7.1 Repeat the analysis in Section 7.4 using the Education Expenditure Data in

Table 5.12.

IT'S FROM : regression-analysis-by-example-5th-edition BY Ali S. Hadi Samprit Chatterjee

\fYOU odd TUES . How mustering responses whe local uclinical . The reader is referred to ) Red tion usten I wood YOU where these methodical the wearing rename Her more plausible way ing to prior the meeting observations according to pri cedasticity in w considerable der sed in Chapter s as an example be analyze data on state a natural and memale recesion tures he EDUCATION EXPENDITURE DATA here on Expenditure data were used in Section 5.7 and it was sue There that these data be looked at across time ( the data are available for 1965 And 1975 ) to check on the stability of the coefficients . Here we use these demonstrate methods of dealing with heteroscedasticity in multiple regressi to analyze the effects of regional characteristics on the regression relationship the present analysis we shall work only with the 1975 data . The objective is the best representation of the relationship between expenditure on education other variables using data for all 50 states . The data are grouped in a natural geographic region . Our assumption is that , although the relationship is stru the same in each region , the coefficients and residual variances may diff gion to region . The different variances constitute a case of heteroscedastic an be treated directly in the analysis The data are presented in Table 7.2 and can be found at the book's The variable names and definitions appear in Table 7.3 . The model is Bot B 1 x 1 + B 2 X 2 + 8 3 X 3 + E The notion of pure error can also be used to obtain a test for lack of fit see , C .8 . Drap re cerro (1998) hapill ww waucegypt . edw / faculty madeRABES