MINITAB Notes for problems



MINITAB Notes for MATH 1342

Discussion of our homework, Chs. 1-8

Mary Parker, mparker@austincc.edu, 7/26/06

This handout includes notes about a few of the homework problems on the last year’s standard syllabus for MATH 1342. These are based on typical questions from students in the past. Since the types of problems we assign aren’t much different in the new edition, I expect that these comments will still be helpful.

Introduction to MINITAB:

Reading the data files: Choose File>Open Worksheet

Here are are two very specific relevant facts about all the datasets for our course.

1. They are all WORKSHEETS, and not Projects. So if you choose File>Open Project that won't work. Keep going down the list until you get to "Open Worksheet."

2. All the data files for our text are in Portable format.  When you first open the dialog box, it doesn’t list that as the default file type, so it looks as if there are no files.   You must change the file type in order to see them.   Notice the difference in the two screens below.  Notice the different file types listed.

|[pic] |[pic] |

Chapter 1

Time Plots: Students often spend too much time trying to get the actual dates along the horizontal axis. In this course that's not necessary. It's just fine to have the horizontal axis numbered 1 to 20 (or whatever number is relevant) to indicate the first time period, the second time period, etc.

Choosing a graph: Don't expect the software package to choose the sort of graph you need. The software package will make the graph, but you have to choose it. Think about what you learned in the chapter about choosing graphs for various types of datasets. Think about the information you want the graph to illustrate and choose a graph that does that.

Chapter 2 

Deleting an observation: Sometimes you may be asked to delete a data point and reanalyze the data.  To delete, just highlight the row and hit the Delete key. If you need the original dataset again, you can just re-load it. (If, for some reason, you save your revised dataset be sure to give it a different name. Otherwise you may over-write the original one and then have to reinstall the data on your hard drive to get the original data.)

Side-by-side boxplots or stemplots: For comparing two distributions. To do side-by-side boxplots, first you must get the two columns of data into the worksheet at the same time.   Open one worksheet and copy the values from the column (using the Edit menu.)   Then open the other worksheet and paste the values you just copied into an adjacent column.   Make sure that the labels for the columns are meaningful so that the labels on the graphs will be useful.

Chapter 3

In this course, we focus on doing normal distribution calculations by hand using the Normal Table rather than using software.

Chapter 4

What if the data isn't already in a file?: Sometimes you have to type in the data yourself. Just go to the worksheet portion of the screen and type in the numbers. Be sure to include labels too, so that you can keep track of what is in each column.

Scatterplots: In earlier versions of MINITAB, the command is called "Plot" rather than "Scatterplot." So you can just remember that Plot is fine.

What's the equivalent of "side-by-side" boxplots for scatterplots?: When we want to contrast the relationship between two variables on two different subgroups of a dataset, we can use different symbols for each of the two subgroups. Your MINITAB Manual shows you how to use these different symbols.

Chapter 5 

Identifying Influential points: asks about influential points.  Look at the Session Window output from Stat > Regression > Regression and see how  MINITAB identifies potential outliers and influential points. You'll need to run this on data that does have outliers and influential points in order to see this. Choose a dataset from the text that has these and use MINITAB to do that regression and then look at the output.

Putting two lines on one scatterplot: It is not worth your time to try to get MINITAB to put both lines on. Use the Stat>Regression>Fitted Line Plot to make one line. Then use MINITAB to do the numerical work for the other line and, by hand, graph it on your output.

If you use the menu item described in the MINITAB Manual to make residual plots, be sure to use the correct option. DO NOT just give all the possible residual plots. Since you haven’t learned to interpret most of those, it is not appropriate to include them.

Chapter 6. MINITAB is useful for producing a summary table of raw data. In the textbook, the data are already presented in these summary tables.

Summarizing raw data into a table: While problems in the textbook show a summary table for the data, please think about what the original data would have looked like. I haven’t seen the datasets for the book at the time I write this, but in the last edition of the text, the data for problem 6.4 was presented, so I believe it will be here too. Their data were from 386 subjects and they measured two categorical variables. So each observation has a value for the variable "major" and a value for the variable "sex." And there are 386 observations. Now look at the dataset file for this problem. Doesn’t it have 386 observations? If so, notice what they look like. Computer software, like MINITAB, is useful for taking a dataset like that and producing a summary table like the one in the text, using Stat > Tables > Cross Tabulation.   Various other options are available for that command, such as percentages. I recommend that you not pursue any of those now because most students find them confusing at this stage. (When you are looking at percentages, it is critical that you understand what was in the denominator and why. That's much easier if you computed the percentage yourself.)

Making graphs: (problem 4) Even if you used MINITAB to get the percentages in the conditional distributions, do the graphs by hand.

Chapter 7

Choosing a random sample: (problem 42) MINITAB makes it much easier to choose a reasonable-sized random sample than doing it by hand with a random number table. In the text for problem 42, they just tell you to find the first three in the sample from each group. But with MINITAB you can easily find the entire sample of 20 men and 20 women. There's no datafile for this, but you don't need one. You don't need the names of the women and men.Just put their ID numbers into the worksheet. So, in one column put the ID numbers from 1 to 290 for the men. (Calc>Make Patterned Data>Simple Set of Numbers) Using the same command, in another column enter ID numbers for the women. Then, for each of those columns, use Calc>Random Data>Sample from Columns to select the sample from that column and put the chosen sample IDs into another column. Because it is a hassle to look at an unsorted set of 20 numbers, then for each sample, use Data>Sort to put the numbers in numerical order. When you write your solution in your homework, be sure to write an explanation of what you did and what it means. Don't just list the numbers.

Notice that your random sample will differ each time you do it because, of course, they are random samples and it would defeat the whole purpose of random sampling if they were the same!

Chapter 8

No MINITAB assignment is given for this chapter. The only use for MINITAB for these topics is for the randomization and that was explained in Chapter 7.

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download