JMP Introductory Guide
Introduction
JMP is a statistical analysis software package developed by the SAS Institute. JMP is
highly visual and provides us with a graphical interface to display and analyze data.
Because it provides us with a GUI, it may be more accessible to people without a
programming or statistics background. JMP organizes our data via a spreadsheet. From
the spreadsheet we can visualize and manipulate our data. JMP allows us to perform
complex analyses and create graphical displays of our data.
Creating a Data Table:
Above is the opening view of JMP. If we select New Data Table, JMP will open a blank
spreadsheet where we can type in our data. If we select Open Data Table, we can import
a data set into JMP.
Creating Histograms and Box Plots:
This Example is from Charles Cwiek, University of Tennesee
Select Distribution under the Analyze tab.
Highlight the variable you want to create a histogram of. Click the Y, Columns button.
Click OK.
This provides a Histogram and Boxplot. In order to change the options click the red
arrow next to your variable name. Highlight Histogram Options and select the
appropriate options.
Go to the Tools menu and select the Grabber tool:
Place the Grabber cursor on top of the histogram bars and click and drag. Dragging up
and down changes the width of the intervals, side to side changes the midpoints.
Creating ScatterPlots:
This Example is from Charles Cwiek, University of Tennesee
Under the Analyze tab, select Fit Y by X.
Highlight the variable you want to appear on the Y axis and click Y, Response. Highlight
the variable you want to appear on the X axis and click X, Factor. Click OK.
Example of Scatter Plot output.
Multiple Regression:
This example comes from James Abbey, Penn State.
Under the Analyze tab select Fit Model
Highlight the response variable and click the Y button. Highlight each predictor variable
and click the Add button. Click Run Model. This provides us with an ANOVA table and
Fit Statistics. Example output is shown below.
Residual Diagnostics:
Residuals can be obtained by clicking the red triangle next to your response variable in
your regression output. Highlight Save Columns and select Residuals. The Residual of
each observation will appear in your data sheet. You can also select Predicted Values.
This will allow you to create scatter plots to test your model assumptions. A QQ plot can
be obtained by creating a histogram of the residuals. Right click on the name of the Y-
Axis and select Normal Quantile Plot.A Confidence Interval for the expected response
of each observation can be obtained by selecting Mean Confidence Interval under the
Save Columns tab. The confidence intervals will be saved in your data table.
ANOVA:
This Example comes from James Abbey, Penn State.
Select Fit Model under the Analyze tab.
Highlight your response variable and click Y. Highlight factors you want to include in the
model and click Add. Click Run Model. Example output is shown below. You can click
the red arrow next to your response variable to save residuals and predicted values as in
the Regression example.
Logistic Regression:
Images from Kent State University. Instructions from James Frederick, UNC.
Select Fit Y by X under the Analyze tab.
Highlight your response variable and click Y, Response. Highlight your continuous
predictor variable and click X, Factor. Click OK. A Fit Nominal/Ordinal Logistic report
will appear containing parameter estimates and Fit Statistics.
References:
SAS Institiute. JMP User’s Guide.
http://www.jmp.com/support/downloads/pdf/jmp_user_guide.pdf. 4-30-2010.
Charles Cwiek, University of Tennesee. JMP Tutorial: Histogram.
http://web.utk.edu/~cwiek/201Tutorials/Histogram/. 4-30-2010.
James Abbey, Penn State University. Using JMP’s Fit Model command for ANOVA.
www.personal.psu.edu/jda188/JMP/FitModelAnova.doc. 4-30-2010.
James Frederick, University of North Carolina. JMP Logistic Regression.
http://www.uncp.edu/home/frederick/DSC510/JMPlogreg.htm. 4-30-2010.
Kent State Universtiy. Graphic: “Fit Y by X”.
http://www.library.kent.edu/images/jmp_pearsons_2u.jpg. 4-30-2010