Page 91 - Basic Statistics

86

Testing criteria:

If this test is used at the significance level of , then
Accept H 0 if F  F(;1;n-2)-table or Pr = P ( F > F-actual )   .

Reject H 0 if F > F(;1;n-2)-table or Pr = P ( F > F-actual ) < . ( R12 )

5.1.3 THE COEFFICIENT OF DETERMINATION, R 2

A measure of how well a linear regression model to explain the
relationship between independent variables with dependent variables, it can be

2
seen from the value of koefisien of determination (R ). In partitioning the total
sum of squares, an indication that the model more appropriate if SS(Reg)

greater approaching the total sum of squares. The coefficient of determination

2
R is defined as the proportion of SS (Reg) to SS(Total). Thus the coefficient of

2
determination has a range of values from zero to one. R value close to 1 means
that the variation of the variable Y increasingly explained by its the linear
relationship with the independent variables.

2
R = SS(Reg) / SS(Total)

ˆ 2
2
= β [  X i - (  X i ) / n ] /  Y j – nY ( R13 )
2
2
2
1

5.1.4 COEFFICIENT TESTING

Investigation directly if there are significant changes in the variable Y by

changes in variable X, made by testing the coefficient  1. The magnitude of the

coefficient  1 is defined as the rate of change in the average value of Y by one

unit change in X. To test whether  1 is equal to zero or not, the test hypothesis

formulated

H 0 :  1 = 0

H 1 :  1  0 ( R14 )

~~* CHAPTER 5 LINEAR REGRESSION MODEL *~~