Three-Sample Study (tap-only, bottle-only, tap-and-bottle)

a. Provide the null and research hypotheses (defining all symbols) for comparing the mean weight gain among the three groups.

b. Create an ANOVA table using weight gain as the dependent variable.

c. Using the information from the ANOVA table part (a). Conduct an overall F-test. What conclusions can be made?

d. In this example, is it appropriate to perform specific hypothesis tests to determine where group differences exist? If so, perform these tests and provide conclusions. If not, explain why the specific hypothesis tests cannot be performed.

4b,c,d The GLM Procedure Class Level Information Class Levels Values group 3 123 Number of Observations Read 979 Number of Observations Used 979LS-Means for group O 42 wtgain LS-Mean 41 O 40 - O 2 3 group4b,c,d The GLM Procedure Dependent Variable: wtgain Sum of Source DF Squares Mean Square F Value Pr > F Model 2 855.67896 427.83948 18.46 <.0001 error corrected total r-square coeff var root mse wtgain mean source df type i ss square f value pr> F group 2 855.6789616 427.8394808 18.46 <.0001 source df type iii ss mean square f value pr> F group 2 855.6789616 427.8394808 18.46 <.0001the glm procedure least squares means wtgain comparisons for group w differences alpha="0.05" not significant significantdistribution of f prob> F <.0001 co o wtgain group4b the glm procedure least squares means standard lsmean group error pr> It| Number 40.0107407 0.2929823 <.0001 least squares means for effect group pr> It| for HO: LSMean(i)=LSMean(j) Dependent Variable: wtgain i/j 2 W <.0001>