How do you quantify the strength of the linear association between the two variables with a number? :confused:

How do you quantify the strength of the linear association between the two variables with a number? :confused:

The correlation coefficient, r²:

http://mathworld.wolfram.com/CorrelationCoefficient.html

They tell you how to calculate r² midway on that page. There's also some nice pictures show how the closer r² is to 1, the stronger the linear relation.

cool, thanks for the link

stuck on this problem:

using the regression equation, predict the number of accidents per 1000 licenses for a city with 19% of its licensed drivers under the age of 21.

regression equation: y = 1.5974 + .287053x

the data:
(City) (% Under 21) (Fatal Accidents per 1000 Licenses)
1 13 2.962
2 12 0.708
3 8 0.885
4 12 1.652
5 11 2.091
6 17 2.627
7 18 3.83
8 8 0.368
9 13 1.142
10 8 0.645
11 9 1.028
12 16 2.801
13 12 1.405
14 9 1.433
15 10 0.039
16 9 0.338
17 11 1.849
18 12 2.246
19 14 2.855
20 14 2.352
21 11 1.294
22 17 4.1
23 8 2.19
24 16 3.623
25 15 2.623
26 9 0.835
27 8 0.82
28 14 2.89
29 8 1.267
30 15 3.224
31 10 1.014
32 10 0.493
33 14 1.443
34 18 3.614
35 10 1.926
36 14 1.643
37 16 2.943
38 12 1.913
39 15 2.814
40 13 2.634
41 9 0.926
42 17 3.256

thanks

I presume u got the regression equation from the data. in that case u don't need the data any more. Just plug in 19% for x and see what comes out.

ok, thanks, i got -1.54285993

Which is impossible.

Thats impossible, all x's are positive and the equation constant is also positive! Tell us what u did.

y = -1.5974 + .287053(.19) = -1.54285993 right? :confused:

Well in the last post, the equation constant was positive. Now its negative! Get rid of the negative sign there and do it again!

it was originally negative. i guess i forgot to add the negative sign. also, excel calculated that equation for me.

Did you plot the percentage of drivers under 21 as decimal fractions or as percentages? Try using 19 as your x value instead of .19 if you did not do this.

i plotted it as regular numbers like 1,2,3...alrighty i guess i'll use 19 instead of .19.