Hello, if you have any need, please feel free to consult us, this is my wechat: wx91due
STA 404
Homework 5 (Ch6 - 60 pts)
Due: 4/25
Please complete the following modified Chapter 6 exercises:
6.5 For part (a), use the raw data and a table of algebraic signs to manually (i.e. not using Minitab) estimate each of the following effects: a, b, ab, c, ac, bc, abc. Comment on which effects you expect to be significant.
Part (b) asks you to fit the full model containing three main effects, three two-way
interactions, and the single three-way interaction. In addition, provide the Pareto plot and the half-normal plot.
Part (c) asks that you write a regression model after obtaining the most parsimonious
model. For example, if you only retain main effects of A and B, plus their interaction, the model would look like:
Y = β0 + β1A + β2 B + β3AB + ε
You do not need to estimate the regression coefficients; just state the model.
Part (d) asks that you fit the reduced model from part (c), where all predictors are categorical variables, and investigate the residuals.
For part (e), provide an interaction plot, then state which levels (low or high) of factors A, B, and C you would recommend.
6.6 Note that if using the DOE menu in Minitab, the factors do not need to be explicitly treated as continuous covariates (fitting a GLM would require use of continuous covariates). It is not enough to simply provide contour plots and the AC surface plot: explanation is required. Convince me that you know what is happening in the plots.
6.32 For part (b), provide the ANOVA table for the most parsimonious model. You do not need to show ANOVA tables for the intermediate steps.
For part (e), provide a cube plot (use Minitab’s DOE menu; don’t sketch anything). Use an interaction plot to recommend how to maximize yield.
6.34 For part (c), analyze the data as if the design were a one-way ANOVA, where the single factor (Brownie Batch) has 8 levels and ni = 8 for i = 1, 2, … ,8. Is there a single batch that seems to be preferred?