Stat Testing and Data Tables<meta http-equiv="Content-Type" content="text/html; charset=UTF-8" />Hello All!
New JMP user here with a Tabulate question. I am trying to achieve a data table look that I have been familiar with up to this point which looks similar to the below:
| Group 1 | Group 2 |
|---|
| Base Size | 300 | 300 |
| A | B |
| Age | | |
18-24 | 10% | 25%A |
| 25-34 | 20% | 20% |
| 35-44 | 30% | 30% |
| 45-54 | 20% | 15% |
| 55+ | 20%B | 10% |
So far I am having luck developing the general look of a table like the above excluding the stat testing.
My question is, can I create a table like the above in JMP? If so, can someone point me towards a resource that goes through the steps?
Re: Stat Testing and Data TablesStan,
This is looking really nice. I will play around with this in a little while and provide any feedback.
Thanks again to both you guys for helping out with this.
Re: Stat Testing and Data TablesHello,
Select Analyze-->Consumer Research-->Categorical
From here select Click the structured tab option.
Drag your gender variable to "Top" green drop zone.
Similarly drag your goals variable to "Side" green drop zone.
Click OK.
This will produce the basic cross tabulation table (two-way tables in compact form) shown previously.
From here go to the red triangle under the Categorical Outline. Here you can specify the statistical tests you seek.
I have selected Cell Chisq and Compare Each Cell. To show the letters select the red triangle under Goals by Gender and select Show Letters.
I hope this helps.
Let me know if this is closer to what you wanted.
Stan
Re: Stat Testing and Data TablesOne more try -- hacking things together here a bit Hope this gives you some ideas.
Julian
Re: Stat Testing and Data TablesThat is what I was afraid of. Thanks much Julian!
Re: Stat Testing and Data TablesHi mcstagger,
My apologies for not interpreting correctly, you want that table exactly! Tabulate is powerful, but you won't be able to move around the sections to make exactly that table. Tabulate is also not built for doing analyses, only tabulating. If you wish to reveal the contingency test, you will need to use Analyze > Fit Y by X, or Analyze > Consumer Research > Categorical. Each of those platforms will make a table close to what you want, but not exactly I'm sorry to say. In case it helps, here is the set up for both of those platforms.
I'm sorry I don't have steps to make that exact table, but I hope this gets you a little bit closer.
Julian
Re: Stat Testing and Data TablesStan,
Thank you for helping out as well. Please see my latest response to Julian. What I am looking for is more consistent formatting to what I had initially posted, along with the steps, if possible, to show column stat testing.
Re: Stat Testing and Data TablesJulian,
What you have shown me is pretty much where I have been able to get. The pieces that I am looking for now are:
- Getting the column base size to show up above the column it represents, i.e. Male (n=227) and Female (n=251)
- Stat testing between column A (males) and B (females)
Re: Stat Testing and Data TablesHello,
Alternatively, if you are using JMP 11 you could use the Categorical platform under the Consumer Research menu.
Using the Children's Popularity data table and using the breakout and stub you suggested will produce the following:
Best,
Stan
Re: Stat Testing and Data TablesHi mcstagger
Of course. Here's a quick video of the steps I would go through to make that sort of table. Does this get you to where you want?
I hope this helps!
Julian
Re: Stat Testing and Data TablesJulian,
Could we use a sample data set provided with JMP?
Using the attached with "Gender" as our banner and "Goals" as our stub should work.
Thank you for the assistance Julian.
Re: Stat Testing and Data TablesHi mcstagger,
Using tabulate you can make a table very much like the one you have posted. The exact steps depend on how your original data table is structured. Would you be willing to post your data, or some sample data with the same format? If so, I would be happy to write out some steps.
Julian
