Stat 401 Distance Learning -- JMP Tutorials and Help Page

ANOVA using Fit Model (51 seconds 18.0MB .avi)

Video File

Shows how to use Analyze->Fit Model to create full ANOVA and t-test/TukeyHSD output. Also runs through the menus to ensure the data is properly formatted. Please read this document for additional discussion of the Fit Model procedure.

Linear Regression on two numeric variables using Fit Model and Fit Y by X (34 and 37 seconds and 6.5 to 7.5 MB .avi)

Video File - Fit Model

Video File - Fit Y by X

Shows how to use Analyze->Fit Model and Analyze->Fit Y by X to create full linear regression based ANOVA for two numeric variables.

Hypothesis Test on a mean (25 seconds 3.77MB .avi)

Video File
(posted 8-26-2005)

Shows how to use Analyze -> Distribution -> Test Mean to find t-test based p-values. Note that you should also understand how to hand compute the test statistic and find the p-value. Do not rely on JMP to do all your quantitative work as you may have to do some calculations on tests.

Matched Pairs Test (39 seconds 6.91MB .avi)

Video File
(posted 8-26-2005)

Shows how to use Analyze -> Matched Pairs to compare means across matched groups as well as generate the default 95% confidence interval. Note that the order of placement in the "Y, Paired Response" box will change the order of comparison. If you want to take the difference Hormone A - Hormone B, you must place Hormone B above Hormone A in the "Y, Paired Response" box.

Data tranformation - Differencing (23 seconds 3.44MB .avi)

Video File
(posted 9-7-2005)

Shows how to use the right-click accessed menu "Formula..." to create difference transformed values in a new column.

Data tranformation - Natural Log (Log base e is called Log in most statistical texts....) (21 seconds 3.19MB .avi)

Video File
(posted 9-7-2005)

Shows how to use the right-click accessed menu "Formula..." to create log transformed values in a new column.

Creating Residuals for Diagnostics using Fit Model (62 seconds 13.2 MB .avi)

Video File
(posted 10-13-2005)

How to create residuals from an ANOVA data set for future use in analyzing ANOVA assumptions. Summary: Analyze -->Fit Model; Save Columns --> Residuals; retrurn to the original data set and use the new column to generate a boxplot and NPP in Analyze --> Distribution. See this help document for more instructions.

Distance Computing Lecture 1 - August 23, 2005 (40 min)

Covers many basic concepts of JMP. Note that the videos above give the same content at a much higher resolution. Lecture slides containing contact information, homework guidelines, advice and more.

Distance Computing Lecture 2 - August 30, 2005 (32 min)

Extensive discussion of homework 2 and how to use JMP for the various problems. Note that the videos, "Hypothesis Test on a mean" and "Matched Pairs Test", above, show JMP work at a much higher resolution. Lecture 2 slides containing contact information and homework tips/advice. Note that for our purposes, we are often interested in randomized experiments since only through randomization can we draw inferences. See pages 6-9 of the text. Also see Dr. Caragea's Notes. Based on student questions, I created this small overview of confidence intervals and hypothesis testing.

No Taped Lecture - September 6, 2005

HW3 contains a fair amount of identification of Observational vs. Experimental Studies. Make certain to read Dr. Caragea's Notes. Chapter 1 of the text also discusses identification of observational vs. experiments (or randomized experiements). Remember, to generate t-statistics based on treatments, use Analyze -> Fit Y by X. Also, the matched pairs question is now on HW3, so watch the distance computing lecture 2 or see the matched pairs video above.

Distance Computing Lecture 3 - September 13, 2005 (35 min)

Extensive discussion of homeworks 2, 3 and 4. Discussion of JMP transformations, fit y by x for a two sample hypothesis test and analyze distribution for a single sample hypothesis test. Lecture 3 slides contain the concept of a randomization distribution, more discussion of std deviation vs. std error, hypothesis testing, confidence intervals, contact information and general homework tips/advice. Based on student questions, I created this small overview of confidence intervals and hypothesis testing. HW4 uses HBE.JMP.

Distance Lab "Lecture 4" - September 19, 2005 (~2hours)

Extensive discussion of homework 4 with comments about homeworks 2 and 3. Lecture 5 slides contain information about the log transformation, experiments vs. observations, the concept of a randomization distribution, more discussion of std deviation vs. std error, hypothesis testing, confidence intervals, contact information and general homework tips/advice. HW4 uses HBE.JMP. HW5 uses Mutant.JMP. See this document for information about a test mean. Notice that the "Wilcoxon Signed Rank" test is an option in the Test Mean Window! Also, see the Test Mean video from above (posted on 8-26-2005).

