } }} = \frac{{ \times 6}}{{ \times 36}} = \times \frac{1}{6} \\
Hence\;gradient\;is\;now\;3 \times \frac{1}{6} = \frac{1}{2} \\
\end{array}[/tex]

I don't really understand how this process works and don't want to assume anything that is wrong

thanks in advance
Steven
 

Answers and Replies

  • #2
HallsofIvy
Science Advisor
Homework Helper
41,833
956
I'm no expert on statistics but I know that the "gradient" of a line is just its slope. I would have done this, ignoring all the statistical stuff, by arguing that y is now yold+ 3 and x is now 6xold so that yold= y- 3 and xold= x/6. Since you are told that yold= 3xold- 12, you now have y-3= 3(x/6)- 12 or simply y= x/2- 15. The slope (gradient) of that line is 1/2.
 
  • #3
118
0
Cheers for that HallsofIvy!

seems like the obvious thing to do looking back lol

thanks!
 
  • #4
hi steven10137,
can you please tell me the name of the book from where you read this formula
of gradient and correlation.
 

Related Threads on Correlation and gradient

  • Last Post
Replies
0
Views
1K
  • Last Post
Replies
2
Views
2K
  • Last Post
Replies
1
Views
1K
  • Last Post
Replies
1
Views
5K
  • Last Post
Replies
4
Views
750
Replies
4
Views
1K
  • Last Post
Replies
3
Views
2K
  • Last Post
Replies
5
Views
5K
Replies
5
Views
3K
  • Last Post
Replies
2
Views
1K
Top