Page 98 - Basic Statistics

93

b. On the F-statistic, F-statistics = 23.06 dan distributed F with the degrees of

freedom of numerator and denominator respectively 1 and 10, giving the
probability Pr (F> F-count) = 0.0007218. For significant level test  = 0.05, this

Pr < . The test results reject H 0. Even H 0 is rejected until a significant level of

0.0007219.

2
c. In the Multiple R-Squared, R = 0.6975. This means that 69.75% of the variation
in fuel consumption can be explained by its the linear relationship with the

level of car sales

d. In the coefficients part, presents each the standard deviasion of value

coefficients, value statistic t, and value of the probability Pr = [ P ( t  -

ˆ
tvalue) + P ( t  tvalue). For the coefficients testing β , given Pr = 0.0007. So
1
that H 0 is rejected at 0.01 significant level, even H 0 is rejected until a

significant level of 0.00071.

Worked Example 5.2 :

Observations on the results of a chemical reaction to temperature

variations are recorded as the following:

Temperature (Co) : 125 125 125 150 150 150 175 175 175 200 200 200

Reaction (Y%) : 77 76 78 84 84 83 88 88 89 94 94 95

If the relationship between the two variables above, the chemical reaction as

dependent variable (Y) and the temperature as independent variable (X), want
to be investigated by linear regression model,

a. Estimate the linear regression equation

b. Perform the significance test for the allegation regression equation through F

test

c. Check the accuracy of this model linier relationship Y with X by the
coefficient of determination R
2

~~* CHAPTER 5 LINEAR REGRESSION MODEL *~~