Generating ordinal data

[#1]

Hello!

Does anyone know how I can generate ordinal data in JMP to simulate a sample?

I mean I need to simulate a sample including ordinal measurements by generating synthetic ordinal  data. 

Thank you in advance for your help. 

[#2]

What is the required distribution of the ordinal data?  Uniform???

Do you want the data to be whole numbers? (no decimals}

How many values do you want?

Do you want a random generation of values?

[#3]

Thank you for your response.
The options for answering the questions are:

• very important
• important
• moderately important
• of little importance
• not important

I want to replace the options by numbers 4, 3, 2, 1, and 0 when analyzing the data. I do not know the probability of each option being chosen by participants who will answer the survey.

[#4]

I like all the other replies. I have another one. You can set up a simulation with a column formula for a specific outcome. For example, suppose that I expect the following outcome:

I can set up a data table with a Response column using the Value Labels and the Formula column properties like this:

The data table with 1000 rows might look like this (top 15 rows):

And the distribution of the ordinal responses might look like this:

This data table is attached here.

[#5]

Thank you for your response.
The options for answering the questions are:

• very important
• important
• moderately important
• of little importance
• not important

I want to replace the options by numbers 4, 3, 2, 1, and 0 when analyzing the data. I do not know the probability of each option being chosen by participants who will answer the survey.

[#6]

Without an idea of what the distribution will be, your simulated data will not likely match reality. But, an easy way to generate some simulated data is to create a new column. Go to the Column Info for this new column. Under Initialize Data click on Missing/Empty and change it to Random. Select Random Integer and specify 0 for the minimum, 4 for the maximum. This will create a uniform distribution of those ordinal values. This is one easy way to simulate some data.

Note that since you are in the Column Info dialog, now would be the time to change the modeling type to Ordinal. You probably would also want to turn on the Value Labels option to specify 0 as Not Important, 1 as Of Little Importance, etc.