Page 169 - Quantitative Data Analysis

Quantitative Data Analysis
Simply Explained Using SPSS

is one the portion of total sum of square. Residual sum of square =
2
SS res=∑ (Y - Y') = 12.786. SS res the amount of variance in a data set
that is not defined by the regression model. The SS res = 12.786 is the
measurement of error remaining between regression function and
data set. Residual sum of square or ‘unexplained sum of square’ is
2
again the amount of total sum of squares (∑y ) that can be
explained in regression model.

Therefore it can be said defined that:
̅
2
2
∑y =∑ (Y - Y') + ∑ (Y' - ) 2
2
Or ∑y =SS res+SS reg
Or Total sum of square = Sum of square explained +
sum of square unexplained

F ratio in the problem #2 is 32.33 with 1 and 18 df (degree of
freedom). For example, the significance level α=0.05, it is found in
the F table with 1 and 18 df is 4.41. As obtained F is greater than
tabulated value, it is concluded that the regression Y on X is
statistically significant.

The Theory and Applications of Statistical Inferences 153