Distance Lab "Lecture 5" - September 26, 2005 (2+ hours)

Distance JMP session - September 26, 2005 (7 minutes)

HW5 files of interest. First, you will want to read how to use JMP for the Rank Sum and Signed Rank tests. Next, you will find the data sets Mutant.JMP and Sparrow.JMP. For question #3, use freeway.JMP. You should use the Wilcoxon Signed Rank test since you are interested in the pre-differenced column of data. Notes from the lecture about a permutation.

Distance Lab "Lecture 6" - October 10, 2005 (1.5 hours)

HW6 has no direct computer use. However, if you want to know how we generally use ANOVA in JMP, read this document on the Fit Model procedure. The document shows how to use Analyze->Fit Model to create full ANOVA and t-test/TukeyHSD output. Also, watch this video on the same topic.

Distance Lab "Lecture 7" - October 17, 2005 (1.8 hours)

Distance JMP session - October 17, 2005 (5 minutes)

HW7 has little direct computer use. Yet, to get the residuals for diagnositcs, you will use Fit Model. The Fit Model document shows how to use Analyze->Fit Model to create full ANOVA output as well as generate residuals for analysis using Analyze-> Distribution. Also, watch this video on the same topic. The topics of the lecture a many and somewhat varied. See the lecture notes for full details.

Distance Lab "Lecture 8" - October 24, 2005 (2 hours)

HW8 has little (or no) direct computer use. Most of the topics are covered in the notes. The Bonferroni Multiple Comparison and Tukey Comparison are heavily discussed in the lab. I suggest you carefully read pages 161 - 164 of your text. Also make certain you understand that we are looking at comparisons (read: differenced s.e.) --> the s.e. for a difference formula as outlined on page 163. Note that for question 2b, you do not use the differenced s.e. as you are simply looking at the CI for a single mean. See the lecture notes details. Also, see the Bonferroni notes from in class.

Distance Lab "Lecture 9" - October 31, 2005 (2 hours)

Distance JMP session - October 31, 2005 (6 minutes)

HW9 once again requires computer use. See the lecture notes for information about the appropriate regression formulas. Also, the lecture notes contain information about correlation and when to use correlation as well as a dicussion of interpolation vs. extrapolation.

JMP Notes for Regression and Multiple Comparisons

• Fit Y by X for:
  • Regression on two numeric variables
  • Multiple Comparions
  • Video (posted in August): Video File
• Fit Model for:
  • Regression on two numeric variables
  • Multiple Comparisons
  • Video (posted in August): Video File
• Data sets:
  • chicken.JMP
  • citytemp.JMP
• Creating Icons for Graphics:
  • Icon Creation in JMP

Distance Lab "Lecture 10 " - November 14, 2005 (1.5 hours)

The lecture and lecture notes cover various topics in multiple regression including interpretation of multiple regression parameters, CIs for the parameters, assessing assumptions, log based regression and more. See the document "Mutliple Regression in JMP" for aid in using JMP for the HW. Also, HW10 uses the datasets Bears (JMP) and BFChange (JMP).

Distance Lab "Lecture 11" - November 28, 2005 (1.5 hours)

Distance JMP session - November 28, 2005 (10 minutes)

HW11 requires extensive computer use. See the lecture notes for information about both indicator variables and stepwise modeling. Note that I augment the notes inside the lecture video, so be certain to watch my discussion of indicator variables. See the materials below for extra aid.

Cooks' D and Other Diagnostics (Extra Material)

• Cooks' D Document
• Anscombe2.JMP

Stepwise Modeling in JMP (HW11)

• Stepwise Help Document
• Mortality (JMP)
• Indicator Variables
• BFChange (JMP)

Factorial Designs and Blocking in JMP

• Blocking Help Document
• Research (JMP)
• FeederFloor (JMP)

• Blocking: In general, you need no special action to create blocks in JMP. Merely enter the variable as you would any normal factor. When interpreting the results, you are usually not concerned with the p-value and tests of the blocking variable.

Distance Lab "Lecture 12" - December 5, 2005 (2 hours)

HW12 is quite long and has many important points for the final exam. My lack of fit notes can be found here. Copies HW12 and the solutions.