A QUICKER INTRODUCTION TO MINITAB

Reading Data Sets for this Course:

All of the data sets we will use here are of Minitab Portable (.MTP) type. This makes it easier to read them into Minitab. If you ever need to read a text file into Minitab, see “A Quick Introduction to Minitab”.

The data sets can be downloaded from the course website, and are also available at Stern on X:\SOR\CHURVICH\MBA\

If you are connected to the course website, and you have Minitab installed on your computer, click on ‘Data Sets’, and then on the name of the data set you want to read. Next, use File  Save As to save the file to your machine, with the extension .MTP. Next, start Minitab and use Open Worksheet as described below.

If you have already started Minitab and you want to read in a data set of Minitab Portable type, use File  Open Worksheet. Then give the name of the data file (for example, C:\Market.MTP).

Making Histograms:

To make a histogram of the data in column C1, use Graph  Histogram Simple. In the box for Graph Variables, enter C1 (either by typing it in, or by clicking on it in the list of variables).

Making Boxplots:

For a single boxplot, use Graph  Boxplot One Y, Simple. Enter the single variable name in the Graph Variables box.

If you want side-by-side boxplots for data stored in a single column but broken into groups according to the value of a separate grouping variable (for example, with values 0,…,0,1,…,1), use Graph  Boxplot  One Y, With Groups. Enter the name of the variable you want to plot and the name of the grouping variable.

For multiple boxplots of data stored in separate columns, use

Graph  Boxplot  Multiple Y’s, Simple. Enter the names of the variables you want to graph.

Making Scatterplots:

Graph  Scatterplotplot  Simple. Enter the Y and X variables in the first row of the dialog box. Once the plot is created, you can double-click any caption or label to change it. You can also double-click the dots to change their color and/or size, and you can double-click the purple outside region to change its color (using “fill pattern”).

Obtaining Summary Statistics:

Click on Stat  Basic Statistics  Display Descriptive Statistics. Enter the name of the variable you want to have summarized. If you need descriptive statistics by groups, click on the ‘By variables’ box and enter the grouping variable.

For the most complete summary, use Stat  Basic Statistics Graphical Summary. This gives a histogram with a superimposed normal curve (you can delete the curve by clicking on it and then hitting the delete key), a horizontal boxplot on the same scale as the histogram, and all the relevant numerical summaries. A small p-value in the Anderson-Darling Normality test indicates that the population is not normally distributed.

Probability Plot:

Click on Graph  Probability Plot  Simple. Enter the name of the variable for which you want the plot. The resulting plot is of “Percent” (the percentiles of a normal random variable with the same mean and standard deviation as the data set) versus the (percentiles of the) data. If the data came from a normal distribution, the plot should be relatively straight. The ideal straight line is plotted (along with two other curves that you can ignore), to help you to gauge how straight the plot is. Non-normality is indicated if the plot shows curvature. High outliers will tend to deviate from the line at the upper right, and low ones at the lower left. For a more formal indication of non-normality, check if the p-value in the box to the right of the plot is less than 0.05. The smaller this p-value, the stronger the indication of non-normality.