I would like to analyse a column with numerical/ordinal data in the range of 1-4. I can analyse the data using the x/y platform and all y is displayed by x variable values 1-4 in four analyses. However, I would like now to compare variable value 1 against all others (2-4) combined.
Despite intensive search in documentation and discussion forums, I have not found a way to do that and would appreciate any help. It seems that all the group by and merge features apply to variables, but not to actual values of a variable. Similarly, I would need a way e.g. to introduce age groups among ages in a column and analyse them separately by age stratum.
How can I group values of a variable for analysis ?
Will the interactive binning JMP Add In written by my colleague Jeff Perkinson accomplish your task? I think if you start with a continuous data type for your ranking, use the interactive binning, then change data type to Ordinal once done binning I think this will get you close?
Here's the link in the JMP User Community to Jeff's Add In:
Recode my be another option for you. Recode groups 2-4 to the same variable name and save in a new column. Based on what you have above stacking your data may also be an option.
First, thank you for your help!
Actually, recoding is the feature which I was looking for. It was however impossible for me to retrieve this information from any documentation or other source.
Interactive binning is nice, but obviously designed for continuous data and less suitable for ordinal/nominal data (although I got it working as well the way you suggested initially).
NB: as a user of the German language version, it is also many times impossible to actually apply the guidance found somewhere or in the documentation - if you find it at all. You have to switch your JMP to English first, then locate the feature you want to use, memorize it, switch back to German and then navigate again to the place. The German translation in the frontend is very cryptic to understand and the manual only available in English....
thank you for your comment on the German language support. I would like to talk to you about the cryptic German translation in more detail to understand your needs better. Then we can bring this to attention to our documentation group. We are always interested in improvig our software and interface in case there is something not obvious to find or misleading.
Please get in contact with me directly under email@example.com. You can write in German, please refer to this thread, so I'm able to recognize you as tz.1