S3N03 - Assignment #1

Statistics 3N03 - Assignment #1

2002-09-25

Due: 2002-10-02 18:00

The following problems and data sets are taken from Montgomery & Runger, Applied Statistics and Probability for Engineers, 2nd edition. Do your graphs and calculations in R. Submit your work as a report, pasting the graphs into a word processor and adding comments and discussion.

You don't have to type in all the data sets, those from Chapters 1-8 are available online.

2-32 (p 45)

Follow the instructions in the text and try a few more things, such as a lag-1 scatter plot, separate time series plots for each month, and comparative box plots to compare months (in chronological order across the graph). Try plotting log(sales) as well as sales; you can get a logarithmic Y-axis in R by adding the option log="y" to the plot() command. Ask someone who knows France why August sales are so different from other months. Give your forecast of monthly champagne sales for the year 1970.

If you are importing the online data file, be sure to check carefully for typos.

2-33 (p 46)

Follow the instructions in the text and also do a lag-1 scatter plot. Is there evidence of trend, a shift in mean or autocorrelation?

Wire Bond Pull Strength - Example 11-15 (p 542)

Do an exploratory data analysis to determine which of the four variables (wire length, die height, post height, loop length) has the greatest effect on pull strength. Use a scatterplot matrix to study relations between the variables and plot histograms to assess normality. Try box plots to see the effects of post height and loop length on pull strength. State your conclusions. Why might time series plots and lag plots be inappropriate for these data?

Hint: If the data are in a data frame called pull in R, then pairs(pull) will give a scatterplot matrix.

12-49 (p 613)

Do graphical analyses using comparative box plots to compare quality scores between the three preparation methods. Repeat on a log scale. State your conclusions. (The question asks for a test of hypothesis and an analysis of residuals but you are not expected to do those for this assignment.)

Hint: Review the notes on making comparative box plots. Note that in R you don't have to make a new column of log(score) to plot on a log scale, you can just add the option log="y" to the boxplot command to transform the Y-axis.

13-9 (p 640)

Do graphical analyses using comparative box plots to compare base current between the anneal conditions and between the two dopings. Give "interaction plots" like the one in Figure 13-10: plot the mean base current against anneal conditions, separately for each doping, and plot the mean base current against doping, separately for each anneal condition. State your conclusions. (The question asks for a two-factor analysis of variance but you are not expected to do that for this assignment.)

Hint: Enter the data as three columns in a data frame, putting the base current in the first column, a code for doping in the second, and a code for anneal condition in the third.