Starting to use StatCrunch: A Step-by-step Introduction to the Basics

Note: A comprehensive help menu is built into StatCrunch 4.0, should you have any other questions.

Contents:

1) Running StatCrunch

2) General Tips

3) Loading a file to StatCrunch

4) Producing a stemplot with StatCrunch

5) Producing a scatterplot with StatCrunch

6) Calculating Correlation using StatCrunch

7) Adding a least-squares regression line to a scatterplot in StatCrunch

8) Producing a probability histogram with StatCrunch

 

Running StatCrunch:

1)  Go to the course homepage and click on StatCrunch (under Statistical Computing), then again on the StatCrunch icon or visit mnstats.morris.umn.edu/statcrunch/statcrunch4.0/ .

2)  You will have to enter a passcode which will be provided to you but the instructor. Enter the passcode and click on the submit button.

3)  Once you have logged in, you will be taken to a page with additional general information and help topics.  You should also see a spreadsheet in the center of the page where you can enter data; if this does not display, try using a different web browser.

 

General Tips:

1)  You can save graphs or results by clicking on "Options" and choosing "Export."  This will save your results under "My Results," so that you can refer back to it.

2)  Once a graph has been saved, you can select it from "My Results" and right-click (control-click) on it to save it as an image on the desktop of your computer.  This should produce an image file that can be renamed and submitted in your homework.

3)  Data can also be saved, though in a slightly different fashion.  Once the numbers have been entered into the spreadsheet, you can select "Data," then "Data table," then "Export."  You can choose columns of data to save, and have several other options.  The default set seems to be fine for most purposes.  Click "Export," and the data will be displayed below, where it can be selected and saved on your computer (or copied into a document).

4)  You can delete all entries in the data table/spreadsheet by clicking "Stat," then "Data table," and selecting "Clear."

 

Loading a file to StatCrunch:

1)  Log in to StatCrunch.

2)  Click "Data" and select "Load data" then "from www".

3)  Type http://mnstats.morris.umn.edu//Data/nameofthefile.xls. Do not forget to uncheck "Use first line as column names", if the data file designed that way. (be careful that "D" in "Data" should be in capital.)

4) Click OK

Now, try to loading the file WPOWER50

In step 3, you need to type http://mnstats.morris.umn.edu//Data/WPOWER50.xls

 

Producing a stemplot with StatCrunch: (NOTE: You can also do this manually, then type it up, especially if splitting or trimming is required, or if you want to produce a back to back stemplot.  StatCrunch may automatically trim or round your data!)

1)  Log in to StatCrunch.

2)  Enter the numbers you wish to display as a stem and leaf, or stemplot, in one of the columns of the spreadsheet.  You can rename the column as wished by clicking on the shaded box with the old name—e.g., "var1"—deleting it, and typing a new name.

3)  Click "Graphics" and select "Stem and Leaf" from the pull-down list.

4)  Select the column (or columns—each will become a separate plot) from the list.

5)  For a simple stem and leaf plot, the "Where" and "Group by" options can be left alone.

6)  Click on "Create Graph!"  Note that the high observations may be listed separately rather than as part of the stemplot.

7)  You can save the stemplot by clicking on "Options" and selecting "Export" from the small pop-up window displaying the stemplot.  This will be saved under "My results" if you want to refer back to it.  To include the stemplot in your homework, simply select the desired text and copy it, then paste into the answer form or document.

 

Producing a scatterplot with StatCrunch:

1)  Log into StatCrunch.

2)  Label your columns ("x" and "y" may help), and enter the data underneath.

3)  Go to "Graphics" and select "Scatter plot."

4)  Select the x and y coordinates from the drop-down menus.  The other options can be ignored or adjusted, as desired.

5)  Clicking on the "Next" button will allow you to specify several things.

               a)  Lines and Points—you can have lines connecting the points, just points, or just lines.

               b)  Axis labels and title—generally, it is a good idea to label your graphs thoroughly.  If you do not specify x- and y-axis labels, the names of the columns will be displayed.

               c)  Graph layout—this is useful if you are creating several graphs simultaneously and want to view them side by side; otherwise, leave the rows and columns per page at 1.  The default color settings are usually fine.

6)  Finally, display your graph by clicking on "Create Graph!"  Remember, you can export the graph (and thus save it) or modify the settings at any time by clicking on the "Options" button at the top.

7)  To put your graph in your homework, first save it as outlined above.  You can then right-click on the graph when it is displayed underneath the spreadsheet (select from under "My Results") and choose "Save Image to Desktop."  This will automatically create an image file, which you can re-name and include in your homework.

 

Calculating correlation using StatCrunch:

1)  Enter the data in two columns, as for a scatterplot.

2)  Select "Stat," then "Summary Stats," then "Correlation."

3)  Select both columns to be compared.

4)  Click "Calculate."

NOTE: Selecting more than two columns will produce a "Correlation matrix" where the correlations are displayed in table format, comparing each pair of data sets.

 

Adding a least-squares regression line to a scatterplot in StatCrunch:

1)  Produce a scatterplot in StatCrunch as directed.

2)  Select "Stat," then "Regression," followed by "Simple Linear."

3)  Choose your x and y variables, then click "Next" until you get to Graphics options, then select "Plot the fitted line" to produce a scatterplot with the least-squares regression line on it.

4)  Click "Calculate."  The first page will display the simple linear regression results, including the equation; clicking "Next" will display the graph.

 

Producing a probability histogram with StatCrunch:

1)  Log into StatCrunch.

2)  Enter the possible values of the random variable X (the sum of the two dice, in this case) in one column.

3)  Enter the probabilities (as decimals) associated with each set of dice in the next column, lined up with the appropriate value of X.

4)  Click on "Graphics," "Bar plot," "With summary."

5)  Choose the values of X (2, 3, 4, etc.) as the "Categories in" column, and the probabilities in the "Counts in" column.

6)  Hitting the "Next" button will produce options similar to other graphs; modify as desired, though most of them can be left alone.  As always, labeling the x- and y-axes and titling the graph is a good idea.

7)  The resulting graph can be saved as previously described (under scatterplot instructions).