Stat 243 Lab 1 - Getting Started in Minitab (Descriptive Statistics and Histograms)

Background:

All labs in this course will involve the use of Minitab, a statistical software package installed on most of the computer labs around campus. This first assignment is an orientation lab to familiarize you with the software. In particular you will:

• Select a location where you are able to use Minitab (this is a statistical software package installed on most of the computer labs around campus including the Ackerman Lab, the Learning Center Lab, and the computer lab at the EOU Library).
• Successfully enter and manipulate data.
• Produce a lab report to be submitted via Canvas.

Getting Started:

In Windows, click the search box beside the Windows icon in the lower left corner. Type minitab and select the Minitab App. This should start the most recent version of the Minitab software which is version 19 (some computers may also have an older version 16 installed, but the instructions that follow assume you will be working with the current version).

Minitab should be divided into windows the window on the bottom is a worksheet window that looks like a spreadsheet. The window above that is called the session window and it is where output will appear. There is also a navigation window off to the left.

Lab 1 Directions:

The goal of this lab is to successfully acquire data from a web page, import the data into a Minitab document, use Minitab to graph and analyze the data, copy this analysis to a Word document, and finally to clean up the copied data into a neat report suitable for submission.

Follow these instructions:

Gathering the Data:

1. Open the web browser on your computer and go to the site http://lib.stat.cmu.edu/datasets/
2. Scan down to the bodyfat data set and click on it. This should take you to a page that gives a summary description of a data set, followed by references, and finally followed by a long list (252 rows) of data.
3. Read the summary information about the data set, paying close attention to the meaning of the first five columns. Use the mouse to highlight all 252 rows of data and right click to copy this information.
4. Open Minitab, and with the worksheet window active and with column C1, row 1 as the active cell, right click to paste the data you copied into the worksheet. Minitab will prompt you to use spaces as delimiters. Click OK.
5. Go back to the web page from which you obtained the data and identify which columns correspond to the variables for Weight and Height. In the Worksheet Window, enter appropriate names for those two columns. There are entered in the areas just above the data and below the column labels C1, C2, etc. (Be sure you are labeling the columns correctly. These are heights in inches and weights in pounds. Double check that the numbers listed in the columns make sense for these quantities).
6. We will not use columns C6 C15. Though it does not hurt to leave these columns there, it will make for a cleaner lab report to delete this unused data. In the worksheet window, click and drag to select columns C6-C15. Then right click and select Delete Cells.
7. We will only use the Weight and Height variables, so use a similar process to delete the other unused columns.

Producing Histograms:

8. It is now time to begin creating histograms. To make the first histogram follow the following procedure
   1. At the upper toolbar select Graph > Histogram > Simple and click OK. Select the weight variable by double-clicking on Weight in the variable list.
   2. Next, to determine the type of histogram, click on the button marked Scale and select Y-Scale Type from the header bar. Next click on Density and finally click OK to exit this sub-window.
   3. Now click OK to direct Minitab to create the histogram.
   4. Note that the y-scale is not in units of percent per pound, but instead lists a decimal that will need to be turned into a percent if a computation is to be done. Without clicking on the histogram, hover the cursor over the tallest bar in the histogram. After a few seconds, you should see the following information displayed on the graph Value = 0.0145503 Bin = 157.5, 172.5. This means that the bin width covers the weights 157.5 Lbs to 172.5 Lbs and the height of the bar is 1.45503 percent per pound. Using our usual sense of area as percentage we could then compute the percentage of data represented by this bar.
9. Right click on graph and select Edit Graph. From the resulting window, right click again and select Edit Bars. Pick Binning, select number of intervals and change this value to 25. Finally click OK.
10. QUESTION 1: What effect does this have on the histogram?
11. QUESTION 2: Using the techniques described above, what percentage of the weight data falls between 185 and 195 Lbs? (Round your answer to the nearest tenth of a percent).
12. Copy this histogram for weight into a Word document. To do this, click OK on the window showing the graph you edited. The modified graph should now show up in the Minitab session window. Right click on this histogram and select Copy Graph. Then right click on your word document and select paste.
13. Repeat the above steps to create a histogram for height. You do not need to change the number of intervals for this histogram, just leave it in the form that Minitab creates. Copy this histogram into a Word document as well.

Getting Descriptive Statistics:

14. From the top menu, select Stat > Basic Statistics > Display Descriptive Statistics. Select the Weight variable and click OK. What Minitab has produced in the session window is a collection of statistics about the weights in the data set. Among the things it has computed are N (the number of data points, which should be 252), the Mean (Average), the Median, and the StDev (Standard Deviation). Copy these descriptive statistics into your Word document.
15. Use a similar command to obtain the descriptive statistics for height. Paste these statistics into your Word document as well.

Editing the Data:

16. Each of the two histograms appears to have an outlier point. One person on the Weight graph is particularly heavy. One on the Height graph is particularly short. We would like to see what effect removing these two people has on the histograms and on the descriptive statistics.
    1. In the worksheet window, look at row 42 of the data. Notice that this person is particularly short.
    2. Click on the row number 42. Right click and select Delete Cells. That row of data should now have been removed from the worksheet.
    3. Next select row 39. The person corresponding to this row is unusually heavy. Delete this row as well, just as you did the other.
17. Produce two new histograms together with descriptive statistics for the modified data sets. We want some way to distinguish the modified data from the original data. When you have produced the histogram, right click to bring up the edit window and then click on the title (Histogram of Weight for example) and change it appropriately (for example, replace it with Histogram of Weight with Outliers Removed). Be sure to do this for both new histograms.
18. Obtain the descriptive statistics for the modified data sets as well, being sure to label them as data with outliers removed as well.
19. QUESTION 3: What effect did removing the outliers have on the statistic reported for N? Why?
20. QUESTION 4: What effect did removing the outlier have on the average Weight of the data set? Why? What effect did it have on the Standard Deviation of the Weights? Why? Which was more affected by the removal of an outlier, the Mean (Average) or the Median? Why?
21. QUESTION 5: What effect did removing the outlier have on the average Height of the data set? Why? What effect did it have on the Standard Deviation of the Heights? Why? Which was more affected by the removal of an outlier, the Mean (Average) or the Median? Why?

Report Guidelines:

1. Begin your lab report by describing what data sets were analyzed. Be sure to include the reference to the web page where you obtained the data. Be sure to include what quantities were being measured and in what units of measurement.
2. Include all four histograms in your report together with the four sets of descriptive statistics for the data set each histogram describes (all appropriately labeled).
3. For the data sets with the outliers removed, use the descriptive statistics to identify the heights and weights corresponding to two standard deviations above and below average. Report these values and explain whether or not the histograms seem consistent the 68/95 rule.
4. Be sure to answer all five questions asked in the lab report.
5. Make the report as organized and clear as possible and submit it via Canvas (either as a pdf, doc, docx, or odt file).