Statistics 3N03 - Exercise #1

Computer Labs: 2001-09-21

"Never on Sunday"

Every morning from 1993-11-16 to 1995-01-19, I recorded the price per litre of regular gasoline at Sarah's Sunoco on Main Street West at Newton Avenue and at the Petro Canada station on Main Street West at Longwood Road. Click here to get the data.

This exercise reviews all the graphical methods from Chapter 2 of Montgomery & Runger. Use MINITAB to plot the graphs. Paste the graphs into MS Word, then add captions and some discussion to make a short report on the data.

  1. Make a scatter plot of the daily Sunoco prices against the daily Petro Canada prices.
  2. Draw a lag 1 scatter plot of the Sunoco data; that is, plot today's price on the Y-axis against yesterday's price on the X-axis, for each day.
  3. Give a comparative box plot showing the Sunoco price against day of the week.
  4. If x(i) is the Sunoco price on day i, compute d(i) = x(i+1) - x(i), i.e. the change in price that occurred on day i. Draw a comparative box plot showing change in price against day of the week.
  5. Display the Sunoco change in price on a histogram.
  6. Display the Sunoco change in price on a stem-and-leaf diagram.
  7. Display the Sunoco price on a time sequence plot.
  8. Display the Sunoco change in price on a time sequence plot.
  9. Plot change in price against price, for the Sunoco data.
  10. Can you think of any other graphs that would help to interpret these data?
  11. Explain what you have learned from each graphical display. Why did I call this "Never on Sunday"?

Statistics 3N03 - Exercise #1 (continued)

Computer Labs: 2001-09-24, 2001-09-26 and 2001-09-28

"Never on Sunday" - Continued

After you have finished Exercise #1, enter the data into SPSS and re-draw the graphs in SPSS.

Next, enter the data into R and re-draw the graphs in R. See RHints to help you get started with this.

Finally, enter the data into Excel (or the spreadsheet of your choice) and re-draw as many of the graphs as you can. (Note that you can't easily do a boxplot, but you can get much the same effect by recoding the days of the week with numbers 1 to 7 and making an X-Y plot, with day number on the X-axis.)

Based on your experience so far, which do you prefer, MINITAB, SPSS or R? What are the advantages and disadvantages of using a spreadsheet rather than dedicated statistical software?

Statistics 3N03