cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
JMP is taking Discovery online, April 16 and 18. Register today and join us for interactive sessions featuring popular presentation topics, networking, and discussions with the experts.
Choose Language Hide Translation Bar

One way ANOVA comparing means.

I've had a look through the forums and can't find an answer to this, what I assume is a fairly easy question ( if someone has already answered it please just point me in the right direction..)

Is there a way to get JMP to give a t value for each level in a student t test and also q value in a comparison of all pairs using Tukey-Kramer HSD?.

hopefully that isn't to silly of a question.

thanks!
1 ACCEPTED SOLUTION

Accepted Solutions

Re: One way ANOVA comparing means.

As stated above:

 

 

Use the Fit Y by X platform. Specify your response and x variable.

Then on the red-traingle menu, select Compare Means > Each Pair, Student's t to do a t test of each pair. Select Compare Means > All Pairs, Tukey HSD to do a Tukey test on each pair. The t values and q values are part of the output. Remember, the t test doesn't correct for multiple comparisons, the Tukey test does. JMP has other multiple comparison procedures on that same menu. JMP 9 has all that plus nonparametric multiple comparisons.

Now that I read your post again, you may be asking about something else. If you are looking for JMP functions that will return quantiles or probabilties, then consider the following functions:

t distribution
t quantile
Tukey HSD Quantile
Tukey HSD P Value

They are found in the formula editor or in JSL.

Duane Hayes

View solution in original post

6 REPLIES 6

Re: One way ANOVA comparing means.

I worked it out. was easy enough.....

Re: One way ANOVA comparing means.

Use the Fit Y by X platform. Specify your response and x variable.

Then on the red-traingle menu, select Compare Means > Each Pair, Student's t to do a t test of each pair. Select Compare Means > All Pairs, Tukey HSD to do a Tukey test on each pair. The t values and q values are part of the output. Remember, the t test doesn't correct for multiple comparisons, the Tukey test does. JMP has other multiple comparison procedures on that same menu. JMP 9 has all that plus nonparametric multiple comparisons.

Now that I read your post again, you may be asking about something else. If you are looking for JMP functions that will return quantiles or probabilties, then consider the following functions:

t distribution
t quantile
Tukey HSD Quantile
Tukey HSD P Value

They are found in the formula editor or in JSL.

Re: One way ANOVA comparing means.

thanks for the reply much appreciated

Re: One way ANOVA comparing means.

As stated above:

 

 

Use the Fit Y by X platform. Specify your response and x variable.

Then on the red-traingle menu, select Compare Means > Each Pair, Student's t to do a t test of each pair. Select Compare Means > All Pairs, Tukey HSD to do a Tukey test on each pair. The t values and q values are part of the output. Remember, the t test doesn't correct for multiple comparisons, the Tukey test does. JMP has other multiple comparison procedures on that same menu. JMP 9 has all that plus nonparametric multiple comparisons.

Now that I read your post again, you may be asking about something else. If you are looking for JMP functions that will return quantiles or probabilties, then consider the following functions:

t distribution
t quantile
Tukey HSD Quantile
Tukey HSD P Value

They are found in the formula editor or in JSL.

Duane Hayes
mrtonyd
Level I

Re: One way ANOVA comparing means.

Sorry for resurrecting an old post, but I have a question on your answer.

 

I can create the fit x by y chart and it creates the graph I'd like. Then I can click the red triangle or right click the graph and add the tukey comparison.

 

My question is, if I have a hundred+ graphs, is there a way to add the tukey comparison to all the graphs? Or do I need to manually go through and add it to all of them?

 

Or is there a way to modify all the fit by x test results at once?

txnelson
Super User

Re: One way ANOVA comparing means.

I suggest that you read through the Tip of the Day, there it covers the ability to Broadcast the options by holding down the CNTL key when you make the Tukey Comparison selection, and it will be broadcast to all of the graphs that were created with Fit Y by X platform 

txnelson_0-1633027512798.png

 

Jim