PROTEUS

              

 

 

 

 

 

 

 

 

Back to ENH440 Problems and solutions

 

#10 - Regression: Graphic view 
  

Here is the graphic view of #10

Note that you can choose ANY point on the x axis for the "pivot" point.  I have chosen x=10. 

When x=10 is entered into the Y' = A + BX  equation, I obtain  Y' = 3.42

CL around B are calculated as   CL(99.9 %)  B  ±  t(0.001)(11df)  x (SB)

                       Thus  CL(99.9 %) =   + 0.270  ±  (4.44)(0.0135)

                  giving the two limits:   + 0.32994    and  + 0.2116

BOTH are the same sign (+) as the original slope so we can reject the null hypothesis at 0.001

To draw the lines on the scatter plot we need the new A values for each CL slope.

     A = Y - BX

     A = 3.42 - (0.32994)(10) = 0.1206

     A = 3.42 - (0.21006)(10) = 1.3194

These are indicated on the Y axis and the line drawn