The results presented in the following tables are based on data from a study by Gruber (1970) to determine how and to what extent changes in blood pressure over time depend on initial blood pressure (at the beginning of the study), the sex of the individual, and the relative weight of the individual. Data were collected on n = 104 persons, for which the k = 7 independent variables were examined by multiple regression. The following variables were used in the study:
Y = SBPSL (estimated slope based on the straight-line regression of an individual's blood pressure over time)
X1 = SBP1 (initial systolic blood pressure)
X2 = SEX (male = 1, female = -1)
X3 = RW (relative weight)
X4 = X1X2, X5 = X1X3
X6 = X2X3, X7 = X1X2X3

Use only the ANOVA table to answer the following questions.
a. Determine the form of the separate fitted regression models of SBPSL (Y) on SBP1 (X1), RW (X3), and SBP1 × RW (X5) for each sex in terms of the estimated regression coefficients for the fitted regression model.
b. Why can't you use this ANOVA table to test whether the two regression equations are either parallel or coincident? Describe the appropriate testing procedure in each case.

Transcribed Image Text:

Variable Partial F(F/Sa) 34.987 0.643 0.149 0.145 0.084 0.017 0.125 XİSBPİ) Х,(SEX) X(RW) 0.045 0.695 0.027 -0.0029 -0.00018 -0.0092 0.00762 0.86644 0.07049 0.00762 0,00062 0.07049 0.00062 0.00022 Intercept) 4.667 MS 5.307 24.988 8.406/102 7.886 .057 .463/100 0.020 0.254 Source d.f. 5.307 37.148 24.988 7.886 Overall regression ". 76.246/96 66 24988 2883" Xi 7.886 80.520/101 1.057 х,х, 9.89 1.057 0.0200025 0.254 0314 2.844--# 3.613 Regression xxx 0.020 79.443/99 0.254 79.189/98 2.844 0,099 76.246 113.394 2.844 76345/97 0.099 0.794 0.099 76.246/96 0.125 Residual Total (corrected F,, 96, 095 2.11, F,.%, as,-3.95. 6. 102,095 96 103 3.94.