This problem involves data from Problem 15 in Chapter 5. Treat LN_BRNTL as the dependent variable and dosage level of toluene (PPM_TOLU) as a categorical predictor (four levels). The experimenter wanted to explore the possibility of using WEIGHT as a control variable.
a. State the appropriate ANACOVA model. Use the 50-ppm exposure group as the reference group when coding dummy variables.
b. Use a computer program to fit the model and provide estimated regression coefficients.
c. Provide adjusted means. Compare these with the unadjusted means. (Do not perform any statistical tests.)
d. Test whether the adjusted means are all equal. Use α = .05. State the null hypothesis in terms of regression coefficients.